instruction stringlengths 34 186 | input stringlengths 2.02k 93.8k | output stringlengths 2 418 | meta_questiontype stringclasses 6 values | meta_inputlicense stringclasses 6 values | meta_pmid stringlengths 8 8 | meta_safetyreportid int64 9.51M 21M | meta_articlepubdate stringlengths 4 10 |
|---|---|---|---|---|---|---|---|
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Cachexia'. | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | DENOSUMAB, NIVOLUMAB, PACLITAXEL, PALBOCICLIB, ZOLEDRONIC ACID | DrugsGivenReaction | CC BY | 33191899 | 17,820,365 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Death'. | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | DENOSUMAB, NIVOLUMAB, PACLITAXEL, PALBOCICLIB, ZOLEDRONIC ACID | DrugsGivenReaction | CC BY | 33191899 | 17,820,365 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Jaundice cholestatic'. | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | DENOSUMAB, NIVOLUMAB, PACLITAXEL, PALBOCICLIB, ZOLEDRONIC ACID | DrugsGivenReaction | CC BY | 33191899 | 17,820,365 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Osteomalacia'. | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | DENOSUMAB, NIVOLUMAB, PACLITAXEL, PALBOCICLIB, ZOLEDRONIC ACID | DrugsGivenReaction | CC BY | 33191899 | 17,820,365 | 2021-02 |
What was the administration route of drug 'CALCIUM CITRATE'? | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | Oral | DrugAdministrationRoute | CC BY | 33191899 | 19,139,574 | 2021-02 |
What was the administration route of drug 'CALCIUM GLUCONATE'? | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | Intravenous (not otherwise specified) | DrugAdministrationRoute | CC BY | 33191899 | 19,139,574 | 2021-02 |
What was the administration route of drug 'SODIUM PHOSPHATE'? | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | Intravenous (not otherwise specified) | DrugAdministrationRoute | CC BY | 33191899 | 19,139,574 | 2021-02 |
What was the dosage of drug 'CALCIUM CITRATE'? | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | 3.8 G (+4 DAYS) | DrugDosageText | CC BY | 33191899 | 19,139,574 | 2021-02 |
What was the outcome of reaction 'Blood parathyroid hormone decreased'? | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | Recovering | ReactionOutcome | CC BY | 33191899 | 19,139,574 | 2021-02 |
What was the outcome of reaction 'Blood phosphorus increased'? | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | Recovering | ReactionOutcome | CC BY | 33191899 | 19,139,574 | 2021-02 |
What was the outcome of reaction 'Death'? | Hypophosphatemia and FGF23 tumor-induced osteomalacia in two cases of metastatic breast cancer.
Tumor-induced osteomalacia (TIO) is a rare paraneoplastic syndrome characterized by factor-induced dysregulation of phosphate and vitamin D metabolism resulting in alterations in bone formation, leading to bone pain and fractures. While the true incidence is likely underestimated, less than 500 cases of TIO have been reported since initial description in 1947. TIO cases have classically been associated with mesenchymal tumors of bone and soft tissue, but have also rarely been linked to malignant tumors, with scant reports implicating non-mesenchymal tumors. TIO is mediated through inappropriate tumor overproduction of fibroblast growth factor 23 (FGF23). Increased FGF23 secretion leads to hypophosphatemia by (1) reduced phosphate reabsorption via activation of the proximal renal tubular epithelial cells to internalize sodium phosphate cotransporters and (2) reduced activation of vitamin D3 via inhibition of the renal enzyme 1-α hydroxylase. Low circulating levels of active vitamin D lead to reduced intestinal phosphate absorption and impaired mineralization of osteoid matrix. TIO in breast cancer poses a distinct diagnostic challenge due to the common adjunct oncologic management with bone protection therapy such as denosumab or bisphosphonates. These agents can be culprits of hypophosphatemia and hypocalcemia, rendering timely diagnosis of TIO difficult. Delay of diagnosis of TIO can result in worsening functional status, and early morbidity and mortality. To date, there has been one prior case report of TIO in breast cancer, and herein we describe two additional cases of TIO in this setting.
pmcIntroduction
Fibroblast growth factor-23 (FGF23) is a phosphaturic humoral factor produced by osteoblasts and osteocytes [1]. First identified two decades ago, mutations in the cleavage of FGF23 cause several inherited renal phosphate wasting diseases leading to rickets in children or osteomalacia in adults [2, 3]. In the paraneoplastic setting, FGF23 oversecretion leads to tumor-induced rickets/osteomalacia (TIO) also known as oncogenic osteomalacia [4]. TIO is typically reported with mesenchymal tumors [5, 6], and is starting to become recognized in patients with liquid [7] and solid organ malignancies [8, 9, 10] as well.
FGF23 is a key regulator of phosphate metabolism. The primary physiologic function is to lower serum phosphate levels which is mediated by FGF receptors (FGFR) and klotho complexes [3]. FGF23 downregulates the expression of cotransporters in the kidney that are essential for the reabsorption of phosphate. Additionally, FGF23 downregulates the expression of enzymes that activate vitamin D which increases intestinal phosphate absorption, thereby indirectly lowering serum phosphate levels [11].
Phosphate is primarily found in bone and is responsible for skeletal strength and rigidity. Low phosphate levels manifest as general muscle weakness, fatigue, and in extreme cases impaired cardiac and respiratory function [12]. These symptoms, in patients with cancer, may be attributed to their malignancy, and the potential diagnosis of TIO may be overlooked, especially with the rarer non-mesenchymal origin tumors. Below are examples of two case reports of patients with metastatic breast cancer with severe hypophosphatemia, phosphaturia and elevated serum FGF23, consistent with TIO. To the best of our knowledge, there is only one other case report of TIO associated with metastatic breast cancer [13]. These cases are particularly challenging given the use of antiresorptive therapy in patients with bone metastasis which can trigger FGF23 overexpression [13] and worsen underlying oncologic osteomalacia.
Case 1
A 47-year-old woman with metastatic breast cancer with liver and bone involvement was referred to the nephrology clinic for persistent hypophosphatemia. Seven years ago patient was diagnosed with left mammary duct carcinoma and underwent partial mastectomy followed by chemotherapy with paclitaxel and tamoxifen. She had a reoccurrence 3 years later and failed multiple lines of chemotherapy including eribulin and vinorelbine with last positron emission tomography (PET) scan showing metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac (Figure 1).
The patient was initiated on monthly denosumab for 1 year (12 doses in total) prior to the current nephrology visit, with last dose 1 month ago, to address metastatic bone involvement. Phosphorous level on consultation was < 0.9 mg/dL (2.5 – 4.5 mg/dL) with no prior levels. Remainder of bloodwork is shown in Table 1 which highlights low calcium 7.4 mg/dL (8.5 – 10.5 mg/dL) and elevated alkaline phosphatase (ALP) of 738 U/L (≤ 130 U/L). The fractional excretion of phosphate (FePhos) in the urine was elevated at 56% (< 5 – 10%).
Etiology for hypophosphatemia was initially thought to be secondary hyperparathyroidism given elevated parathyroid hormone (PTH) of 488 pg/mL (12 – 88 pg/mL) due to hypocalcemia in the setting of recent denosumab administration. Phosphorous levels remained low despite oral calcium and phosphate repletion and oral calcitriol administration (Table 1). Given persistent hypophosphatemia, FGF23 was checked, and levels returned strikingly elevated at 2,430 RU/mL (≤ 180 RU/mL) suggesting an FGF23 secreting tumor as the most likely cause for severe hypophosphatemia. Unfortunately, the patient passed away within 1 month due to disease progression.
Case 2
A 55-year-old woman with triple negative invasive ductal breast cancer, who achieved remission 10 years ago presented with progressive weakness. She was found to have relapsed disease involving the liver, lung, and bone (vertebral, acetabulum, and ilium) 1 year ago (Figure 2), and subsequently received chemotherapy including palbociclib, nivolumab, and abraxane as well as 4 monthly doses of zoledronate, followed by 10 monthly treatments of denosumab. She last received bone-stimulating therapy and chemotherapy 3 months prior to admission. She had no other comorbidities, nor a history of additional medications or herbal supplements. She was a lifetime nonsmoker. She was admitted for obstructive jaundice due to progression of disease. During the course of her admission, she complained of severe lower extremity bone pain limiting ambulation. Prior to admission, the patient’s electrolytes were within normal limits.
Upon admission, she was cachectic (body mass index < 18), with hypophosphatemia of 1.6 mmol/L (2.5 – 4.5 mmol/L). Nephrology was called for further evaluation. Remainder of lab studies are shown in Table 2 and include a normal corrected calcium of 9.5 mmol/L (8.5 – 10.5 mmol/L), low 25-hydroxyvitamin D of 15 ng/dL (20 – 50 ng/dL), elevated PTH of 287.3 pg/mL (12 – 88 pg/mL), and elevated ALP of 635 U/L (≤ 130U/L). FePhos was 78% (< 5 – 10%), consistent with phosphate wasting. Of note, 1,25-dihydroxyvitamin D was elevated at 83 pg/mL (20 – 50 pg/mL) despite not being on calcitriol.
Given elevated urine phosphate, an oncologic osteomalacia was suspected and FGF23 was checked and was elevated at 548 (< 180) RU/mL. Due to aggressive supplementation, serum phosphate increased to a peak value of 3.8 mmol/L; PTH decreased to 44, but FGF23 and FePhos remained elevated at 424 and 72%, respectively. The patient continued to decline and passed away within 2 weeks.
Discussion
FGF23 is a glycoprotein part of the FGF family which is subdivided into 7 subfamilies with 22 members reported in humans [14]. FGF23 belongs to the FGF19 subfamily which has also been called the endocrine FGFs due to the inner protein structure allowing it to function as a circulating hormone [15]. FGF23 is derived from bones, and under physiologic conditions, its production is stimulated by extracellular phosphate. Once secreted from osteoblasts and osteocytes, FGF23 plays a pleiotropic role which links the bone with several organ systems including the kidney, heart, and cells part of the immune system [1]. FGF23 signaling contributes to regulation in cellular proliferation, survival, and differentiation making it an attractive pathway to hijack by cancer cells [16].
FGF23 renal pathophysiology
With respect to the kidney, the main function of FGF23 is to lower serum phosphate levels as shown in Figure 3. This is established through direct inhibition of phosphate reabsorption at the level of the proximal tubular cells, and indirectly by downregulation of enzymes necessary to activate vitamin D. Direct actions involve the binding of circulating FGF23 to FGF receptors (FGFRs) and coreceptor klotho on the basolateral surface of the proximal tubular cells. This results in decreased expression of two sodium-phosphate cotransporters called NaPi-2a and NaPi-2c. These transporters, located on the apical surface of the proximal tubular cell are responsible for renal phosphate reabsorption. Decreased expression of NaPi-2a and NaPi-2c is therefore a direct cause of phosphaturia [17].
FGF23 also indirectly lowers serum phosphate levels by inhibiting renal 1-α-hydroxylase which is necessary to activate vitamin D. Further, FGF23 also increases the expression of 24-hydroxylase which degrades the active form of vitamin D into inactive metabolites. These actions collectively reduce active levels of vitamin D leading to decreased intestinal reabsorption of phosphate [18]. This relationship has been demonstrated in animal studies where a single injection of recombinant FGF23 resulted in reduction of serum phosphate and 1,25 (OH) 2D levels independent of PTH levels [11]. During the experiment, PTH levels remained low, and the hypophosphatemia was reproduced by injection of FGF23 in parathyroidectomized rats [11].
FGF23 mode of inheritance
Both genetic and acquired mechanisms of FGF23-related hypophosphatemic disease have been described. Genetic mechanisms vary by mode of inheritance. Autosomal dominant hypophosphatemic rickets (ADHR) is caused by mutations in FGF23 gene [2]. The autosomal recessive variant is caused by mutations in dentrin matrix protein 1 (DMP1) [19]. The X-linked dominant form occurs due to mutations in phosphate-regulating gene (PHEX) [20].
An acquired FGF23 hypophosphatemic disease is associated with the administration of intravenous iron, specifically the saccharated ferric oxide and iron polymaltose. Evaluation of these patients showed elevated FGF23 levels with the exact mechanism not known [21]. TIO is another example of an acquired form of FGF23 hypophosphatemic disease [17] which is reviewed in greater detail below.
Tumor-induced osteomalacia
TIO is a rare paraneoplastic disease, first described in 1947 by Robert McCance who reported a patient with pain and weakness in the setting of low phosphate levels. His symptoms persisted despite being treated with vitamin D, and eventually improved only after a tumor found in the femur bone was resected [22]. Animal experiments have supported the presence of the humoral factor leading to hypophosphatemia [23]. The earliest evidence to support this in humans was done by Miyauchi et al. [24] where tumor removal in a patient with osteomalacia and injection into healthy mice lead to hypophosphatemia.
Tumors associated with TIO are usually mesenchymal in origin [17]. Within the reported cases of TIO, 40% occur in the bone and 55% occur in soft tissues. The thigh and femur are the most common sites of involvement with the pelvis reported in only 8% of cases, and only 2% of cases reported as involving more than one site [25]. These tumors can be histologically polymorphous, but in 1991 Weidner [26] proposed a classification system to divide them into four morphologic patterns including phosphaturic mesenchymal tumor mixed connective tissue variant (PMTMCT), osteoblastoma-like variant, non-ossifying fibroma-like variant, and ossifying fibroma-like variant. PMTMCT comprises 70 – 80% of cases of TIO and typically begins in bone or soft tissues [5].
Non-mesenchymal tumors with TIO manifestations are now being recognized and reported in leukemia [7], B cell non-Hodgkin’s lymphoma [8], sarcoma [10], and other solid organ cancers including lung [27], prostate [28], and colon cancer [29]. There is only one case of TIO reported in metastatic breast cancer [13] with the two cases above resulting in a total of three. During malignancy, abnormal FGF signaling has been shown to induce cell proliferation and angiogenesis thereby promoting metastasis [16]. In breast cancer specifically, molecular alternations in FGFR1 and FGFR2 receptors are the most common reported [16]. Clinical trials support this data where phase I trials showed hyperphosphatemia as the most common adverse effect when novel tyrosine kinase inhibitors targeted FGF signaling [30].
Diagnostic evaluation of TIO should start with a comprehensive metabolic panel to check serum phosphorous and calcium levels which are typically low. Alkaline phosphatase may be elevated as in case 1 (738 U/L) and case 2 (635 U/L) due to osteoblast hyperactivity. Vitamin D levels should be checked and are typically low due to the inhibitory effect of FGF23. This was seen in our cases where vitamin D levels were 8 ng/mL and 15 ng/mL in case 1 and 2, respectively. PTH levels may be variable and increased at times as part of a normal feedback response to low vitamin D levels and subsequently hypocalcemia. In both cases, the elevation in PTH (488 pg/mL and 287 pg/mL) was likely multifactorial; initially as a feedback to hypocalcemia in the setting of denosumab. Secondary hyperparathyroidism has been demonstrated in patients receiving denosumab as a result of prolonged hypocalcemia caused by this drug [31], leading to renal phosphate wasting in some patients. This mechanism may have contributed to pathogenesis of hypophosphatemia in our patients. However in case 1, phosphorus remained low despite aggressive supplementation. Persistent hypophosphatemia however should also raise concern for an FGF23 secreting tumor. For case 2, denosumab was given 3 months prior to recognition of hypophosphatemia. Furthermore, FGF23 remained elevated, and phosphaturia continued despite PTH normalization. Therefore, denosumab likely did not play a major role in the FGF23 elevation or renal phosphate wasting. Along with serum FGF23, urine studies including urine creatinine and urine phosphorous must be checked to calculate the fractional excretion of phosphate and tubular reabsorption of phosphate. In the setting of TIO, one would expect a high fractional excretion of phosphate (> 10%) and low tubular reabsorption of phosphate (< 75%) due to inhibition of sodium phosphate transporters at the proximal tubules and low vitamin D. Dihydroxyvitamin D-1,25 was low in case 1 as expected due to suppressed activation by FGF23. However, in case 2, dihydroxyvitamin D-1,25 was elevated in the absence of calcitriol. Although in patients with chronic kidney disease and hyperphosphatemia FGF23 is elevated leading to suppression of vitamin D 1,25 production, we hypothesize that perhaps in some patients with hypophosphatemia, other mechanisms may be responsible for higher vitamin D 1,25 levels to counteract effects of low phosphorus levels.
Several imaging modalities can be used to identify the tumor, including magnetic resonance imaging (MRI) and PET scan. Somatostatin receptors (SSTR) based functioning imaging can also be performed since some of these tumors express SSTRs [32]. However, clinicians have to be mindful that inflammatory reactions can cause a false positive SSTR imaging [32]. In cases where tumor is identified, the treatment of choice is resection. Once FGF23 levels decline in circulation, serum phosphate levels return to normal, as early as five days post operatively [33]. In cases where the tumor is inoperable, medical management may be attempted with phosphate supplementation and calcitriol as recommended in our cases of metastatic disease. Octreotide is another potential treatment, given link with SSTR. Targeted antibodies against FGF23 have shown promise in animal models [34].
Conclusion
TIO can be a challenging diagnosis to make, especially in patients with malignancy other than mesenchymal origin, as symptoms of hypophosphatemia are nonspecific and could be easily attributed to the underlying cancer. In fact, the average time from recognition of osteomalacia to identifying the associated tumor is ~ 5 years [35]. We recommend more frequent testing of serum phosphorous since it is not part of the routine basic metabolic panel. Furthermore, in breast cancer specifically, patients are frequently managed with bone-targeted therapy such as bisphosphonates and denosumab which can further exacerbate hypophosphatemia. Antiresorptive therapy during malignancies should be carefully weighed with degree of hypophosphatemia and risk of skeletal-related events. Patients with TIO should be evaluated for resection, which can be curative when involving a solitary lesion. It is reasonable to check FGF23 levels in oncologic patients with persistent hypophosphatemia despite adequate supplementation of phosphorus and vitamin D and discontinuation of the drugs known to cause renal phosphate wasting. In patients with several lesions or metastatic cancer such as described above, systemic oncologic therapy and supplementation of phosphorous, calcium, and vitamin D can be attempted to improve the quality of life.
Funding
This research was supported by National Institute of Health grant award P30CA008748.
Conflict of interest
Ilya Glezerman owns Pfizer Stock. Remaining authors have nothing to disclose.
Figure 1 PET scan showing progression of disease for case 1. Metastasis to the liver, sternum, and sclerotic osseous lesions to the spine and right iliac.
Table 1. Case 1. Sequence of laboratory findings and treatment for hypophosphatemia.
–12 months to –1 month –10 days –4 days Nephrology
consult (day 0) +10 days
Treatment
Denosumab (mg) 120 mg/monthly × 10 doses
Potassium-phosphate/sodium-phosphate (mg) 250-45-298
t.i.d. 250-45-298
t.i.d.
Calcitriol (mcg) 0.25 b.i.d. 0.25 b.i.d.
Laboratory studies
Serum phosphate (mg/dL) < 0.9 1 1.1
Serum calcium* (mg/dL) Range 8.7 – 10.5 8.1 8.4 7.9 9.1
Alkaline phosphatase (U/L) Range 97 – 506 504 690 738 619
Serum PTH (pg/mL) 488
Serum FGF23 (RU/mL) 2,430
Serum 25-OH Vit D (ng/mL) 8
Urine sodium (mEq/L) 22
Urine calcium (mg/dL) < 1
Urine phosphate (mg/dL) 214
Urine creatinine (mg/dL) 229
FePhos** 56%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Table 2. Case 2. Sequence of laboratory findings and treatment for hypophosphatemia,
–12 months to –3 months –10 days –4 days Nephrology
consult (day 0) +3 days +4 days
Treatment
Denosumab (mg) 120 mg/monthly
× 10 doses
Potassium-phosphate/Sodium-phosphate (mg) 250-45-298 once 250-45-298
TID 250-45-298
QID
IV Phosphate (mmol) 30 30 15
PO calcium citrate (g) 3.8
IV calcium gluconate (g) 4 2
Laboratory studies
Serum phosphate (mg/dL) 2.6 (month –3) 1.6 1.4 1.4 3.8 2.4
*Serum calcium (mg/dL) 9.2 – 10.1 range 9.0 8.7 8.0 9.4 9.3
Alkaline phosphatase (U/L) 138 – 253 range 516 581 712 677 664
Serum PTH (pg/mL) 287.3 44.3
Serum FGF23 (RU/mL) 548 424
Serum 25-OH Vit D (ng/mL) 15
Serum 1,25-Dihydroxyvitamin D (pg/mL) 82
Urine sodium (mEq/L) < 20
Urine calcium (mg/dL) 3.1
Urine phosphate (mg/dL) 175 416
Urine creatinine (mg/dL) 80 99
**FePhos 78% 72%
*Corrected calcium = total calcium (mg/dL) + 0.8 (4.0-serum albumin [g/dL]), where 4.0 represents the average albumin level. **FePhos = (urine phosphorus/serum phosphorus) × (serum creatinine/urine creatinine). PTH = parathyroid hormone; FGF23 = fibroblast growth factor 23; FePhos = fractional excretion of phosphorus.
Figure 2 PET Scan showing progression of disease for case 2. Metastasis to the liver, right acetabulum, thoracic vertebrae, and right ilium.
Figure 3 Bone-kidney axis and phosphaturic effects of FGF23. FGF23 is produced in bone by osteocytes in response to high serum phosphorous. In malignant bone, FGF23 is produced regardless of serum phosphorous. One of FGF23 targets is the kidney. FGF23 binds to FGR receptors and complexes with klotho on the basolateral surface of proximal tubular cells. This causes a decrease in expression of sodium-phosphorus co-transporters (Na-PO42-) whose role is renal phosphate reabsorption. Indirect effects include inhibition of 1-α-hydroxylase levels which are necessary to activate vitamin D and increased expression of 24-hydroxylase which degrades active vitamin D. The net effect is a decrease in serum phosphorous. | Fatal | ReactionOutcome | CC BY | 33191899 | 17,820,365 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Anuria'. | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | LEVETIRACETAM | DrugsGivenReaction | CC BY-NC-ND | 33194552 | 19,598,806 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Hydronephrosis'. | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | LEVETIRACETAM, SODIUM ACETATE, SODIUM CHLORIDE | DrugsGivenReaction | CC BY-NC-ND | 33194552 | 19,605,126 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Nephrolithiasis'. | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | LEVETIRACETAM, SODIUM ACETATE, SODIUM CHLORIDE | DrugsGivenReaction | CC BY-NC-ND | 33194552 | 19,506,348 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Pyrexia'. | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | LEVETIRACETAM | DrugsGivenReaction | CC BY-NC-ND | 33194552 | 19,598,806 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Ureterolithiasis'. | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | LEVETIRACETAM | DrugsGivenReaction | CC BY-NC-ND | 33194552 | 19,598,806 | 2021-01 |
What was the outcome of reaction 'Hydronephrosis'? | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | Recovered | ReactionOutcome | CC BY-NC-ND | 33194552 | 19,605,126 | 2021-01 |
What was the outcome of reaction 'Nephrolithiasis'? | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | Recovered | ReactionOutcome | CC BY-NC-ND | 33194552 | 19,506,348 | 2021-01 |
What was the outcome of reaction 'Ureterolithiasis'? | Hypertonic saline-induced urolithiasis presenting as acute renal failure in a child with traumatic brain injury: A case report.
We report the case of a 4 year old female with severe traumatic brain injury who developed bilateral obstructing ureteral stones after hypertonic saline treatment. She developed calcium phosphate stones after two weeks of hypertonic saline therapy, and was successfully treated with ureteral stents and ureteroscopy. She has remained stone-free since that time. We postulate that an incomplete type 1 renal tubular acidosis made her intolerant to the acid and sodium load of the saline, and discuss other lithogenic factors of her presentation.
Introduction
The incidence of nephrolithiasis is increasing, with pediatric disease rising 6–10% annually over the last 25 years.1 Children may not present with classic stone symptoms due to poor communication abilities, and a higher index of suspicion may be required to diagnose them. Accurate 24-h urine collection is difficult, and they are lost to adult nephrologists and urologist as they age. The majority of stone-forming children have an underlying metabolic or infectious etiology, and a family history and bilateral stone burden are commonly seen in those with metabolic derangements.1,2 Immobility, medications, and other etiologies may precipitate stones or unmask underlying abnormalities, as evidenced by the following case.
Case presentation
A previously healthy 4 year old female presented to our hospital after sustaining a traumatic brain injury in a high-energy motor vehicle collision. A head CT scan revealed intracranial hemorrhage, for which she underwent an emergent craniotomy and external ventricular drain placement. Her workup included a CT abdomen and pelvis, which demonstrated normal kidneys and bladder without stones or hydronephrosis. Her family history was significant for nephrolithiasis in her mother and maternal grandmother.
She received 5 days of 3% hypertonic saline (HTS) as treatment for cerebral edema, with serum sodium in the 150's and a peak of 161mmol/L. She then received 6 days of sodium acetate, during which time her urine pH was 9. HTS was then restarted at 7%, with urinary output of 2–3.5ml/kg/hr over the following days with no urinary pH measurements. She received levetiracetam 20mg/kg BID from hospital days 5–25, and received enteral feeding throughout her hospital stay.
On hospital day 15, she was noted to be anuric and febrile. Her BUN was 58mg/dL with a creatinine rising to 2.47mg/dL from a baseline of 0.27. Labs showed a serum sodium of 150mmol/L, serum corrected calcium of 10.0mg/dL, and a serum phosphate of 5.3mg/dL. Renal ultrasound demonstrated bilateral hydroureteronephrosis and hyperechoic debris with shadow artifact in the collecting systems (Fig. 1). Non-contrast CT revealed bilateral stones throughout both collecting systems ranging from 2 to 5mm in diameter (Fig. 2).Fig. 1 Preoperative longitudinal ultrasound of right kidney demonstrating shadowing debris in renal pelvis.
Fig. 1Fig. 2 Maximum Intensity Projection coronal non-contrast CT image demonstrating stones in bilateral renal pelvises and ureters (arrows).
Fig. 2
She was taken urgently for cystoscopy and bilateral ureteral stent placement. Her creatinine returned to baseline two days later. Stone analysis revealed calcium phosphate. Urine cultures and analysis showed no evidence of infection and repeat renal ultrasound showed resolution of hydroureteronephrosis (Fig. 3).Fig. 3 Postoperative longitudinal ultrasound of the right kidney demonstating stent, resolution of caliectasis, and absence of debris.
Fig. 3
Ureteroscopy two months later revealed no stones, and a large coagulum within the renal pelvis. She has remained stone-free without medical therapy with normal renal function.
Discussion
This is the first reported case of nephrolithiasis in the setting of traumatic brain injury and hypertonic saline administration.
A review of renal physiology is crucial in understanding the potential mechanism of stone formation. Filtered sodium passes through the proximal tubule of the nephron and is then reabsorbed by the NKCC2 channel of the thick ascending limb with potassium and chloride. Some potassium leaks back into the tubular lumen through the renal outer medullary potassium (ROMK) channels, increasing luminal charge and driving calcium back into the peritubular capillaries through the paracellular route. Increased plasma sodium modulates both of these transport proteins, ultimately leading to less calcium reabsorption and thus increased calciuria. We believe that her increased sodium levels could have led to her hypercalciuria.
Hypertonic saline has a pH of approximately 5.0, and as such would have induced an iatrogenic metabolic acidemia. Bone releases calcium and phosphate to buffer acidic loads through direct ion exchange and modulation of osteoclast/blast activity, as well as in response to immobilization, both factors found in this patient.3 Corrected serum calcium levels remained within normal limits throughout her stay, but these would have underestimated her free serum calcium due to calcium-hydrogen exchange on the anionic albumin binding sites in acidemia. It seems reasonable that our patient with such a high salt load would develop calciuria through these mechanisms. Her stones were calcium phosphate, which forms in a high urinary pH, which corresponds to her pH of 9 that was previously mentioned.
Distal renal tubular acidosis (dRTA) is likely the underlying mechanism for her calcium phosphate stones. dRTA is caused by an inability to reabsorb bicarbonate, which leads to chronically alkaline urine. The proximal tubule responds by increasing citrate reabsorption, leaving an alkalotic urine low in citrate: a perfect milieu for calcium phosphate stone formation. This patient had no history of symptoms indicative of the hyperchloremic acidosis and severe hypokalemia often seen with dRTA (polydipsia, polyuria, weakness), which makes incomplete dRTA a potential diagnosis. Patients with incomplete dRTA only reveal their defect when challenged with an acid load, much like this patient.4 Failure to acidify urine during acid loading is diagnostic for dRTA, and may prompt the initiation of potassium citrate supplementation to lower urinary calcium excretion and increase urinary citrate levels.1
An additional factor to consider in this patient is the levetiracetam she received. Some anti-epileptic drugs (AEDs) are known to increase stone risk. It is postulated that this is due to urine alkalization.5 Our patient was on levetiracetam, which has not been studied regarding urinary pH or nephrolithiasis. It is reasonable, to suspect that her levetiracetam use may have increased her risk of stones by alkylinizing her urine, in the setting of a dRTA. It is also possible that partial precipitation of the drug may have caused the organic coagulum found at stent removal.
Conclusion
This case of HTS-induced obstructing nephrolithiasis presents a novel mechanism of iatrogenic stone formation and unmasking of RTA, and invites review of renal physiology and iatrogenic stone formation. Our patient returned for a follow-up with pediatric urology, at which time she continued to be stone-free with normal labs and blood pressure The patient will be periodically monitored by our multidisciplinary pediatric stone clinic to ensure she has no further stone episodes. | Recovered | ReactionOutcome | CC BY-NC-ND | 33194552 | 19,598,806 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Condition aggravated'. | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | IMIQUIMOD | DrugsGivenReaction | CC BY-NC-ND | 33194554 | 20,768,881 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Off label use'. | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | GEMCITABINE, PRALATREXATE | DrugsGivenReaction | CC BY-NC-ND | 33194554 | 19,598,815 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Urine flow decreased'. | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | IMIQUIMOD | DrugsGivenReaction | CC BY-NC-ND | 33194554 | 20,768,881 | 2021-01 |
What was the administration route of drug 'IMIQUIMOD'? | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | Topical | DrugAdministrationRoute | CC BY-NC-ND | 33194554 | 19,627,157 | 2021-01 |
What was the dosage of drug 'BRENTUXIMAB VEDOTIN'? | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | 1.8 MILLIGRAM/KILOGRAM | DrugDosageText | CC BY-NC-ND | 33194554 | 19,567,167 | 2021-01 |
What was the outcome of reaction 'Condition aggravated'? | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | Recovered | ReactionOutcome | CC BY-NC-ND | 33194554 | 20,768,881 | 2021-01 |
What was the outcome of reaction 'Drug ineffective'? | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | Fatal | ReactionOutcome | CC BY-NC-ND | 33194554 | 19,627,157 | 2021-01 |
What was the outcome of reaction 'Urine flow decreased'? | Complete response of penile mycosis fungoides with systemic brentuximab therapy.
Mycosis fungoides with penile involvement is extremely rare. Previous reports have shown successful treatment with imiquimod or a combination of beam radiation and chemotherapy. We present a patient with mycosis fungoides and penile involvement. The penile lesions were initially treated with topical imiquimod; however, he developed worsening glandular lesions and discharge. Therefore the treatment was discontinued. Subsequent treatment with brentuximab (anti-CD30) targeted therapy resulted in complete resolution of the penile lesions. To our knowledge, this represents the first case of a complete penile mycosis fungoides response to brentuximab therapy. Brentuximab may be considered for refractory penile mycoses fungoides.
Abbreviations
1. MFMycosis Fungoides
2. CTCLCutaneous T-cell lymphoma
3. CHOPCyclophosphamide, doxorubicin hydrochloride, vincristine sulfate, prednisone
Introduction
Mycosis fungoides (MF) is the predominant form of cutaneous T-cell lymphoma (CTCL) and accounts for 50–65% of cases.1 MF is rare, with an incidence in the United States of only 0.3–1 cases per 100,000. MF can be identified as erythematous patches or plaques, most often in a non-sun-exposed (so-called bathing trunk) distribution. Most patients present with patches or plaques, representing early stage disease; rarely, MF can progress to more advance stages with tumors or erythroderma.
Treatment of early stage MF commonly includes topical therapies such as corticosteroids, nitrogen mustard, phototherapy, imiquimod, retinoids and local radiation.2,3 Advanced disease requires systemic treatment, such as bexarotene, methotrexate, histone deacetylase inhibitors, pralatrexate, brentuximab vedotin (anti-CD30), and/or allogeneic stem cell transplantation.2,4 Early diagnosis and treatment is generally associated with a good prognosis, and 5-year survival rates approach 90%.1 Penile involvement with MF is rare with only 2 cases reported in the literature.3,5 We present the first report of complete response of penile MF with systemic brentuximab therapy.
Case presentation
A 71 year old uncircumcised Caucasian male with history of diabetes mellitus and MF, presented with new onset lesions on his glans and penile shaft. He had been originally diagnosed with MF, stage IB, in 2014. At the time of the diagnosis, he had plaques involving the lower abdomen, left flank and proximal left thigh. His lesions were initially under control with narrow-band ultraviolet therapy and methotrexate; however, 3 years later, he developed new erythematous plaques on his glans and shaft associated with a decreased urinary stream. Oral antibiotic and topical antifungal treatment provided only temporary improvement of his symptoms. He denied dysuria, penile discharge, fevers or weight loss. Physical examination demonstrated an uncircumcised penis with phimotic foreskin, a 1.5cm erythematous, non-tender, mobile lesion on the right inner preputial surface, and inflammation on the glans penis surrounding the urethral meatus. There were no lesions or stricture involving the urethral mucosa. Biopsies were obtained from the abnormal areas.
Histology showed a dense mononuclear inflammatory infiltrate involving the epidermis and dermis. Numerous atypical, enlarged lymphocytes with hyperchromatic nuclei were noted in the epidermis, superficial and mid-dermis with evidence of fibroplasia (Fig. 1). The cells were predominantly CD3 positive T-cells, with rare CD30 cells. These findings were consistent with his previous biopsies and a diagnosis of MF. Following the biopsy a urethral dilation was performed, which resulted in improvement of the patient's stream.Fig. 1 Histologic features showed a dense mononuclear inflammatory infiltrate in the epidermis and mid-dermis (A), H&E 4x; Note the presence of enlarged atypical lymphocytes with hyperchromatic nuclei in the epidermis (B), H&E, 40x.
Fig. 1
In order to improve hygiene he underwent a circumcision, after which he developed purulent exudates on the glans penis. Topical imiquimod 5% was started; however, the patient discontinued treatment after 3 weeks due to increased purulent secretions (Fig. 2A) and further weakening of urinary stream secondary to involvement of the urethral meatus. Shortly thereafter, he presented with swelling over the left eyelid consistent with progression of MF and underwent local radiation therapy as treatment. Unfortunately, he progressed further with lesions to the right orbit, lateral mouth, and chest. At this point systemic brentuximab therapy was initiated at 1.8mg/kg every three weeks. After two infusions, the patient noted a marked improvement in his urinary stream, and physical exam demonstrated normal epidermal skin on the glans (Fig. 2B). Initial side effects of brentuximab included mild GI upset and fatigue. After 3 infusions peripheral neuropathy was noted. Treatment was discontinued after 7 infusions due to refractory neuropathy. Eight months following cessation of brentuximab infusions, no further lesions were noted on the penis. However, his diseased recurred with large cell transformation, sparing the penis, while he was off treatment. The patient underwent further treatment with pralatrexate and gemcitabine, but unfortunately passed away from refractory disease 4 months after restarting chemotherapy.Fig. 2 Gross images of penile mycosis fungoides before (A) and after (B) 2 brentuximab infusions.
Fig. 2
Discussion
Penile involvement in MF is rare, with only 2 case reports in the literature to date. Reports of MF of the penis have shown tissue preserving remission with immune system modifiers and combinations of electron beam radiation and chemotherapy.3,5 Chiam and Chan reported a case of a 32 year old healthy man from Bangladesh with a pink plaque on the glans that had been present for 15 years.3 After 6 weeks of treatment with topical clobetasol propionate, which failed to provide improvement, he was switched to 5% imiquimod. After 4–5 months of therapy the patient was in complete remission. During treatment he developed pain at the application site and a skin erosion requiring cessation of therapy. A second case reported by O'Brien et al. involved a 64 year old with an ulcer on the penile meatus.5 The patient initially underwent surgical excision of the lesion, and was later treated with 15 fractions of 27 Gy radiotherapy and mini-CHOP. This combination treatment provided a complete response.
Our case is unique, as complete response of penile MF to brentuximab systemic therapy has not yet been reported. Brentuximab vedotin is an anti-CD30 monoclonal antibody that has been shown to provide effective targeted therapy for cutaneous T-cell lymphomas.2 Brentuximab was initially approved for the treatment of Hodgkin's lymphoma to target CD30 on Reed-Sternberg cells.4 The Food Drug Administration approved its use for the treatment of patients with CTCL in 2017. Initially, it was used to treat patients with CD30 expressing CTCL, such as primary cutaneous anaplastic large cell lymphoma and CD30 expressing MF. However, Brentuximab has shown efficacy in CTCL patients with low CD30 expression, as it was the case in our patient.2 Potential side effects of brentuximab therapy include neuropathy, leukopenia and fatigue.4 Additionally there are multiple reports indicating that progressive multifocal leukoencephalopathy is associated with treatment. Our patient developed neuropathy leading to gait instability after 7 cycles, at which point therapy was stopped. In patients with refractory Hodgkin lymphoma, brentuximab chemotherapy is commonly followed by hematopoietic stem cell transplant.4 Our patient was not a suitable candidate for stem cell transplantation due to underlying comorbidities. As brentuximab was the likely cause of the patient's neuropathy, hematopoietic stem cell transplant may have provided a more tolerable treatment.
Conclusion
In summary, brentuximab systemic targeted therapy may be considered as a treatment option for patients with penile MF refractory to more traditional topical therapies.
Consent
The patient provided consent for this case report including the use of images.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Declaration of competing interest
The authors have no conflicts of interest to disclose.
Acknowledgements
Thank you to the patient and to the supporting clinical staff. | Recovering | ReactionOutcome | CC BY-NC-ND | 33194554 | 20,768,881 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Maternal exposure before pregnancy'. | A Pharmacoepidemiologic Approach to Evaluate Real-world Effectiveness of Hormonal Contraceptives in the Presence of Drug-drug Interactions.
Accurate estimation of conception is critical in the assessment of the effects of drugs used during pregnancy or to prevent pregnancy. In a novel application, we studied the effectiveness of oral contraceptives (OCs), where misclassification of conception relative to OC exposure may obscure effect estimates.
We studied OC failure, in a large claims database, among women who used antiepileptic drugs with metabolizing enzyme-inducing properties (carbamazepine or oxcarbazepine), which reduce OC's effectiveness or enzyme-neutral properties (lamotrigine or levetiracetam), with no expected impact on OC effectiveness. We compared conception rates in women 12-48 years of age concomitantly using OCs and enzyme-inducing drugs with rates in concomitant users of OCs and enzyme-neutral drugs. We measured conception with a validated algorithm that estimates gestational age based on pregnancy endpoints. We estimated relative and attributable risk using generalized estimating equation models after standardized mortality ratio weighting.
We identified 89,777 concomitant use episodes with adjusted contraceptive failure rates of 1.6 (95% confidence interval (CI) = 1.4, 1.8) per 100 person-years among users of enzyme-neutral drugs and 18,964 episodes with a rate of 2.3 (1.9, 2.8) among users of enzyme-inducing drugs. The relative risk of conception for enzyme-inducing group was 1.4 (1.1, 1.8), and the rate difference was 0.7 (0.2, 1.2).
OCs in combination with antiepileptic drugs that interact with metabolic enzymes were associated with increased contraceptive failure rates. Measurement of conception in claims data had adequate accuracy to uncover a strong drug-drug interaction, offering promise for broader application in comparative effectiveness studies on hormonal contraceptives to inform clinical and regulatory decisionmaking.
Observational study designs employing real-world data are commonly used to evaluate the safety and effectiveness of medications during pregnancy.1,2 To ensure accurate timing of drug exposure, pregnancy episodes are usually determined via a specific pregnancy endpoint, estimation of gestational age at the time of the endpoint, and imputation of the pregnancy start date.3–5 Although the International Classification of Disease (ICD) coding offers detail on gestational age at delivery, the codes are not consistently applied to medical encounter claims, thus leaving some uncertainty about the exact date of conception. Such inaccuracies are more pronounced for preterm deliveries and other adverse pregnancy endpoints such as stillbirths and abortions.6,7 Inappropriate timing of conception will in turn result in misclassification of drug exposure, and introduce bias. Misclassification bias will be most pronounced if the exposure window to be studied is close to conception (such as when evaluating effects of first-trimester exposure on the risk for malformation), or exposure pattern varies over time.8 It will be even more prominent in scenarios where the pregnancy was not intended or when studying drugs with safety concerns regarding use during pregnancy, and thus exposure is terminated as soon as pregnancy is discovered, leading to only a short period of fetal exposure during pregnancy.
Another research area where accurate timing of conception is critical is in the evaluation of contraceptive effectiveness. Contraceptive failure is operationalized as conception during exposure to the contraceptive agent, and if such failure occurs, the contraceptive use will be discontinued as soon as pregnancy is discovered. Delayed conception estimates may inaccurately conclude that the contraceptive was effective. On the other hand, estimates that time conception too early may conclude that the contraceptive failed even though it may have been discontinued because intentions to prevent pregnancy had changed. Thus, even small errors of conception estimation of only a few weeks, which would be expected for some live births and to a larger extent for nonlive pregnancies, may yield claims data unusable to evaluate the real-world effectiveness of contraceptives. In the previous study, we developed a pregnancy identification algorithm based on several validation studies6,7 to evaluate the effectiveness of regulatory requirements designed to prevent maternal exposure to teratogenic medications.9 We have recently expanded the algorithm to incorporate diagnosis and procedure codes from the ICD-10. In light of the above-described concerns about conception timing, we aimed to investigate whether pregnancy identification algorithm can accurately identify contraception failure. We chose a clinical scenario where a well-documented drug–drug interaction modifies the failure rate of hormonal contraceptives. In this scenario, possible misclassifications of exposure or outcome could obscure causal inferences when evaluating the drug–drug interaction. However, if shown to be sensitive to detect the drug–drug interaction, this approach would offer opportunities to study a broad range of clinical risk factors for unintended pregnancy using real-world data, including interactions involving hormonal contraceptives, and thus advance pharmacoepidemiologic methods.
METHODS
Study Design and Data Source
We conducted a cohort study using the IBM MarketScan Commercial Claims Databases (2005–2017). This database includes data on inpatient and outpatient medical encounters and pharmacy dispensing claims for a large sample of the privately insured population in the United States. The beneficiaries have encrypted identifiers in the database, which allows for longitudinal follow-up. This database is certified as de-identified data, and the present study was approved as exempt by the Institutional Review Board at the University of Florida (IRB approval number: 201801093).
Clinical Scenario
Carbamazepine and oxcarbazepine are among the first-generation antiepileptic drugs with several approved and off-label indications, including epilepsy and bipolar disorder.10 Both drugs are inducers of Cytochrome P450 3A4 (CYP3A4), with well-documented evidence for reduction of estrogen/progestin plasma levels of oral contraceptives (OCs), and ovulation pattern disruption.11–13 The enzyme induction effect of carbamazepine and oxcarbazepine is comparable and may result in approximately 50% change in area under the curve (AUC) of OC products. 12,14 Therefore, and because both antiepileptic drugs are associated with neural tube defects, clinical guidelines recommend against the use of OCs when using carbamazepine or oxcarbazepine to avoid unintended pregnancy.15,16 In contrast, newer antiepileptic drugs, including lamotrigine and levetiracetam, have minimal effect on CYP3A4 and are not expected to reduce OC efficacy.17,18 In the present study, we aimed to compare the rate of OC failure in two cohorts of OC users who had concomitant use of either a CYP3A4-inducer (i.e., carbamazepine or oxcarbazepine) or a CYP3A4-neutral (i.e., lamotrigine or levetiracetam) antiepileptic drugs.
Study Cohorts
We identified female patients of childbearing age (12–48 years old) who had at least one pharmacy claim for a combined OC with low-dose estrogen (<50 µg) or a progestin-only OC, referred to as OCs. We identified pharmacy claims for lamotrigine or levetiracetam to create a cohort of OC users without drug–drug interaction and extracted claims for carbamazepine or oxcarbazepine to create a cohort of OC users with drug–drug interaction. These four antiepileptic drugs were selected based on their profile of CYPA34 activity and potential clinical uses to create comparable study cohorts with regard to baseline clinical characteristics and pregnancy rates. We defined the cohort entry date (i.e., index date) as the first day of concomitant use of the drugs of interest and OCs and defined a look-back period of six months before the index date with continuous insurance enrollment to ascertain drug indications and other covariates. Patients were required to have at least one medical claim for epilepsy, bipolar disorder, or personality disorder (i.e., indications for antiepileptic drugs) during the look-back period. The indications were measured using coding algorithms developed by the Center for Medicare and Medicaid Services, Chronic Conditions Data Warehouse based on the ICD, ninth, and tenth versions, clinical modification (ICD9/10-CM) codes.19 We excluded patients if they had medical diagnoses for infertility, ovary dysfunction, or hirsutism in their look back period to rule out off-label indications for OCs. The list of ICD codes is provided in eAppendix 1; http://links.lww.com/EDE/B746.
Concomitancy Definition
We assumed drug exposure started at the prescription dispensing date and ended on the last day of the pharmacy-entered dispensed days’ supply. We defined “concomitancy” as overlapping exposure periods regardless of the order of drug dispensing for antiepileptic drugs or OCs. 20 We excluded concomitancy periods that had exposure to valproate sodium, topiramate, or phenytoin because of potential CYP3A4 activity and teratogenic effects, which may encourage patients to use a second contraceptive method (e.g., barrier methods). We also excluded concomitancy periods where we observed any other hormonal contraceptive agents, including long-acting reversible contraceptives (e.g., intrauterine devices), injectables, or high-dose estrogen OCs. For users of CYP3A4-neutral antiepileptic drugs, we excluded the concomitancy periods with CYP3A4-inducer drugs. Then we created concomitant use episodes for patients in each study group, and a gap of ≥14 days for either of the medications during an episode was allowed. If we observed a gap of more than 14 days concomitant use, the concomitancy episode ended on the last day of concomitant use. Patients were allowed to reenter the cohort if they had subsequent concomitant use episodes after their first observation period and met all inclusion criteria, including the availability of the 6-months look-back period to allow re-evaluation of baseline characteristics.
Outcome Definition
The study outcome was contraception failure defined as conception during a concomitancy period. Conception was estimated via the pregnancy identification algorithm that uses medical encounters with ICD-9/10-CM, Current Procedural Terminology, and Healthcare Common Procedure Coding System codes to identify specific pregnancy endpoints, including live birth, ectopic pregnancies, stillbirth, terminations, and prenatal screening visits.6,7,21,22 Once pregnancy episodes were identified, the algorithm estimated gestational age to calculate the last menstrual period (LMP). The conception date was assumed to be 14 days after the estimated LMP date. We provide more details on the pregnancy identification algorithm in eAppendix 1; http://links.lww.com/EDE/B746.
Covariates
We measured several demographic and clinical variables at baseline to assess the comparability of study cohorts, including patient age, residence region, insurance plan type, and relationship to the employee covered by the health plan (spouse, employee, children/other). Clinical variables measured during the 6-months look-back period included recent pregnancy history (based on any pregnancy endpoint), use of teratogenic drugs (with or without mandated pregnancy prevention programs), and a variety of clinical conditions that may affect OC efficacy.
Study Follow-up
All patients were required to have insurance enrollment for a minimum of 90 days after their concomitancy episode ended. This requirement allows the pregnancy identification algorithm to capture pregnancy-related medical encounters, which are then used to date conception. This minimum number of days was defined based on our previous work that showed approximately 90% of live deliveries in our database have prenatal visits within the first 90 days after conception. We followed each patient from the index date of each concomitancy episode until conception, infertility, ovary dysfunction, hirsutism diagnoses, initiation of a teratogenic drug, end of concomitancy, maximum of 3 years’ follow-up, or end of study (December 31, 2017).
Statistical Analysis
The unit of analysis was a concomitancy episode. We compared baseline demographic and clinical characteristics of each study cohort using a threshold of an absolute standardized difference (ASD) higher than 10% as clinically significant.23 To account for confounding, we used a logistic regression model to create an exposure propensity score and selected covariates into the model based on a literature review on potential risk factors for the study outcome.24 We used the common support region of the score to create weights to estimate the average treatment effect among the treated, also known as the standardized mortality ratio (SMR) weighting method.23,25 In this weighted pseudo-population, the confounding effect of measured covariates is eliminated, and the effect estimates can approximate the causal effect. We used these SMR weights in a generalized estimating equation model with a Poisson distribution and offset of follow-up time to compare contraception failure rates among the study cohort. We used a robust variance estimator to account for the clustering of episodes within the same patient. We conducted all data management and analyses using SAS 9.4 and SAS/STAT 15.1 (Cary, NC).
Sensitivity Analysis
We performed several sensitivity analyses to evaluate the robustness of the study findings. For the concomitancy definition, we changed the gap allowance to 1 or 7 days instead of 14 days. We also excluded concomitancy periods that overlapped with other moderate or strong inducers or inhibitors of CYP3A4 for more than 14 days (see eAppendix-1; http://links.lww.com/EDE/B746: eTable 1; http://links.lww.com/EDE/B746 and eTable 2; http://links.lww.com/EDE/B746). For the outcome definition, we varied the estimated conception date by ±14 days, limited the analysis to episodes indexed before 2015 (confining coding to ICD-9-CM), limited the outcome definition to only live birth episodes (which allows the most accurate conception date estimation). To evaluate the impact of more homogeneous comparison groups, we restricted the maximum follow-up time to 6 months, limited the analysis to only the first episode of concomitancy for each patient, and restricted drug initiation sequence to patients who initiated OC initiation while on antiepileptic drug treatment.
RESULTS
In the main analysis, we identified 89,777 concomitancy episodes involving CYP3A4-neutral antiepileptic drug and 18,964 episodes with the CYP3A4 inducers. Cohorts had similar age distributions with a mean age of 26.3 ± 8.5 years for the CYP3A4-neutral and 25.5 ± 9.2 for the CYP3A4-inducing drugs (Table 1). About 70% of women in each cohort had a bipolar disorder diagnosis, and 30% had an epilepsy diagnosis. We observed a high prevalence of anxiety (29% vs. 29%) and teratogenic medication use (30% vs. 28%) in the baseline period. The baseline covariates were balanced between study cohorts except for age, beneficiary status, and paralysis diagnosis (ASD < 15%). SMR weighting successfully balanced all measured characteristics (Table 1). Propensity score distributions, SMR weights, and hazard plots are available in eAppendix 1; http://links.lww.com/EDE/B746.
TABLE 1. Demographic and Clinical Characteristics of the Study Cohorts
Covariates Before SMR Weighting After SMR Weighting
Concomitant OC Plus Enzyme-neutral AED Episodesa (N = 89,777) Concomitant OC Plus Enzyme-inducing AED Episodesb (N = 18,964) ASD (%) Concomitant OC Plus Enzyme-neutral AED Episodesa (N = 18,973) Concomitant OC Plus Enzyme-inducing AED Episodesb (N = 18,964) ASD (%)
Age n (%) n (%) n (%) n (%)
<20 23,297 (26) 6337 (33) 16 6359 (34) 6337 (33) 0
20–29 36,339 (41) 6748 (36) 10 6741 (36) 6748 (36) 0
30–39 21,618 (24) 3918 (21) 8 3910 (21) 3918 (21) 0
≥40 8523 (10) 1961 (10) 3 1964 (10) 1961 (10) 0
Hypertension 3732 (4) 946 (5) 4 958 (5) 946 (5) 0
Hyperlipidemia 4114 (5) 1011 (5) 3 1021 (5) 1011 (5) 0
Obesity 1277 (1) 357 (2) 4 358 (2) 357 (2) 0
Epilepsy 23,309 (26) 5816 (31) 10 5853 (31) 5816 (31) 0
Bipolar disorder 65,116 (73) 13,081 (69) 8 13,057 (69) 13,081 (69) 0
Schizophrenia 2428 (3) 852 (5) 10 853 (4) 852 (5) 0
Depression 50,723 (57) 10,219 (54) 5 10,215 (54) 10,219 (54) 0
Personality Disorder 5606 (6) 1209 (6) 1 1210 (6) 1209 (6) 0
Anxiety 26,043 (29) 5544 (29) 1 5542 (29) 5544 (29) 0
Substance Use Disorder 4575 (5) 1284 (7) 7 1288 (7) 1284 (7) 0
Recent pregnancy (live birth) 2248 (3) 309 (2) 6 308 (2) 309 (2) 0
Recent pregnancy (termination) 507 (1) 104 (1) 0 103 (1) 104 (1) 0
Teratogenic drug without REMS 24,925 (28) 5588 (30) 4 5620 (30) 5588 (30) 0
Teratogenic drug with REMS 414 (1) 95 (1) 1 96 (0) 95 (1) 0
Charlson Comorbidity Index (CCI)3
≤1 87,223 (97) 18,93 (95) NA 18,127 (95) 18,093 (95) NA
1< 2554 (3) 871 (5) NA 846 (5) 871 (5) NA
Comorbidities
Myocardial infarction 31 (0) < 11 0 <11 < 11 0
Congestive Heart Failure 175 (0) 50 (0) 1 51 (0) 50 (3) 0
Vascular Disorder 187 (0) 51 (0) 1 54 (0) 51 (3) 0
Cerebrovascular Disorder 892 (1) 197 (1) 1 202 (1) 197 (1) 0
Pulmonary Disorders 7211 (8) 1705 (9) 3 1716 (9) 1705 (9) 0
Dementia 86 (0) 30 (0) 2 31 (2) 30 (0) 0
Paralysis 613 (1) 380 (2) 11 388 (2) 380 (2) 0
Diabetes w/o complications 1945 (2) 432 (2) 1 434 (2) 432 (2) 0
Diabetes with complications 175 (0) 40 (0) 0 40 (0) 40 (0) 0
Renal Disorders 263 (0) 64 (0) 1 67 (3) 64 (0) 0
Mild Liver Disorders 765 (1) 192 (1) 2 194 (1) 192 (1) 0
Severe Liver Disorders 22 (0) <11 0 <11 <11 0
Peptic Ulcer 193 (0) 58 (0) 2 60 (3) 58 (0) 0
Rheumatoid Disorders 585 (1) 156 (1) 2 160 (8) 156 (1) 0
AIDS 22 (0) <11 0 <11 <11 0
Malignancy 793 (1) 170 (1) 0 173 (9) 170 (1) 0
Metastatic Malignancy 39 (0) <11 0 <11 <11 0
Beneficiary status
Employee 32,469 (36) 5702 (30) 13 5698 (30) 5702 (30) 0
Spouse 14,208 (16) 2764 (15) 3 2677 (15) 2764 (15) 0
Child/other 43,100 (48) 10,498 (55) 15 10,509 (55) 10,498 (55) 0
Residence region
Northeast 16,228 (18) 3045 (17) 5 3047 (16) 3045 (16) 0
Northcentral 19,394 (22) 4363 (23) 4 4370 (23) 4363 (23) 0
South 35,322 (39) 8074 (43) 7 8071 (42) 8074 (43) 0
West 17,594 (20) 3225 (17) 7 3228 (17) 3225 (17) 0
Unknown 1239 (1) 257 (1) 0 257 (1) 257 (1) 0
Health plan type
COM 1525 (2) 479 (2) 6 482 (3) 479 (3) 0
HMO 12,955 (14) 2749 (14) 0 2733 (14) 2749 (15) 0
PPO 53,996 (60) 11,362 (60) 1 11,373 (60) 11,362 (60) 0
POS 6853 (8) 1594 (8) 0 1597 (8) 1594 (8) 0
CDHP 5887 (6) 1216 (6) 1 1218 (6) 1216 (6) 0
Other 8561 (9) 1564 (8) 5 1570 (8) 1564 (8) 0
ASD, absolute standardized difference; COM, comprehensive; HMO, health maintenance organization; PPO, preferred provider organization; POS, noncapitated point-of-service; CDHP, consumer-driven health plan; Other: includes capitated/partially capitated point-of-service, exclusive provider organization, high deductible health plan; REMS, risk evaluation and mitigation strategy.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam). After SMR weighting, the size of the pseudo-population in cohort A becomes comparable to cohort B.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
Episodes involving concomitant use of enzyme-neutral antiepileptic drugs had a slightly larger mean follow-up time of 96 days (vs. 79 days among women who used enzyme-inducers). Concomitancy periods ended with 400 conceptions among women who used enzyme-neutral antiepileptic drugs, resulting in a crude contraception failure rate of 1.7 events per 100 person–years. Women with concomitant use of OCs and enzyme-inducing antiepileptic drug had 94 conceptions with a crude contraception failure rate of 2.3 per 100 person–years (Table 2). Figure 1 shows unadjusted survival plots for contraception failure outcome. Approximately two-thirds of all conceptions were identified based on liveborn deliveries (both groups 63%), whereas abortions were the second prevalent pregnancy endpoint (28% in the enzyme-neutral group vs. 27% in the enzyme-inducing group) (Table 3).
TABLE 2. Relative and Absolute Risk of Oral Contraceptive Failure in the Presence or Absence of Drug–drug Interaction
Study Cohort Events Total Follow-up Time (Person-years) Incidence Rate (per 100 Person-years) Relative Risk Risk Difference
Unadjusted analysis
Concomitant OC plus enzyme-neutral AED episodesa 400 23,647 1.7 (1.5, 1.9) REF REF
Concomitant OC plus enzyme-inducing AED episodesb 94 4102 2.3 (1.9, 2.8) 1.4 (1.1, 1.7) 0.6 (0.1, 1.1)
Adjusted Analysis
Concomitant OC plus enzyme-neutral AED episodesa 400 23,647 1.6 (1.4, 1.8) REF REF
Concomitant OC plus enzyme-inducing AED episodesb 94 4102 2.3 (1.9, 2.8) 1.4 (1.1, 1.8) 0.7 (0.2, 1.2)
ASD, absolute standardized difference; OC, oral contraceptive.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam)—Reference cohort.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
TABLE 3. Pattern of Pregnancy Episodes that Contributed to Measurement of Contraception Failure
Study Cohort Full-term
n (%) Preterm
n (%) Post-term
n (%) Ectopic
n (%) Stillbirth
n (%) Spontaneous abortion
n (%) Induced abortion
n (%) Unknown outcome
n (%) Total
n
Concomitant OC plus enzyme-neutral AED episodesa 233 (58) 16 (4) 3 (1) 4 (1) 3 (1) 65 (16) 46 (12) 30 (8) 400
Concomitant OC plus enzyme-inducing AED episodesb 54 (57) 4 (4) 1 (1) 1 (1) 0 (0) 14 (15) 11 (12) 9 (10) 94
ASD, absolute standardized difference; OC, oral contraceptive.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam)—Reference cohort.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
The adjusted contraceptive failure rates were 1.6 (95% CI = 1.4, 1.8) per 100 person–years among users of enzyme-neutral drugs and 2.3 (1.9, 2.8) among users of enzyme-inducing drugs. The marginal models with SMR weights increased the unadjusted relative risk for contraception failure slightly from 1.4 (1.1, 1.7) to 1.4 (1.1, 1.8), comparing women who used enzyme-inducing antiepileptic drugs to those who used enzyme-neutral antiepileptic drug. Concomitant use of enzyme-inducing antiepileptic drug and OCs resulting in an additional 0.7 conceptions (0.2, 1.2) per 100 person–years of concomitant use.
All sensitivity analyses corroborated our findings (Table 4). Analyses with conceivably superior measurement of confounding (analysis number 6, 7, 9 in Table 4) showed as expected slightly larger relative risk estimates with 1.6 (1.2, 2.2) when restricting to the first concomitancy episode per patient, 1.5 (1.1, 2.1) with a fixed sequence of drug initiation, and 1.5 (1.2, 1.9) after eliminating concomitant use of all other potential CYP3A4 inducers/inhibitors.
TABLE 4. Sensitivity Analyses on the Definitions of Concomitancy, Outcome Measurement, and Study Design Features
Row Sensitivity Analysis Incidence Rate (per 100 Person-year) Adjusted Rate Ratio Adjusted Rate Difference
Concomitant OC Plus Enzyme-neutral AED Episodesa Concomitant OC Plus Enzyme-inducing AED Episodesb
1 Conception date altered by +14 days 1.5 (1.3, 1.7) 1.9 (1.5, 2.4) 1.3 (1.0, 1.6) 0.4 (0.0, 0.9)
2 Conception date altered by −14 days 1.9 (1.7, 2.1) 2.5 (2.1, 3.0) 1.3 (1.1, 1.7) 0.6 (0.1, 1.1)
3 Restriction to episodes with index date before 2015 (ICD-9 era) 1.6 (1.4, 1.8) 2.3 (1.9, 2.9) 1.6 (1.1, 1.8) 0.7 (0.2, 1.3)
4 Conception ascertained based on live birth only 1.0 (0.9, 1.2) 1.4 (1.1, 1.8) 1.4 (1.1, 1.9) 0.4 (0.2, 8.0)
5 Maximum follow-up time restricted to 6 months 1.7 (1.5, 1.9) 2.3 (1.9, 2.9) 1.4 (1.1, 1.8) 0.6 (0.1, 1.2)
6 Episodes restricted to first episode per patient 1.6 (1.4, 1.8) 2.5 (1.9, 3.3) 1.6 (1.2, 2.2) 0.9 (0.2, 1.6)
7 Episodes restricted to those where OC initiation follows AED use 1.5 (1.3, 1.8) 2.3 (1.8, 3.1) 1.5 (1.1, 2.1) 0.8 (0.1, 1.5)
8 Permissable gap in concomitancy 7 days 1.6 (1.5, 1.8) 2.2 (1.8, 2.7) 1.4 (1.1, 1.7) 0.6 (0.1, 1.0)
9 Permissable gap in concomitancy 1 day 1.6 (1.5, 1.8) 2.1 (1.7, 2.6) 1.3 (1.0, 1.6) 0.5 (0.0, 0.9)
10 Exclusion of all follow-up time with other CYP3A4 inducers or inhibitors 1.6 (1.4, 1.8) 2.4 (1.9, 2.9) 1.5 (1.2, 1.9) 0.7 (0.2, 1.3)
ASD, absolute standardized difference; ICD, International Classification of Disease; OC, oral contraceptive.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam)—Reference cohort.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
DISCUSSION
Our study found that women who concomitantly used OCs and CYP3A4-inducing antiepileptic drugs were 40% more likely to experience contraceptive failure compared to CYP3A4-neutral antiepileptic users. Our findings suggest that conception estimation was sufficiently accurate to identify this well-documented strong drug–drug interaction, and thus, our approach may be useful to generate real-world evidence on mechanisms of contraceptive failure, including the examination of interactions.
Because respective drug approval requirements are largely confined to pharmacokinetic studies, limited information exists on the clinical significance of drug–drug interaction involving OCs. A meta-analysis published in 2010 identified only pharmacokinetic studies that examined the potential for contraceptive failure among patients who use antiepileptic drugs.26 We identified one small cohort study published in 1979 that followed 41 epilepsy patients with concomitant use of antiepileptic drugs and OCs and reported ~2.9 failures per 100 person–years, which is slightly higher than in our cohorts.27 In our study, we observed a failure rate of 1.6 or 2.3 per 100-person years, depending on the type of antiepileptic drugs in terms of enzyme-inducing properties. The observed magnitude of the drug–drug interaction impact on failure rates in our study is biologically plausible based on evidence from pharmacokinetic evaluations, which recommend higher OC doses when used concomitantly with carbamazepine (e.g., 80–100 mcg of ethinyl estradiol that far exceeds the observed doses of 20–35 mcg observed in our data). 11,12 A clinical trial on healthy volunteers showed a 46.1% reduction in the AUC for levonorgestrel, and 44.5% for ethinyl estradiol when the OC was administrated with carbamazepine (600 mg). The study also reported more ovulations (5/10 cycles vs. 1/10 cycles) among carbamazepine users versus OC use alone.12 The diminishing effects of CYP3A4 inducers, including carbamazepine and oxcarbazepine, on the efficacy of OC is also emphasized in the guidance document for labeling of combined hormonal contraceptives.28
Because clinical trials on contraceptive failure are typically infeasible, observational studies, especially prospective cohorts, have a pivotal role.29–31 Broader availability of real-world data and advancements in pharmacoepidemiologic methods can facilitate comparative effectiveness studies and help to translate mechanistic findings into clinically significant outcomes. To the best of our knowledge, the present study is the first attempt to investigate contraceptive failure in claims data. To evaluate the performance of this approach, we designed our study based on a clinical scenario of decreased OC efficacy in the presence of a known drug–drug interaction with a well-studied potent enzyme inducer. We successfully replicated mechanistic findings regarding a significant interaction between OC use and the perpetrator drug (a CYP3A4-inducer), resulting in contraceptive failure. Therefore, we envision that this approach could serve as a novel platform to evaluate the comparative effectiveness of hormonal contraceptives, considering different routes of administration and differences in pharmacokinetic profiles, among diverse patient populations with comorbidities or other concomitant medications. However, researchers should be vigilant about potential misclassification biases in exposure (contraceptive use) or outcome (conception) measurements and their consequences on the ability to make causal inferences, especially, for quantifying the magnitude of risk, and the possibility of inadequate sensitivity of our approach to detect weaker drug–drug interaction effects.
We operationalized exposure in our claims data by using “days of supply” recorded on the pharmacy claims. Although this measurement approach is more reliable than patient self-report, it may not be fully reflective of the actual medication consumption.32 For instance, women might discontinue OC treatment before their supply is exhausted to plan for pregnancy. This decision would possibly result in misclassification of exposure and overestimation of the contraception failure rate. The use of active comparator groups that exhibit similar demographics and comorbidities, such as demonstrated in our cohort, may mitigate some of these concerns, but even nondifferential exposure misclassification could bias relative risk estimates either away or toward the null hypothesis (eAppendix 2; http://links.lww.com/EDE/B747).
Regarding outcome measurement, we relied on an algorithm to infer pregnancy episodes from medical encounters with the healthcare system that use validated coding to specify pregnancy endpoints, but without an actual recording of LMP. Based on previous literature, estimation of conception will have varying degrees of accuracy for each type of pregnancy endpoint.6,22 In our pregnancy identification algorithm, we assigned gestational age for live-birth episodes based on an algorithm of ICD codes indicating gestational age or preterm status, and the overall agreement against birth certificates is reported to be >93% in the Medicaid database.22 For ectopic pregnancy, induced/spontaneous abortion, and stillbirth, we used fixed values for gestational age (8, 10, and 28 weeks, respectively) similar to previous literature.6,7 This approach has shown a moderate agreement both against medical charts (70% for ectopic pregnancy and 67% for spontaneous abortions within four weeks)7, and a cohort of in-vitro fertilization patients (77% agreement for ectopic pregnancy, 47% for stillbirth, and 36% for abortions cases).6 Validation studies on pregnancy identification algorithms typically report the agreement between the estimated LMP and the gold standard (e.g., the clinical estimate of gestational age on the birth certificate) within a prespecified number of weeks (e.g., 2 weeks) as the margin of error.4,6,22 It should be noted; however, that such a margin might need to be varied according to the pregnancy endpoint that was used to estimate LMP. For example, previous studies have shown that the capture of preterm status among deliveries improves with decreasing gestational age.33 Thus, considering the relationship between measurement sensitivity and gestational age as well as the pronounced left-skewed distribution of gestational age among preterm infants, error margins of 2–4 weeks used in sensitivity analyses may be appropriate. In contrast, stillbirth with a flat gestational age frequency distribution ranging from 20 to 42 weeks may require broader margins that should be tested.34 We should also note that available conception algorithms typically estimate LMP, following long-established conventions in timing gestational age, and thus, conception must be imputed as LMP + 14 days to operationalize contraceptive failure.35 In our study, we conducted sensitivity analyses by varying the estimated conception date, limiting the analysis to the ICD-9-CM era with previously validated pregnancy endpoint definitions, and restricting the events to pregnancies with the liveborn outcome. The two latter analyses aimed to increase the specificity of the outcome measure.
Our adjustment for potential confounding was based on a literature review on risk factors for contraceptive failure, but we acknowledge that several predictive factors were not or were only partially measurable in claims data. For example, sexual activity may be an important factor for contraceptive failure but is not available in claims data. However, we believe that our use of an active comparator group successfully balanced for the majority of risk factors, as exemplified by fairly well-balanced comparison groups before SMR weighting. We should also acknowledge that race and socioeconomic status were unmeasured in our dataset and could act as confounding factors. However, the study drugs are available on the market as generic products, yielding channeling for economic reasons unlikely. Finally, we should note that our adjustment for confounding further increased the observed relative risk estimates, thus suggesting that enzyme inducers were slightly more prevalent among patients with fewer risk factors for contraceptive failure. Thus, to explain our findings, any unmeasured covariate would need to be distributed in the opposite direction than the measured risk factors. Researchers could apply probabilistic bias analysis methods to evaluate the impact of exposure, outcome, and confounder misclassifications, simultaneously.36,37
In conclusion, women who use OCs in combination with carbamazepine or oxcarbazepine should be aware of increased contraceptive failure rates. Our study showed that measurement of conception in claims data has adequate accuracy to reveal the effect of a known drug–drug interaction with hormonal contraceptives. Our pharmacoepidemiologic approach is promising for comparative effectiveness studies on hormonal contraceptives to generate real-world evidence and inform clinical and regulatory decision-making.
FIGURE 1. Survival curves for contraception failure during follow-up in each study cohort.
ACKNOWLEDGMENTS
None.
Supplementary Material
A.S., J.B., S.S., and A.W. participated in study conceptualization and design. A.S., J.B., A.G., B.C., C.H., P.P., P.S., and A.W. involved in data acquisition, analysis, and interpretation. A.S. participated in drafting the article. A.S., J.B., A.G., B.C., C.H., P.P., P.S., and A.W. participated in the critical revision and final approval of the manuscript.
The results reported herein correspond to specific aims of grant #OPP1185454 to Stephan Schmidt and Joshua Brown from the Bill & Melinda Gates Foundation.
A.G.W. has active research support from Merck & Co, which is unrelated to this project. The other authors have no conflicts to report.
The IBM MarketScan database is available via a data use agreement with the company. | LEVETIRACETAM | DrugsGivenReaction | CC BY | 33196560 | 18,620,305 | 2021-03-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Pregnancy on contraceptive'. | A Pharmacoepidemiologic Approach to Evaluate Real-world Effectiveness of Hormonal Contraceptives in the Presence of Drug-drug Interactions.
Accurate estimation of conception is critical in the assessment of the effects of drugs used during pregnancy or to prevent pregnancy. In a novel application, we studied the effectiveness of oral contraceptives (OCs), where misclassification of conception relative to OC exposure may obscure effect estimates.
We studied OC failure, in a large claims database, among women who used antiepileptic drugs with metabolizing enzyme-inducing properties (carbamazepine or oxcarbazepine), which reduce OC's effectiveness or enzyme-neutral properties (lamotrigine or levetiracetam), with no expected impact on OC effectiveness. We compared conception rates in women 12-48 years of age concomitantly using OCs and enzyme-inducing drugs with rates in concomitant users of OCs and enzyme-neutral drugs. We measured conception with a validated algorithm that estimates gestational age based on pregnancy endpoints. We estimated relative and attributable risk using generalized estimating equation models after standardized mortality ratio weighting.
We identified 89,777 concomitant use episodes with adjusted contraceptive failure rates of 1.6 (95% confidence interval (CI) = 1.4, 1.8) per 100 person-years among users of enzyme-neutral drugs and 18,964 episodes with a rate of 2.3 (1.9, 2.8) among users of enzyme-inducing drugs. The relative risk of conception for enzyme-inducing group was 1.4 (1.1, 1.8), and the rate difference was 0.7 (0.2, 1.2).
OCs in combination with antiepileptic drugs that interact with metabolic enzymes were associated with increased contraceptive failure rates. Measurement of conception in claims data had adequate accuracy to uncover a strong drug-drug interaction, offering promise for broader application in comparative effectiveness studies on hormonal contraceptives to inform clinical and regulatory decisionmaking.
Observational study designs employing real-world data are commonly used to evaluate the safety and effectiveness of medications during pregnancy.1,2 To ensure accurate timing of drug exposure, pregnancy episodes are usually determined via a specific pregnancy endpoint, estimation of gestational age at the time of the endpoint, and imputation of the pregnancy start date.3–5 Although the International Classification of Disease (ICD) coding offers detail on gestational age at delivery, the codes are not consistently applied to medical encounter claims, thus leaving some uncertainty about the exact date of conception. Such inaccuracies are more pronounced for preterm deliveries and other adverse pregnancy endpoints such as stillbirths and abortions.6,7 Inappropriate timing of conception will in turn result in misclassification of drug exposure, and introduce bias. Misclassification bias will be most pronounced if the exposure window to be studied is close to conception (such as when evaluating effects of first-trimester exposure on the risk for malformation), or exposure pattern varies over time.8 It will be even more prominent in scenarios where the pregnancy was not intended or when studying drugs with safety concerns regarding use during pregnancy, and thus exposure is terminated as soon as pregnancy is discovered, leading to only a short period of fetal exposure during pregnancy.
Another research area where accurate timing of conception is critical is in the evaluation of contraceptive effectiveness. Contraceptive failure is operationalized as conception during exposure to the contraceptive agent, and if such failure occurs, the contraceptive use will be discontinued as soon as pregnancy is discovered. Delayed conception estimates may inaccurately conclude that the contraceptive was effective. On the other hand, estimates that time conception too early may conclude that the contraceptive failed even though it may have been discontinued because intentions to prevent pregnancy had changed. Thus, even small errors of conception estimation of only a few weeks, which would be expected for some live births and to a larger extent for nonlive pregnancies, may yield claims data unusable to evaluate the real-world effectiveness of contraceptives. In the previous study, we developed a pregnancy identification algorithm based on several validation studies6,7 to evaluate the effectiveness of regulatory requirements designed to prevent maternal exposure to teratogenic medications.9 We have recently expanded the algorithm to incorporate diagnosis and procedure codes from the ICD-10. In light of the above-described concerns about conception timing, we aimed to investigate whether pregnancy identification algorithm can accurately identify contraception failure. We chose a clinical scenario where a well-documented drug–drug interaction modifies the failure rate of hormonal contraceptives. In this scenario, possible misclassifications of exposure or outcome could obscure causal inferences when evaluating the drug–drug interaction. However, if shown to be sensitive to detect the drug–drug interaction, this approach would offer opportunities to study a broad range of clinical risk factors for unintended pregnancy using real-world data, including interactions involving hormonal contraceptives, and thus advance pharmacoepidemiologic methods.
METHODS
Study Design and Data Source
We conducted a cohort study using the IBM MarketScan Commercial Claims Databases (2005–2017). This database includes data on inpatient and outpatient medical encounters and pharmacy dispensing claims for a large sample of the privately insured population in the United States. The beneficiaries have encrypted identifiers in the database, which allows for longitudinal follow-up. This database is certified as de-identified data, and the present study was approved as exempt by the Institutional Review Board at the University of Florida (IRB approval number: 201801093).
Clinical Scenario
Carbamazepine and oxcarbazepine are among the first-generation antiepileptic drugs with several approved and off-label indications, including epilepsy and bipolar disorder.10 Both drugs are inducers of Cytochrome P450 3A4 (CYP3A4), with well-documented evidence for reduction of estrogen/progestin plasma levels of oral contraceptives (OCs), and ovulation pattern disruption.11–13 The enzyme induction effect of carbamazepine and oxcarbazepine is comparable and may result in approximately 50% change in area under the curve (AUC) of OC products. 12,14 Therefore, and because both antiepileptic drugs are associated with neural tube defects, clinical guidelines recommend against the use of OCs when using carbamazepine or oxcarbazepine to avoid unintended pregnancy.15,16 In contrast, newer antiepileptic drugs, including lamotrigine and levetiracetam, have minimal effect on CYP3A4 and are not expected to reduce OC efficacy.17,18 In the present study, we aimed to compare the rate of OC failure in two cohorts of OC users who had concomitant use of either a CYP3A4-inducer (i.e., carbamazepine or oxcarbazepine) or a CYP3A4-neutral (i.e., lamotrigine or levetiracetam) antiepileptic drugs.
Study Cohorts
We identified female patients of childbearing age (12–48 years old) who had at least one pharmacy claim for a combined OC with low-dose estrogen (<50 µg) or a progestin-only OC, referred to as OCs. We identified pharmacy claims for lamotrigine or levetiracetam to create a cohort of OC users without drug–drug interaction and extracted claims for carbamazepine or oxcarbazepine to create a cohort of OC users with drug–drug interaction. These four antiepileptic drugs were selected based on their profile of CYPA34 activity and potential clinical uses to create comparable study cohorts with regard to baseline clinical characteristics and pregnancy rates. We defined the cohort entry date (i.e., index date) as the first day of concomitant use of the drugs of interest and OCs and defined a look-back period of six months before the index date with continuous insurance enrollment to ascertain drug indications and other covariates. Patients were required to have at least one medical claim for epilepsy, bipolar disorder, or personality disorder (i.e., indications for antiepileptic drugs) during the look-back period. The indications were measured using coding algorithms developed by the Center for Medicare and Medicaid Services, Chronic Conditions Data Warehouse based on the ICD, ninth, and tenth versions, clinical modification (ICD9/10-CM) codes.19 We excluded patients if they had medical diagnoses for infertility, ovary dysfunction, or hirsutism in their look back period to rule out off-label indications for OCs. The list of ICD codes is provided in eAppendix 1; http://links.lww.com/EDE/B746.
Concomitancy Definition
We assumed drug exposure started at the prescription dispensing date and ended on the last day of the pharmacy-entered dispensed days’ supply. We defined “concomitancy” as overlapping exposure periods regardless of the order of drug dispensing for antiepileptic drugs or OCs. 20 We excluded concomitancy periods that had exposure to valproate sodium, topiramate, or phenytoin because of potential CYP3A4 activity and teratogenic effects, which may encourage patients to use a second contraceptive method (e.g., barrier methods). We also excluded concomitancy periods where we observed any other hormonal contraceptive agents, including long-acting reversible contraceptives (e.g., intrauterine devices), injectables, or high-dose estrogen OCs. For users of CYP3A4-neutral antiepileptic drugs, we excluded the concomitancy periods with CYP3A4-inducer drugs. Then we created concomitant use episodes for patients in each study group, and a gap of ≥14 days for either of the medications during an episode was allowed. If we observed a gap of more than 14 days concomitant use, the concomitancy episode ended on the last day of concomitant use. Patients were allowed to reenter the cohort if they had subsequent concomitant use episodes after their first observation period and met all inclusion criteria, including the availability of the 6-months look-back period to allow re-evaluation of baseline characteristics.
Outcome Definition
The study outcome was contraception failure defined as conception during a concomitancy period. Conception was estimated via the pregnancy identification algorithm that uses medical encounters with ICD-9/10-CM, Current Procedural Terminology, and Healthcare Common Procedure Coding System codes to identify specific pregnancy endpoints, including live birth, ectopic pregnancies, stillbirth, terminations, and prenatal screening visits.6,7,21,22 Once pregnancy episodes were identified, the algorithm estimated gestational age to calculate the last menstrual period (LMP). The conception date was assumed to be 14 days after the estimated LMP date. We provide more details on the pregnancy identification algorithm in eAppendix 1; http://links.lww.com/EDE/B746.
Covariates
We measured several demographic and clinical variables at baseline to assess the comparability of study cohorts, including patient age, residence region, insurance plan type, and relationship to the employee covered by the health plan (spouse, employee, children/other). Clinical variables measured during the 6-months look-back period included recent pregnancy history (based on any pregnancy endpoint), use of teratogenic drugs (with or without mandated pregnancy prevention programs), and a variety of clinical conditions that may affect OC efficacy.
Study Follow-up
All patients were required to have insurance enrollment for a minimum of 90 days after their concomitancy episode ended. This requirement allows the pregnancy identification algorithm to capture pregnancy-related medical encounters, which are then used to date conception. This minimum number of days was defined based on our previous work that showed approximately 90% of live deliveries in our database have prenatal visits within the first 90 days after conception. We followed each patient from the index date of each concomitancy episode until conception, infertility, ovary dysfunction, hirsutism diagnoses, initiation of a teratogenic drug, end of concomitancy, maximum of 3 years’ follow-up, or end of study (December 31, 2017).
Statistical Analysis
The unit of analysis was a concomitancy episode. We compared baseline demographic and clinical characteristics of each study cohort using a threshold of an absolute standardized difference (ASD) higher than 10% as clinically significant.23 To account for confounding, we used a logistic regression model to create an exposure propensity score and selected covariates into the model based on a literature review on potential risk factors for the study outcome.24 We used the common support region of the score to create weights to estimate the average treatment effect among the treated, also known as the standardized mortality ratio (SMR) weighting method.23,25 In this weighted pseudo-population, the confounding effect of measured covariates is eliminated, and the effect estimates can approximate the causal effect. We used these SMR weights in a generalized estimating equation model with a Poisson distribution and offset of follow-up time to compare contraception failure rates among the study cohort. We used a robust variance estimator to account for the clustering of episodes within the same patient. We conducted all data management and analyses using SAS 9.4 and SAS/STAT 15.1 (Cary, NC).
Sensitivity Analysis
We performed several sensitivity analyses to evaluate the robustness of the study findings. For the concomitancy definition, we changed the gap allowance to 1 or 7 days instead of 14 days. We also excluded concomitancy periods that overlapped with other moderate or strong inducers or inhibitors of CYP3A4 for more than 14 days (see eAppendix-1; http://links.lww.com/EDE/B746: eTable 1; http://links.lww.com/EDE/B746 and eTable 2; http://links.lww.com/EDE/B746). For the outcome definition, we varied the estimated conception date by ±14 days, limited the analysis to episodes indexed before 2015 (confining coding to ICD-9-CM), limited the outcome definition to only live birth episodes (which allows the most accurate conception date estimation). To evaluate the impact of more homogeneous comparison groups, we restricted the maximum follow-up time to 6 months, limited the analysis to only the first episode of concomitancy for each patient, and restricted drug initiation sequence to patients who initiated OC initiation while on antiepileptic drug treatment.
RESULTS
In the main analysis, we identified 89,777 concomitancy episodes involving CYP3A4-neutral antiepileptic drug and 18,964 episodes with the CYP3A4 inducers. Cohorts had similar age distributions with a mean age of 26.3 ± 8.5 years for the CYP3A4-neutral and 25.5 ± 9.2 for the CYP3A4-inducing drugs (Table 1). About 70% of women in each cohort had a bipolar disorder diagnosis, and 30% had an epilepsy diagnosis. We observed a high prevalence of anxiety (29% vs. 29%) and teratogenic medication use (30% vs. 28%) in the baseline period. The baseline covariates were balanced between study cohorts except for age, beneficiary status, and paralysis diagnosis (ASD < 15%). SMR weighting successfully balanced all measured characteristics (Table 1). Propensity score distributions, SMR weights, and hazard plots are available in eAppendix 1; http://links.lww.com/EDE/B746.
TABLE 1. Demographic and Clinical Characteristics of the Study Cohorts
Covariates Before SMR Weighting After SMR Weighting
Concomitant OC Plus Enzyme-neutral AED Episodesa (N = 89,777) Concomitant OC Plus Enzyme-inducing AED Episodesb (N = 18,964) ASD (%) Concomitant OC Plus Enzyme-neutral AED Episodesa (N = 18,973) Concomitant OC Plus Enzyme-inducing AED Episodesb (N = 18,964) ASD (%)
Age n (%) n (%) n (%) n (%)
<20 23,297 (26) 6337 (33) 16 6359 (34) 6337 (33) 0
20–29 36,339 (41) 6748 (36) 10 6741 (36) 6748 (36) 0
30–39 21,618 (24) 3918 (21) 8 3910 (21) 3918 (21) 0
≥40 8523 (10) 1961 (10) 3 1964 (10) 1961 (10) 0
Hypertension 3732 (4) 946 (5) 4 958 (5) 946 (5) 0
Hyperlipidemia 4114 (5) 1011 (5) 3 1021 (5) 1011 (5) 0
Obesity 1277 (1) 357 (2) 4 358 (2) 357 (2) 0
Epilepsy 23,309 (26) 5816 (31) 10 5853 (31) 5816 (31) 0
Bipolar disorder 65,116 (73) 13,081 (69) 8 13,057 (69) 13,081 (69) 0
Schizophrenia 2428 (3) 852 (5) 10 853 (4) 852 (5) 0
Depression 50,723 (57) 10,219 (54) 5 10,215 (54) 10,219 (54) 0
Personality Disorder 5606 (6) 1209 (6) 1 1210 (6) 1209 (6) 0
Anxiety 26,043 (29) 5544 (29) 1 5542 (29) 5544 (29) 0
Substance Use Disorder 4575 (5) 1284 (7) 7 1288 (7) 1284 (7) 0
Recent pregnancy (live birth) 2248 (3) 309 (2) 6 308 (2) 309 (2) 0
Recent pregnancy (termination) 507 (1) 104 (1) 0 103 (1) 104 (1) 0
Teratogenic drug without REMS 24,925 (28) 5588 (30) 4 5620 (30) 5588 (30) 0
Teratogenic drug with REMS 414 (1) 95 (1) 1 96 (0) 95 (1) 0
Charlson Comorbidity Index (CCI)3
≤1 87,223 (97) 18,93 (95) NA 18,127 (95) 18,093 (95) NA
1< 2554 (3) 871 (5) NA 846 (5) 871 (5) NA
Comorbidities
Myocardial infarction 31 (0) < 11 0 <11 < 11 0
Congestive Heart Failure 175 (0) 50 (0) 1 51 (0) 50 (3) 0
Vascular Disorder 187 (0) 51 (0) 1 54 (0) 51 (3) 0
Cerebrovascular Disorder 892 (1) 197 (1) 1 202 (1) 197 (1) 0
Pulmonary Disorders 7211 (8) 1705 (9) 3 1716 (9) 1705 (9) 0
Dementia 86 (0) 30 (0) 2 31 (2) 30 (0) 0
Paralysis 613 (1) 380 (2) 11 388 (2) 380 (2) 0
Diabetes w/o complications 1945 (2) 432 (2) 1 434 (2) 432 (2) 0
Diabetes with complications 175 (0) 40 (0) 0 40 (0) 40 (0) 0
Renal Disorders 263 (0) 64 (0) 1 67 (3) 64 (0) 0
Mild Liver Disorders 765 (1) 192 (1) 2 194 (1) 192 (1) 0
Severe Liver Disorders 22 (0) <11 0 <11 <11 0
Peptic Ulcer 193 (0) 58 (0) 2 60 (3) 58 (0) 0
Rheumatoid Disorders 585 (1) 156 (1) 2 160 (8) 156 (1) 0
AIDS 22 (0) <11 0 <11 <11 0
Malignancy 793 (1) 170 (1) 0 173 (9) 170 (1) 0
Metastatic Malignancy 39 (0) <11 0 <11 <11 0
Beneficiary status
Employee 32,469 (36) 5702 (30) 13 5698 (30) 5702 (30) 0
Spouse 14,208 (16) 2764 (15) 3 2677 (15) 2764 (15) 0
Child/other 43,100 (48) 10,498 (55) 15 10,509 (55) 10,498 (55) 0
Residence region
Northeast 16,228 (18) 3045 (17) 5 3047 (16) 3045 (16) 0
Northcentral 19,394 (22) 4363 (23) 4 4370 (23) 4363 (23) 0
South 35,322 (39) 8074 (43) 7 8071 (42) 8074 (43) 0
West 17,594 (20) 3225 (17) 7 3228 (17) 3225 (17) 0
Unknown 1239 (1) 257 (1) 0 257 (1) 257 (1) 0
Health plan type
COM 1525 (2) 479 (2) 6 482 (3) 479 (3) 0
HMO 12,955 (14) 2749 (14) 0 2733 (14) 2749 (15) 0
PPO 53,996 (60) 11,362 (60) 1 11,373 (60) 11,362 (60) 0
POS 6853 (8) 1594 (8) 0 1597 (8) 1594 (8) 0
CDHP 5887 (6) 1216 (6) 1 1218 (6) 1216 (6) 0
Other 8561 (9) 1564 (8) 5 1570 (8) 1564 (8) 0
ASD, absolute standardized difference; COM, comprehensive; HMO, health maintenance organization; PPO, preferred provider organization; POS, noncapitated point-of-service; CDHP, consumer-driven health plan; Other: includes capitated/partially capitated point-of-service, exclusive provider organization, high deductible health plan; REMS, risk evaluation and mitigation strategy.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam). After SMR weighting, the size of the pseudo-population in cohort A becomes comparable to cohort B.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
Episodes involving concomitant use of enzyme-neutral antiepileptic drugs had a slightly larger mean follow-up time of 96 days (vs. 79 days among women who used enzyme-inducers). Concomitancy periods ended with 400 conceptions among women who used enzyme-neutral antiepileptic drugs, resulting in a crude contraception failure rate of 1.7 events per 100 person–years. Women with concomitant use of OCs and enzyme-inducing antiepileptic drug had 94 conceptions with a crude contraception failure rate of 2.3 per 100 person–years (Table 2). Figure 1 shows unadjusted survival plots for contraception failure outcome. Approximately two-thirds of all conceptions were identified based on liveborn deliveries (both groups 63%), whereas abortions were the second prevalent pregnancy endpoint (28% in the enzyme-neutral group vs. 27% in the enzyme-inducing group) (Table 3).
TABLE 2. Relative and Absolute Risk of Oral Contraceptive Failure in the Presence or Absence of Drug–drug Interaction
Study Cohort Events Total Follow-up Time (Person-years) Incidence Rate (per 100 Person-years) Relative Risk Risk Difference
Unadjusted analysis
Concomitant OC plus enzyme-neutral AED episodesa 400 23,647 1.7 (1.5, 1.9) REF REF
Concomitant OC plus enzyme-inducing AED episodesb 94 4102 2.3 (1.9, 2.8) 1.4 (1.1, 1.7) 0.6 (0.1, 1.1)
Adjusted Analysis
Concomitant OC plus enzyme-neutral AED episodesa 400 23,647 1.6 (1.4, 1.8) REF REF
Concomitant OC plus enzyme-inducing AED episodesb 94 4102 2.3 (1.9, 2.8) 1.4 (1.1, 1.8) 0.7 (0.2, 1.2)
ASD, absolute standardized difference; OC, oral contraceptive.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam)—Reference cohort.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
TABLE 3. Pattern of Pregnancy Episodes that Contributed to Measurement of Contraception Failure
Study Cohort Full-term
n (%) Preterm
n (%) Post-term
n (%) Ectopic
n (%) Stillbirth
n (%) Spontaneous abortion
n (%) Induced abortion
n (%) Unknown outcome
n (%) Total
n
Concomitant OC plus enzyme-neutral AED episodesa 233 (58) 16 (4) 3 (1) 4 (1) 3 (1) 65 (16) 46 (12) 30 (8) 400
Concomitant OC plus enzyme-inducing AED episodesb 54 (57) 4 (4) 1 (1) 1 (1) 0 (0) 14 (15) 11 (12) 9 (10) 94
ASD, absolute standardized difference; OC, oral contraceptive.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam)—Reference cohort.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
The adjusted contraceptive failure rates were 1.6 (95% CI = 1.4, 1.8) per 100 person–years among users of enzyme-neutral drugs and 2.3 (1.9, 2.8) among users of enzyme-inducing drugs. The marginal models with SMR weights increased the unadjusted relative risk for contraception failure slightly from 1.4 (1.1, 1.7) to 1.4 (1.1, 1.8), comparing women who used enzyme-inducing antiepileptic drugs to those who used enzyme-neutral antiepileptic drug. Concomitant use of enzyme-inducing antiepileptic drug and OCs resulting in an additional 0.7 conceptions (0.2, 1.2) per 100 person–years of concomitant use.
All sensitivity analyses corroborated our findings (Table 4). Analyses with conceivably superior measurement of confounding (analysis number 6, 7, 9 in Table 4) showed as expected slightly larger relative risk estimates with 1.6 (1.2, 2.2) when restricting to the first concomitancy episode per patient, 1.5 (1.1, 2.1) with a fixed sequence of drug initiation, and 1.5 (1.2, 1.9) after eliminating concomitant use of all other potential CYP3A4 inducers/inhibitors.
TABLE 4. Sensitivity Analyses on the Definitions of Concomitancy, Outcome Measurement, and Study Design Features
Row Sensitivity Analysis Incidence Rate (per 100 Person-year) Adjusted Rate Ratio Adjusted Rate Difference
Concomitant OC Plus Enzyme-neutral AED Episodesa Concomitant OC Plus Enzyme-inducing AED Episodesb
1 Conception date altered by +14 days 1.5 (1.3, 1.7) 1.9 (1.5, 2.4) 1.3 (1.0, 1.6) 0.4 (0.0, 0.9)
2 Conception date altered by −14 days 1.9 (1.7, 2.1) 2.5 (2.1, 3.0) 1.3 (1.1, 1.7) 0.6 (0.1, 1.1)
3 Restriction to episodes with index date before 2015 (ICD-9 era) 1.6 (1.4, 1.8) 2.3 (1.9, 2.9) 1.6 (1.1, 1.8) 0.7 (0.2, 1.3)
4 Conception ascertained based on live birth only 1.0 (0.9, 1.2) 1.4 (1.1, 1.8) 1.4 (1.1, 1.9) 0.4 (0.2, 8.0)
5 Maximum follow-up time restricted to 6 months 1.7 (1.5, 1.9) 2.3 (1.9, 2.9) 1.4 (1.1, 1.8) 0.6 (0.1, 1.2)
6 Episodes restricted to first episode per patient 1.6 (1.4, 1.8) 2.5 (1.9, 3.3) 1.6 (1.2, 2.2) 0.9 (0.2, 1.6)
7 Episodes restricted to those where OC initiation follows AED use 1.5 (1.3, 1.8) 2.3 (1.8, 3.1) 1.5 (1.1, 2.1) 0.8 (0.1, 1.5)
8 Permissable gap in concomitancy 7 days 1.6 (1.5, 1.8) 2.2 (1.8, 2.7) 1.4 (1.1, 1.7) 0.6 (0.1, 1.0)
9 Permissable gap in concomitancy 1 day 1.6 (1.5, 1.8) 2.1 (1.7, 2.6) 1.3 (1.0, 1.6) 0.5 (0.0, 0.9)
10 Exclusion of all follow-up time with other CYP3A4 inducers or inhibitors 1.6 (1.4, 1.8) 2.4 (1.9, 2.9) 1.5 (1.2, 1.9) 0.7 (0.2, 1.3)
ASD, absolute standardized difference; ICD, International Classification of Disease; OC, oral contraceptive.
aCohort A: concomitant use of oral contraceptives and CYP3A4-neutral drugs (lamotrigine or levetiracetam)—Reference cohort.
bCohort B: concomitant use of oral contraceptives and CYP3A4-inducer drugs (carbamazepine or oxcarbazepine).
DISCUSSION
Our study found that women who concomitantly used OCs and CYP3A4-inducing antiepileptic drugs were 40% more likely to experience contraceptive failure compared to CYP3A4-neutral antiepileptic users. Our findings suggest that conception estimation was sufficiently accurate to identify this well-documented strong drug–drug interaction, and thus, our approach may be useful to generate real-world evidence on mechanisms of contraceptive failure, including the examination of interactions.
Because respective drug approval requirements are largely confined to pharmacokinetic studies, limited information exists on the clinical significance of drug–drug interaction involving OCs. A meta-analysis published in 2010 identified only pharmacokinetic studies that examined the potential for contraceptive failure among patients who use antiepileptic drugs.26 We identified one small cohort study published in 1979 that followed 41 epilepsy patients with concomitant use of antiepileptic drugs and OCs and reported ~2.9 failures per 100 person–years, which is slightly higher than in our cohorts.27 In our study, we observed a failure rate of 1.6 or 2.3 per 100-person years, depending on the type of antiepileptic drugs in terms of enzyme-inducing properties. The observed magnitude of the drug–drug interaction impact on failure rates in our study is biologically plausible based on evidence from pharmacokinetic evaluations, which recommend higher OC doses when used concomitantly with carbamazepine (e.g., 80–100 mcg of ethinyl estradiol that far exceeds the observed doses of 20–35 mcg observed in our data). 11,12 A clinical trial on healthy volunteers showed a 46.1% reduction in the AUC for levonorgestrel, and 44.5% for ethinyl estradiol when the OC was administrated with carbamazepine (600 mg). The study also reported more ovulations (5/10 cycles vs. 1/10 cycles) among carbamazepine users versus OC use alone.12 The diminishing effects of CYP3A4 inducers, including carbamazepine and oxcarbazepine, on the efficacy of OC is also emphasized in the guidance document for labeling of combined hormonal contraceptives.28
Because clinical trials on contraceptive failure are typically infeasible, observational studies, especially prospective cohorts, have a pivotal role.29–31 Broader availability of real-world data and advancements in pharmacoepidemiologic methods can facilitate comparative effectiveness studies and help to translate mechanistic findings into clinically significant outcomes. To the best of our knowledge, the present study is the first attempt to investigate contraceptive failure in claims data. To evaluate the performance of this approach, we designed our study based on a clinical scenario of decreased OC efficacy in the presence of a known drug–drug interaction with a well-studied potent enzyme inducer. We successfully replicated mechanistic findings regarding a significant interaction between OC use and the perpetrator drug (a CYP3A4-inducer), resulting in contraceptive failure. Therefore, we envision that this approach could serve as a novel platform to evaluate the comparative effectiveness of hormonal contraceptives, considering different routes of administration and differences in pharmacokinetic profiles, among diverse patient populations with comorbidities or other concomitant medications. However, researchers should be vigilant about potential misclassification biases in exposure (contraceptive use) or outcome (conception) measurements and their consequences on the ability to make causal inferences, especially, for quantifying the magnitude of risk, and the possibility of inadequate sensitivity of our approach to detect weaker drug–drug interaction effects.
We operationalized exposure in our claims data by using “days of supply” recorded on the pharmacy claims. Although this measurement approach is more reliable than patient self-report, it may not be fully reflective of the actual medication consumption.32 For instance, women might discontinue OC treatment before their supply is exhausted to plan for pregnancy. This decision would possibly result in misclassification of exposure and overestimation of the contraception failure rate. The use of active comparator groups that exhibit similar demographics and comorbidities, such as demonstrated in our cohort, may mitigate some of these concerns, but even nondifferential exposure misclassification could bias relative risk estimates either away or toward the null hypothesis (eAppendix 2; http://links.lww.com/EDE/B747).
Regarding outcome measurement, we relied on an algorithm to infer pregnancy episodes from medical encounters with the healthcare system that use validated coding to specify pregnancy endpoints, but without an actual recording of LMP. Based on previous literature, estimation of conception will have varying degrees of accuracy for each type of pregnancy endpoint.6,22 In our pregnancy identification algorithm, we assigned gestational age for live-birth episodes based on an algorithm of ICD codes indicating gestational age or preterm status, and the overall agreement against birth certificates is reported to be >93% in the Medicaid database.22 For ectopic pregnancy, induced/spontaneous abortion, and stillbirth, we used fixed values for gestational age (8, 10, and 28 weeks, respectively) similar to previous literature.6,7 This approach has shown a moderate agreement both against medical charts (70% for ectopic pregnancy and 67% for spontaneous abortions within four weeks)7, and a cohort of in-vitro fertilization patients (77% agreement for ectopic pregnancy, 47% for stillbirth, and 36% for abortions cases).6 Validation studies on pregnancy identification algorithms typically report the agreement between the estimated LMP and the gold standard (e.g., the clinical estimate of gestational age on the birth certificate) within a prespecified number of weeks (e.g., 2 weeks) as the margin of error.4,6,22 It should be noted; however, that such a margin might need to be varied according to the pregnancy endpoint that was used to estimate LMP. For example, previous studies have shown that the capture of preterm status among deliveries improves with decreasing gestational age.33 Thus, considering the relationship between measurement sensitivity and gestational age as well as the pronounced left-skewed distribution of gestational age among preterm infants, error margins of 2–4 weeks used in sensitivity analyses may be appropriate. In contrast, stillbirth with a flat gestational age frequency distribution ranging from 20 to 42 weeks may require broader margins that should be tested.34 We should also note that available conception algorithms typically estimate LMP, following long-established conventions in timing gestational age, and thus, conception must be imputed as LMP + 14 days to operationalize contraceptive failure.35 In our study, we conducted sensitivity analyses by varying the estimated conception date, limiting the analysis to the ICD-9-CM era with previously validated pregnancy endpoint definitions, and restricting the events to pregnancies with the liveborn outcome. The two latter analyses aimed to increase the specificity of the outcome measure.
Our adjustment for potential confounding was based on a literature review on risk factors for contraceptive failure, but we acknowledge that several predictive factors were not or were only partially measurable in claims data. For example, sexual activity may be an important factor for contraceptive failure but is not available in claims data. However, we believe that our use of an active comparator group successfully balanced for the majority of risk factors, as exemplified by fairly well-balanced comparison groups before SMR weighting. We should also acknowledge that race and socioeconomic status were unmeasured in our dataset and could act as confounding factors. However, the study drugs are available on the market as generic products, yielding channeling for economic reasons unlikely. Finally, we should note that our adjustment for confounding further increased the observed relative risk estimates, thus suggesting that enzyme inducers were slightly more prevalent among patients with fewer risk factors for contraceptive failure. Thus, to explain our findings, any unmeasured covariate would need to be distributed in the opposite direction than the measured risk factors. Researchers could apply probabilistic bias analysis methods to evaluate the impact of exposure, outcome, and confounder misclassifications, simultaneously.36,37
In conclusion, women who use OCs in combination with carbamazepine or oxcarbazepine should be aware of increased contraceptive failure rates. Our study showed that measurement of conception in claims data has adequate accuracy to reveal the effect of a known drug–drug interaction with hormonal contraceptives. Our pharmacoepidemiologic approach is promising for comparative effectiveness studies on hormonal contraceptives to generate real-world evidence and inform clinical and regulatory decision-making.
FIGURE 1. Survival curves for contraception failure during follow-up in each study cohort.
ACKNOWLEDGMENTS
None.
Supplementary Material
A.S., J.B., S.S., and A.W. participated in study conceptualization and design. A.S., J.B., A.G., B.C., C.H., P.P., P.S., and A.W. involved in data acquisition, analysis, and interpretation. A.S. participated in drafting the article. A.S., J.B., A.G., B.C., C.H., P.P., P.S., and A.W. participated in the critical revision and final approval of the manuscript.
The results reported herein correspond to specific aims of grant #OPP1185454 to Stephan Schmidt and Joshua Brown from the Bill & Melinda Gates Foundation.
A.G.W. has active research support from Merck & Co, which is unrelated to this project. The other authors have no conflicts to report.
The IBM MarketScan database is available via a data use agreement with the company. | LEVETIRACETAM | DrugsGivenReaction | CC BY | 33196560 | 18,620,305 | 2021-03-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Death'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, CARBOPLATIN, CISPLATIN, ETOPOSIDE, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, OXALIPLATIN, PACLITAXEL, PANITUMUMAB | DrugsGivenReaction | CC BY-NC | 33197299 | 18,711,884 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Disease progression'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, LEUCOVORIN CALCIUM, OXALIPLATIN | DrugsGivenReaction | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Gene mutation'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, CARBOPLATIN, CISPLATIN, ETOPOSIDE, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, OXALIPLATIN, PACLITAXEL, PANITUMUMAB | DrugsGivenReaction | CC BY-NC | 33197299 | 18,711,884 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Malignant neoplasm progression'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, LEUCOVORIN CALCIUM, OXALIPLATIN | DrugsGivenReaction | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Metastatic squamous cell carcinoma'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, CARBOPLATIN, CISPLATIN, ETOPOSIDE, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, OXALIPLATIN, PACLITAXEL, PANITUMUMAB | DrugsGivenReaction | CC BY-NC | 33197299 | 18,711,884 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Neuroendocrine carcinoma metastatic'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, CARBOPLATIN, CISPLATIN, ETOPOSIDE, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, OXALIPLATIN, PACLITAXEL, PANITUMUMAB | DrugsGivenReaction | CC BY-NC | 33197299 | 18,711,884 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Neuroendocrine carcinoma'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, LEUCOVORIN CALCIUM, OXALIPLATIN | DrugsGivenReaction | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Product use in unapproved indication'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, LEUCOVORIN CALCIUM, OXALIPLATIN | DrugsGivenReaction | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Squamous cell carcinoma'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, LEUCOVORIN CALCIUM, OXALIPLATIN | DrugsGivenReaction | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Therapeutic product effect incomplete'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, CARBOPLATIN, CISPLATIN, ETOPOSIDE, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, OXALIPLATIN, PACLITAXEL, PANITUMUMAB | DrugsGivenReaction | CC BY-NC | 33197299 | 18,711,884 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Therapeutic response decreased'. | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | BEVACIZUMAB, FLUOROURACIL, IRINOTECAN, LEUCOVORIN, LEUCOVORIN CALCIUM, OXALIPLATIN | DrugsGivenReaction | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
What was the dosage of drug 'BEVACIZUMAB'? | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | UNK(4 COURSES) | DrugDosageText | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
What was the dosage of drug 'FLUOROURACIL'? | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | UNK(4 COURSES) | DrugDosageText | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
What was the dosage of drug 'IRINOTECAN'? | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | UNK(4 COURSES) | DrugDosageText | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
What was the dosage of drug 'LEUCOVORIN CALCIUM'? | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | UNK(4 COURSES) | DrugDosageText | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
What was the dosage of drug 'OXALIPLATIN'? | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | UNK(4 COURSES) | DrugDosageText | CC BY-NC | 33197299 | 18,952,120 | 2021-01 |
What was the outcome of reaction 'Death'? | Mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon: detailed molecular characterisation of two cases indicates a distinct colorectal cancer entity.
We present two rare cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon. A literature search revealed only three published cases with similar histology but none of these reports provided profound molecular and mutational analyses. Our two cases exhibited a distinct, colon-like immunophenotype with strong nuclear CDX2 and β-catenin expression in more than 90% of the tumour cells of both components. We analysed the two carcinomas regarding microsatellite stability, RAS, BRAF and PD-L1 status. In addition, next-generation panel sequencing with Ion AmpliSeq™ Cancer Hotspot Panel v2 was performed. This approach revealed mutations in FBXW7, CTNNB1 and PIK3CA in the first case and FBXW7 and RB1 mutations in the second case. We looked for similar mutational patterns in three publicly available colorectal adenocarcinoma data sets, as well as in collections of colorectal mixed neuroendocrine-non-neuroendocrine neoplasms (MiNENs) and colorectal neuroendocrine carcinomas. This approach indicated that the FBXW7 point mutation, without being accompanied by classical adenoma-carcinoma sequence mutations, such as APC, KRAS and TP53, likely occurs at a relatively high frequency in mixed neuroendocrine and squamous cell carcinoma and therefore may be characteristic for this rare tumour type. FBXW7 codifies the substrate recognition element of an ubiquitin ligase, and inactivating FBXW7 mutations lead to an exceptional accumulation of its target β-catenin which results in overactivation of the Wnt-signalling pathway. In line with previously described hypotheses of de-differentiation of colon cells by enhanced Wnt-signalling, our data indicate a crucial role for mutant FBXW7 in the unusual morphological switch that determines these rare neoplasms. Therefore, mixed large cell neuroendocrine and a squamous cell carcinoma can be considered as a distinct carcinoma entity in the colon, defined by morphology, immunophenotype and distinct molecular genetic alteration(s).
Introduction
Neuroendocrine carcinomas of the colorectum are rare and highly aggressive tumours with poor clinical outcome. Their incidence is 0.1–0.6% [1, 2]. The percentage of pure squamous cell carcinoma among all colorectal carcinomas is even lower [3, 4]. Here we present two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colon. Previously, only three cases with an identical histology were described in the caecum, rectum and the descending colon [5, 6, 7], but extensive immunohistochemical and molecular profiling was not performed. This is the first report of this rare type of carcinoma that also defines its typical molecular genetic features. Combined neuroendocrine and squamous cell carcinomas also occur in organs with original squamous epithelium, such as the maxillary sinus or the oesophagus [8, 9]. Such neoplasms biologically present tumour development via stages of increasing atypia. On the contrary, mixed neuroendocrine and squamous cell carcinomas in the colon represent a different kind of tumour emergence. In our opinion, these rare carcinomas might be the outcome of progressive malignant transformation of mixed neuroendocrine‐non‐neuroendocrine neoplasms (MiNENs), formerly termed mixed adenoneuroendocrine carcinomas (MANECs) [10]. In accordance with this hypothesis, single cases with an additional squamous carcinoma component are known among high‐grade MiNENs in the colorectum [11]. Alongside accurate morphological evaluation, molecular classification of colorectal cancers with high grade morphology, via immunohistochemistry of mismatch repair proteins and mutational analyses of BRAF and other genes, has proven essential to provide best guidance for patient treatment and therapeutic outcome. Hence, we carefully analysed the present lesions morphologically and immunohistochemically. In order to better understand the pathophysiological mechanisms underlying these rare neoplasms, we additionally applied next‐generation sequencing and compared the mutational results to data sets of classical colorectal adenocarcinoma as well as MiNEN and neuroendocrine carcinomas of the colorectum. Based on next‐generation panel sequencing data and immunohistochemical analyses, our data indicate that mixed neuroendocrine and squamous cell carcinoma may be a distinct new colon cancer entity.
Materials and methods
Tumour specimens, histology and immunohistochemistry
This study was conducted according to the recommendations of the ethics committee of the Medical Faculty of the Ludwig‐Maximilians‐University Munich, Germany and the standards set in the declaration of Helsinki 1975. Archival tissue from two formalin‐fixed and paraffin‐embedded (FFPE) cases of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma were accessed from the Institute of Pathology in Bayreuth as well as from a practice of pathology in Munich. The neoplasms were resected in 2014 (first case) and 2017 (second case). Sections of 5 μm were cut, deparaffinised and stained with H&E for histological preparation. For immunohistochemistry, sections were incubated with prediluted mouse anti‐β‐catenin (14, ready to use, Ventana), rabbit mouse anti‐CK5/6 (D5/16B4, ready to use, Ventana), mouse anti‐MSH‐2 (G219‐1129, ready to use, Ventana), rabbit anti‐MSH‐6 (SP93, ready to use, Ventana), mouse anti‐PMS‐2 (A16‐4, ready to use, Ventana), rabbit anti‐PDL‐1 (SP263, ready to use, Ventana), mouse anti‐CD56 (123C3, ready to use, Ventana), rabbit anti‐synaptophysin (MRQ‐40, ready to use, Ventana), mouse anti‐chromogranin A (LK2H10, ready to use, Ventana), mouse anti‐neuron‐specific enolase (NSE; BBS/NC/VI‐H14, 1:200, Dako, Santa Clara, CA, USA), rabbit anti‐CDX2 (EPR2764y, 1:50, Medac; Bio‐Genex), mouse anti‐MLH‐1 (ES05, 1:100, Leica, Wetzlar, Germany), rabbit anti‐NUT (C52B1, 1:75, Cell Signaling), mouse anti‐p63 (BC4A4, 1:100, Zytomed; Biocare Medical, Pacheco, CA, USA), mouse anti‐p40 (BC28, 1:100, Zytomed, Berlin, Germany), mouse anti‐TTF‐1 (8G7G3/1, 1:200, Agilent, Santa Clara, CA, USA), or mouse anti‐Ki67 antibody (MIB‐1, 1:150, Dako). For staining, a Ventana Benchmark XT autostainer was used. Detection was performed with either ultraView Universal DAB detection kits or optiView DAB IHC detection kits (Ventana Medical Systems, Tuscon, AZ, USA).
DNA extraction and pyrosequencing
To identify tumour areas, we used sections stained with H&E, which were subsequently used as templates to isolate areas of the combined large cell neuroendocrine and squamous cell carcinoma under microscopic control from deparaffinised serial sections using sterile scalpel blades. Neuroendocrine and squamous components were not micro‐dissected separately. Tumour DNA was extracted with QIAamp DNA Micro Kits and GeneRead DNA FFPE Kits (Qiagen, Hilden, Germany) for consecutive analyses of KRAS, NRAS and BRAF V600E gene mutations as well as panel sequencing, respectively. The mutational status of KRAS exon 2–4, NRAS exon 2–4 and BRAF V600E was analysed by pyrosequencing on a PyroMark Q24 Advanced instrument (Qiagen), as previously described [12].
Panel sequencing
The Ion AmpliSeq Cancer Hotspot Panel v2, covering the mutation hotspots of 50 oncogenes and tumour suppressor genes (Life Technologies, Calsbad, CA, USA), was used for next‐generation panel sequencing following the manufacturer's protocol. 10 ng of Qubit quantified DNA was used for library generation with Ion AmpliSeq Library Kits and Ion Xpress Barcode Adapters (Thermo Fisher, Calsbad, CA, USA). After emulsion PCR and bead purification, multiplexed libraries were then loaded onto 318 chips, and sequenced on an Ion Personal Genome Machine (all Thermo Fisher). For data analysis, sequence reads were mapped to human reference genome hg19 and filtered for non‐synonymous variants using Ion reporter software v5.0 (Thermo Fisher). Annotations, information on pathogenesis and population allele frequencies were retrieved from Ensembl VEP (www.ensembl.org/Homo_sapiens/Tools/VEP).
Results
Case presentations
Case 1
Clinical data and pathological findings
A 51 year old male patient with known ulcerative colitis presented with rectal bleeding and diarrhoea, leading to the diagnosis of a tumour in the sigmoid colon followed by complete surgical resection. The 8 cm large, ulcerated tumour caused luminal stenosis and infiltration of the entire wall into the surrounding adipose tissue. Histology revealed lymphangiosis carcinomatosa, venous invasion and three lymph node metastases. Resection margins were free of tumour cells. Samples showed no signs of ulcerative colitis.
The carcinoma showed a solid growth pattern without gland formation or mucin production. In central areas, the tumour cells exhibited distinct squamous differentiation, whereas large tumour cells in the marginal zone exhibited no specific differentiation. Profound atypia, high rates of apoptosis, and numerous atypical mitoses, with Ki‐67 labelling index up to 90%, were present. Immunohistochemistry revealed strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells. Cells with squamous differentiation were positive for cytokeratin 5/6 and p63, whereas the large tumour cells without specific differentiation showed strong positivity for synaptophysin and neuron specific enolase (NSE). Morphological and immunhistochemical findings are shown in Figure 1 and supplementary material, Figure S1. All tumour cells were negative for CD56, chromogranin A, p40 and TTF‐1. To distinguish the lesion from NUT (nuclear protein in testis) midline carcinoma (NMC), we performed NUT immunohistochemistry, which was negative. Immunohistochemistry for hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in all tumour cells, characterising the neoplasm as a microsatellite stable tumour. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon, pT3, pN1a (3/17), V1, L1, Pn0 was diagnosed.
Figure 1 Morphological and immunohistochemical characteristics of the first case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is marked with a dotted line and foci of keratinisation are highlighted by arrows (E). The neoplasm shows intense staining of CDX2 (F). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Within the following months of disease, distant metastasis to the liver and the abdominal wall occurred (pM1c [HEP, OTH]) resulting in a final UICC‐stage IVC. Therapy with three courses of panitumumab plus FOLFOX 6, two courses of cisplatin and etoposide and later four courses of bevacizumab and FOLFOXIRI was performed.
Molecular pathology
Because of insufficient therapeutic response, immunohistochemistry for PDL1 and molecular genetic analysis were carried out. PDL1 expression was not detectable in carcinoma cells or in the surrounding stroma. No mutations were present in exons 2, 3 and 4 of the KRAS and NRAS genes and in exon 15 of the BRAF gene. Next‐generation sequencing analysis surveying hotspot regions of 50 oncogenes and tumour suppressor genes detected CTNNB1 (c.110C>G, p.Ser37Cys), PIK3CA (c.1173A>G, p.Ile391Met) and FBXW7 (c.1393C>T, p.Arg465Cys) mutations.
Follow up
The tumour progressed rapidly under bevacizumab plus FOLFOXIRI therapy. Chemotherapy was changed to paclitaxel, carboplatin and palliative care. The patient died 1 year after initial diagnosis of the tumour.
Case 2
Clinical data and pathological findings
A 46 year old female patient without relevant pre‐existing conditions underwent colonoscopy due to diarrhoea with admixed blood. A tumour in the sigmoid colon was found and complete surgical resection performed. The resection specimen showed a 2.5 cm ulcerated tumour. Histology revealed a high‐grade carcinoma with solid growth devoid of glandular differentiation. The transmural infiltration involved the serosa. Five regional lymph node metastases were detected. Lymphangiosis carcinomatosa and venous invasion were present. Resection margins were free of tumour cells. PET‐CT scanning showed diffuse liver metastases.
The histology of the carcinoma exhibited clusters of squamous tumour cells showing immunohistochemical expression of cytokeratin 5/6, but not p63 or p40. A second tumour component showed solid and trabecular growth of large carcinoma cells with strong immunohistochemical expression of synaptophysin and CD56, but negativity for chromogranin A and NSE. All tumour cells exhibited strong cytoplasmic expression of nuclear β‐catenin and CDX2. The mitotic rate was high and the Ki‐67 proliferation index was 80% of tumour cells (Figure 2). No TTF‐1 and NUT expression was detectable by immunohistochemistry. Analysis of hMLH1, hMSH2, hMSH6 and hPMS2 showed nuclear expression in tumour cells. In summary, a mixed large cell neuroendocrine and squamous cell carcinoma of the sigmoid colon devoid of microsatellite instability was diagnosed. The following staging was reported: pT4a, pN2a (5/19), cM1a (HEP), L1, V1, Pn0, R0, UICC‐stage IVA.
Figure 2 Morphological and immunohistochemical characteristics of the second case of colorectal combined large cell neuroendocrine carcinoma and squamous cell carcinoma pictured in overview (A) and close‐up view (B–H). Examples of neuroendocrine differentiation are shown by immunostaining for synaptophysin (accentuated in marginal areas; C). Tumour cells exhibit strong expression of β‐catenin (D). The squamous component is again marked with dotted lines (E). The overview shows intense staining of CDX2 in tumor and remaining normal colon mucosa (F; asterisk). Examples of squamous differentiation as well as proliferation are shown by immunostaining for CK5/6 (accentuated in central areas; G) and Ki67 (H), respectively.
Molecular pathology
Next‐generation sequencing analysis revealed a FBXW7 (c.1393C>T, p.Arg465Cys) point mutation, as was also true for the first analysed case. In addition, a RB1 (c.2284C>T, p.Gln762Ter) mutation was found. In contrast to the first case, no CTNNB1 and PIK3CA mutations were detected.
Follow up
In accordance with standard guidelines and results from the NORDIC NEC study [13], therapy with five cycles of cisplatin and etoposide followed. Follow‐up PET‐CT scanning showed complete remission of liver metastasis. Three years later one new liver metastasis with strong immunohistochemical expression of NSE was successfully ablated by local brachytherapy.
Data set analyses
Genomic data analysis on three publicly available colorectal adenocarcinoma cohort data sets was performed, employing the cBioPortal as a cancer genomics tool. The TCGA Nature 2012 Study, the updated TCGA Pan Cancer Atlas Study on CRC, and the MSKCC 2018 Cancer Cell Study for metastatic colorectal cancer [14, 15, 16, 17, 18] were screened for other cases with FBXW7, CTNNB, PIK3CA and RB1 mutations. Our search revealed 5–8% CTNNB1 mutations, 13–17% FBXW7 mutations, 20–28% PIK3CA mutations and 3–5% RB1 mutations, respectively. As expected, the classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, outnumber those findings by far (Table 1). In addition, we screened for significant co‐occurrences or mutual exclusivities between FBXW7, CTNNB1, PIK3CA and RB1 mutations in all three data sets, which mostly consist of classic adenocarcinoma cases, in order to explore possible mutational correlations that could potentially also occur in the scarce mixed neoplasms described here. Here again we included most common classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, for comparison. Referring to these, we detected significant co‐occurrence of APC and KRAS and APC and TP53 in two of three data sets. In addition, mutations in the genes coding for APC and CTNNB1 as well as TP53 and PIK3CA related to the classical adenoma–carcinoma sequence were found to be mutually exclusive. Importantly, significant co‐occurrence of FBXW7 and PIK3CA as well as FBXW7 and RB1 mutations, as was found in the scarce neoplasm type described here, was identified in two of the three data sets (Table 2). This points to functional importance of these two mutational interactions also in classical adenocarcinomas. To define similarities and differences between classical colorectal adenocarcinomas, mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum, colorectal MANECs and pure colorectal neuroendocrine carcinomas, we compared frequencies of genetic alterations between those entities (Table 3). In the two cases of mixed large cell neuroendocrine and squamous cell carcinoma described here, and in contrast to MiNENs and classic adenocarcinomas, we noted the absence of APC, KRAS and TP53 mutations, as well as the occurrence of mutations in the FBXW7 gene in both tumours. The frequency of mutations in FBXW7 in particular was markedly lower (16–25%) in classic adenocarcinomas and MiNENs (Table 3), although we cannot exclude the existence of FBXW7 wild‐type, mixed neuroendocrine and squamous cell carcinoma cases from our case report on only two individuals affected by this very rare tumour type. Given that tissue images of colorectal carcinoma cases with FBWX7 mutation were available via cBioPortal within the TCGA Nature 2012 study, these were screened for unusual morphology, such as squamous or neuroendocrine differentiation. However, only two of the reviewed 35 cases showed a tendency toward neuroendocrine differentiation, and none of those had relevant morphological features which would have pointed towards squamous differentiation. Hence, other factors, such as the cell of tumour origin or epigenetic peculiarities might also be needed which, presumably in collaboration with mutant FBXW7, contribute to the occurrence of this very rare, mixed colorectal cancer entity.
Table 1 Gene alteration frequencies in colorectal adenocarcinoma data sets.
Genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC
76 75 76
CTNNB1
5 7 8
FBXW7
17 17 13
KRAS
42 42 45
PIK3CA
20 28 20
TP53
53 60 73
RB1
3 5 3
Values indicate the frequency of gene alterations (in percent) in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). Classical adenoma–carcinoma sequence mutations, such as APC, KRAS and TP53, are highlighted in orange.
Table 2 Co‐occurrences and mutual exclusivities of mutated genes in colorectal adenocarcinoma data sets.
Significant co‐occurrence Significant mutual exclusivity
Mutated genes TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study TCGA Nature 2012 Study TCGA Pan Cancer Atlas Study MSKCC 2018 Cancer Cell Study
APC and CTNNB1
0 0 0 0 1 (0.014) 1 (<0.001)
APC and KRAS
0 1 (<0.001) 1 (0.014) 0 0 0
APC and PIK3CA
0 0 1 (0.019) 0 0 0
APC and TP53
0 1 (<0.001) 1 (0.022) 0 0 0
CTNNB1 and FBXW7
0 1 (<0.001) 0 0 0 0
CTNNB1 and PIK3CA
0 1 (<0.001) 0 0 0 0
CTNNB1 and RB1
0 1 (<0.001) 0 0 0 0
FBXW7 and KRAS
0 0 1 (0.001) 0 0 0
FBXW7 and PIK3CA
0 1 (0.012) 1 (<0.001) 0 0 0
FBXW7 and TP53
0 0 0 0 0 1 (0.013)
FBXW7 and RB1
0 1 (0.014) 1 (0.001) 0 0 0
KRAS and PIK3CA
1 (<0.001) 1 (<0.001) 1 (<0.001) 0 0 0
KRAS and TP53
0 0 0 0 0 1 (<0.001)
PIK3CA and TP53
0 0 0 0 1 (<0.001) 1 (<0.001)
Values indicate the existence (1) or non‐existence (0) of significant co‐occurrence, or significant mutual exclusivity between the listed mutated genes in three different data sets according to The Cancer Genome Atlas Program 2012 (TCGA, [16]), TCGA Pan Cancer Atlas Study [17] and Memorial Sloan Kettering Cancer Center Study (MSKCC, [18]). No significant finding is shown in red, significant correlation in one data set is marked in orange and significant findings in two or more data sets are highlighted in green. P values are indicated in parenthesis.
Table 3 Mutations in colorectal neoplasms.
Entity AC MiNEN MiNEN NEC NEC Combined large cell neuroendocrine carcinoma and squamous cell carcinoma
Source TCGA, 2012 Woischke et al, 2017 Jesinghaus et al, 2017 Woischke et al, 2017 Jesinghaus et al, 2017 Present study
Number of cases 269 6 19 4 8 2
Mutations
AKT1 0 0 25 0
APC 61 83 16 75 63 0
ATM 4 0 14 50 0
BRAF 8 16 37 25 25 0
CTNNB1 1 (1 out of 2 cases)
EGFR 2 16 25 0
ERBB4 0 0 25 0
FBXW7 12 16 16 25 (2 out of 2 cases)
FGFR2 0 0 25 0
FLT3 5 0 25 0
GNAS 0 0 25 0
HRAS 0 0 25 0
IDH1 0 16 0 0
IDH2 1 0 25 0
JAK2 1 0 25 0
KDR 0 16 25 0
KRAS 35 83 21 100 25 0
MET 0 33 50 0
NOTCH1 0 33 25 0
PIK3CA 16 50 5 25 (1 out of 2 cases)
PTEN 5 0 11 0 0
PTPN11 1 0 25 0
RB1 1 16 50 (1 out of 2 cases)
RET 0 33 0 0
SMAD4 10 0 5 25 0
SMO 0 0 25 0
TP53 45 100 47 75 63 0
VHL 0 16 25 0
Frequencies of genetic alterations (in percent) of colorectal adenocarcinomas (AC), MiNENs, neuroendocrine carcinomas (NEC) in three studies (The Cancer Genome Atlas Program 2012 (TCGA, [16]), Jesinghaus et al [48] and Woischke et al [47]) in comparison with the genetic alterations of the two cases of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma. Regarding TCGA cases, only putative driver mutations are included. Frequencies are highlighted by a coloured scale ranging from 0% (yellow) to 100%, or out of two for the category of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma (green).
Discussion
In this study, we analysed two mixed large cell neuroendocrine and squamous cell carcinomas of the colorectum by next‐generation sequencing and compared the results with data from three publicly available colorectal adenocarcinoma data sets, as well as from cohorts of colorectal MiNENs and colorectal neuroendocrine carcinomas. This approach revealed a shared FBXW7 mutation and a lack of classical adenoma–carcinoma sequence mutations in both of our cases. This is in contrast to classic adenocarcinomas and MiNENs and therefore represents a molecular signature, which, together with the unique morphological features, may distinguish mixed neuroendocrine carcinoma and squamous carcinoma of the colorectum from other colorectal cancer types. Neuroendocrine carcinomas of colorectal origin represent very rare but highly aggressive tumours with a poor prognosis [1, 2]. Nevertheless, pure squamous cell carcinomas have been reported at an even lower incidence [3, 4, 19]. Since the first pure squamous cell carcinoma in the colorectum was reported by Schmidtmann in 1919 [20], profound literature research provided only 75 more cases to date, stating this neoplasm as extremely rare, with frequencies of 0.1–0.25% of all colorectal carcinomas [3, 4, 19]. Possible causes for this squamous colonic carcinoma are chronic inflammation in the context of ulcerative colitis, schistosomiasis, human papillomavirus infection, abdominal sinus or fistula, or pelvic radiation [4, 21]. Associations between neuroendocrine carcinomas or MiNEN of the colon and ulcerative colitis, as seen in case 1, are sporadically reported [22, 23]. The combination of the two neoplasm types in the colorectal region is highly exceptional and so far very little is known about the underlying mutational landscape of such combined carcinomas. In accordance with the new World Health Organization Classification from 2019, mixed large cell neuroendocrine carcinoma and squamous cell carcinoma in the colorectum is subsumed under the category of MiNENs, formerly named MANECs, in which each component accounts for ≥30% of the neoplasm [24]. Although three case reports of mixed neuroendocrine carcinoma and squamous cell carcinoma of the colorectum in literature do exist [5, 6, 7], only one of those has been assessed for microsatellite stability. In addition, one study examined the mutational status of KRAS and BRAF [5]. However, none of these cases has been analysed regarding its underlying genetic background via next‐generation sequencing. Thus, we performed for the first time next‐generation sequencing‐based multigene panel analysis of mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon.
Our two cases contain several remarkable similarities. One is the striking morphology, showing squamous carcinoma cells in central areas and poorly differentiated large cell neuroendocrine carcinoma in marginal areas, each component accounting for >30% of the tumour. The squamous cell differentiation was demonstrated not only by morphological features, such as intercellular bridges and focal keratinisation, but also by immunohistochemical expression of cytokeratin 5/6 and/or p63, with p63 being positive only in case 1. Cytokeratin 5/6 shows a sensitivity of 84% and a specificity of 79% in the diagnosis of squamous cell carcinoma, and p63 exhibits similar diagnostic performance, with a sensitivity of 81–84% and specificity of 85% [25, 26]. Neuroendocrine differentiation was confirmed by strong immunohistochemical positivity for synaptophysin, which has been approved as the best single marker for neuroendocrine tumours [27]. In accordance with one previous study, we found remarkably strong nuclear expression of CDX2 and β‐catenin in over 90% of tumour cells of both carcinoma cases as well as in both components (neuroendocrine and squamous) of the tumours [7]. The high nuclear abundance of β‐catenin detected here in large cell neuroendocrine carcinomas is very exceptional, but has been reported previously [11]. Besides clinical and morphological aspects, the strong nuclear CDX2 expression detected in the vast majority of carcinoma cells indicates the colon as the primary origin of the lesion, since CDX2 is known as a reliable marker for cancers of intestinal origin [28]. Despite the young age of the patients, both carcinomas were microsatellite stable (MSS), excluding Lynch syndrome.
In one of the cases, we identified a CTNNB1 mutation, which is a key factor in the Wnt signalling pathway and well described in the development of colorectal carcinomas [29, 30]. In one of our cases, there was a mutation in the tumour suppressor gene RB1, which are present in 5.8% of all colorectal cancers (14, 15). To date, no statistically significant impact of RB1 gene mutations on patient prognosis in colorectal cancer has been shown [31]. In addition to CTNNB1 and RB1, a PIK3CA mutation was found in one of the two neoplasms. Mutations in PIK3CA can be detected in various cancer types and have been associated with more aggressive metastatic behaviour in colorectal cancer [32]. However the PIK3CA (c.1173A>G, p.Ile391Met) mutation found here was a variant of uncertain significance (VUS) at the time of diagnosis but is now considered benign [33]. Through analyses of PIK3CA mutations in three colorectal carcinoma data sets we detected a significant co‐occurrence of PIK3CA and KRAS, which supports previous findings on that correlation [34].
The most important common feature of the two cases is the FBXW7 point mutation c.1393C>T(p.Arg465Cys). The FBXW7 gene codes for the substrate recognition component of a SCF (SKP1‐CUL1‐F‐box protein) E3 ubiquitin–protein ligase complex, which functions as an ubiquitin ligase marking several dominant oncogenic proteins, including c‐myc, cyclin E, notch and β‐catenin for ubiquitin mediated proteasomal degradation [35, 36]. Loss of function FBXW7 mutations, like the R465C gene variant described here, occur in approximately 11% of colorectal cancers [37]. Mono‐allelic missense alterations, which affect crucial arginine residues, have been reported to be the most common mutant genotypes, even though bi‐allelic inactivation mutations occur [38]. In 2017, Korphaisarn et al showed data suggesting a greater emphasis of FBXW7 missense mutation in comparison to other gene aberrations for patient outcome, linking these mutations, like those found in the above presented two cases, with a strong negative prognostic association [39]. Additional to its role as a key player in maintaining the balance between stem cell resting state and self‐regeneration [40], FBXW7 is a known regulator of Wnt/β‐catenin signalling in pancreatic cancer [41]. Although the latter has not yet been shown in colorectal cancer cells, the concept of FBXW7 controlling Wnt/β‐catenin signalling in colorectal cancer seems plausible, as a correlation between FBXW7 status and Wnt/β‐catenin signalling has been demonstrated in various cancer types [41, 42, 43]. Therefore, we suppose that the detected FBXW7 mutation resulted in malfunctioning of β‐catenin depletion with subsequent β‐catenin accumulation in the nucleus, leading to extreme overactivation of Wnt‐signalling. Due to this excessive activation of the Wnt/β‐catenin pathway, tumour cells in the colon may gain a pronounced plasticity, which may cause the critical switch towards this special combined morphology. Consistent with this hypothesis, de‐differentiation of colon cells by soluble Wnt‐ligand was recently shown by others [44]. Furthermore studies indicated the induction of squamous transdifferentiation through activation of β‐catenin signalling in various tissues [45]. Additionally, this hypothesis is supported by the findings of Davis et al, who showed reinforced Wnt‐signalling through FBXW7 propeller tip mutation and hence a driven tumorigenesis in mouse models [46]. Notably, the R465 gene variant found in our two cases also represents a propeller tip mutation.
Of note, Wnt activating mutations in FBXW7 and CTNNB1 are not restricted to the rare colorectal cancer type identified here, but also occur in classical adenocarcinoma. However, it is widely accepted that the intestinal epithelial cell subtype of cancer origin has a major influence on ultimate tumour characteristics. In neuroendocrine tumours, these cells are most likely represented by neural crest‐derived, precursor (entero)endocrine cells [47]. Different subtypes of these secretory precursor cells localise close to the crypt base, show mixed expression of secretory and bona‐fide intestinal stem cell markers, and possess a high degree of plasticity when confronted by regenerative signals, such as pathway Wnt activation [48, 49]. Importantly, a study by Wang et al revealed that aberrant Wnt activation at an early stage of neurogenin three‐dependent enteroendocrine cell differentiation induces small intestinal adenomas positive for serotonin expression in mice [50]. Given the low frequency of enteroendocrine cells (1–2%), and the short lifespan of their early precursors, this might explain the rare occurrence of neuroendocrine tumours, and the mixed neuroendocrine and squamous cell carcinomas described here, in colorectal cancer patients. Future studies on animal models should clarify if the propeller mutation in FBXW7 alone or in combination with alterations in RB1 or CTNNB1, when occurring in distinct (neuro)endocrine precursor cells of the adult colon, gives rise to the mixed cancer type characterised in our study.
In summary, these data seem to be a first important hint for the tumorigenesis of the mixed neuroendocrine and squamous carcinoma subtype. The underlying FBXW7 mutation might be the connecting element and the trigger for the crucial morphological switch, via overactivation of the canonical Wnt/β‐catenin signalling pathway. Its special relevance is also highlighted by the fact that it appears to reveal co‐occurrence with two mutations, specifically RB1 and PIK3CA, which were also detected in the presented cases. Other genes related to neuroendocrine differentiation, like ASCL1, may also play a role in the development of the neuroendocrine component, especially since ASCL1 is involved in the Notch‐Hes1 axis, which is analogous to the Wnt‐beta catenin signalling pathway, altered by the FBXW7 mutation [51, 52, 53]. Our findings may expedite the understanding of combined tumour development in the colon and in addition help establish awareness for such rare neoplasms, although continuing research, especially with regard to divergent differentiation of neuroendocrine‐ and squamous‐related genes, is necessary to fully decode the development of this combined neoplasm.
In the past, we and others provided evidence that MiNEN do have a monoclonal origin and are not stochastically neighbouring tumours [54, 55]. Furthermore, we found key mutations such as KRAS, TP53 and APC in both tumour components of MiNEN, which indicated a tumour progression similar to the well‐known classical adenoma–carcinoma sequence of colorectal adenocarcinomas [54]. We assume that the large cell neuroendocrine carcinoma, after originating from an adenoma or an adenocarcinoma, developed squamous structures via transdifferentiating processes and hence resulted in a combined large cell neuroendocrine carcinoma and squamous cell carcinoma, in which the original glandular component vanished or was no longer detectable. Interestingly, the initial colon biopsy of the first case showed parts of an ulcerated carcinoma in addition to colon mucosa with distinct serrated morphology, which supports this hypothesis. A different option in the development of the combined morphology, such as chemotherapy‐induced transdifferentiation, as reported in lung cancer, has to be considered as well [56]. However, in our cases chemotherapy took place after the microscopic characterisation of the resected specimen was completed and thus a chemotherapy‐induced switch resulting in the combined morphology seems unlikely.
In conclusion, a mixed large cell neuroendocrine carcinoma and squamous cell carcinoma of the colon can occur, even if it is extremely rare. Furthermore, we provide the histological and genetic evidence for a primary origin of this combined carcinoma in the colon and our data indicate that tumour development might occur via FBXW7 mutation‐triggered tumorigenesis, and very intensive Wnt‐signalling pathway enhancement. In combination with the absence of classical mutations of the adenoma–carcinoma sequence, as well as the notable morphology, this could be a first hint toward a distinct entity and novel subtype of colorectal carcinoma.
Author contributions statement
CW conceived and carried out experiments, drafted the article and contributed substantially to conception and design of the study and interpretation of data. TK and JN contributed substantially to conception of the study and interpretation of data and revised the article critically for important intellectual content. PJ, AJ, JK, SE, CJA and MV carried out experiments, analysed data and revised the article critically. All authors were involved in writing the paper and had final approval of the submitted and published versions.
Supporting information
Figure S1. Morphological characteristics from case 1 in close‐up view
Click here for additional data file.
Acknowledgement
We thank G Charell and J Kövi for excellent technical assistance. Open access funding enabled and organized by Projekt DEAL. | Fatal | ReactionOutcome | CC BY-NC | 33197299 | 18,711,884 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Adrenocorticotropic hormone deficiency'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Arthritis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Asthma'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Bile duct stone'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Cholangitis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Colitis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Diarrhoea'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Eosinophilic fasciitis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Hepatitis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Infectious pleural effusion'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Infusion related reaction'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Latent tuberculosis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Myocarditis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Nephritis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Pneumonia'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Pneumonitis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Pyrexia'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Rash'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Rhinitis allergic'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Thyroid disorder'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Uveitis'. | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | ATEZOLIZUMAB, NIVOLUMAB, PEMBROLIZUMAB | DrugsGivenReaction | CC BY | 33201587 | 18,564,141 | 2021-01 |
What was the outcome of reaction 'Pneumonitis'? | Outcome and risk factor of immune-related adverse events and pneumonitis in patients with advanced or postoperative recurrent non-small cell lung cancer treated with immune checkpoint inhibitors.
Non-small cell lung cancer (NSCLC) patients with pre-existing respiratory diseases have been excluded in clinical trials of immune checkpoint inhibitor (ICI) therapy, and it is unknown whether the same degree of response can be expected as that in patients without pre-existing respiratory diseases and if they are associated with increased risk for various immune-related adverse events (irAEs) and ICI pneumonitis. This study aimed to evaluate predictive factors of clinical response, prognostic factors, risk factors of irAEs, and ICI pneumonitis in NSCLC patients with or without pre-existing respiratory diseases.
We conducted a retrospective study of 180 NSCLC patients who received ICI monotherapy of nivolumab, pembrolizumab, or atezolizumab from 1 January 2016 to 31 March 2019.
A total of 119 patients had pre-existing respiratory diseases, including 20 with pre-existing idiopathic interstitial pneumonias (IIPs). A total of 85 patients experienced irAEs, of which ICI pneumonitis was the most frequent adverse event, occurring in 27 patients. Of the three patients who died from irAEs, all from ICI pneumonitis, two had pulmonary emphysema and one had pre-existing IIP. In multivariate analyses, irAEs were associated with objective response rate (ORR) and favorable OS, and IIPs were associated with increased risk for ICI pneumonitis. However, IIPs were not associated with low ORR or poor OS.
Pre-existing IIPs were a risk factor for ICI pneumonitis. However, this study showed that ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Significant findings of the study: Pre-existing IIPs were a risk factor for ICI pneumonitis, but objective response rate and prognosis of patients with IIPs were similar to those of other patients.
In patients with pre-existing IIPs, ICI pneumonitis should be noted. However, ICI therapy can be offered to patients with pre-existing respiratory diseases with the expectation of the same degree of response as that in patients without pre-existing respiratory diseases.
Introduction
Immune checkpoint inhibitors (ICIs), including programmed cell death‐1 (PD‐1) inhibitor and programmed cell death ligand‐1 (PD‐L1) inhibitor, have become a standard treatment for patients with unresectable advanced or recurrent non‐small cell lung cancer (NSCLC). Nivolumab and pembrolizumab are PD‐1 inhibitors, and atezolizumab is a PD‐L1 inhibitor. In phase III trials, nivolumab, pembrolizumab, and atezolizumab as second‐line treatment provided longer overall survival (OS) than docetaxel in NSCLC patients.
1
,
2
,
3
,
4
Additionally, pembrolizumab as a first‐line treatment provided longer OS than platinum‐based chemotherapy in NSCLC patients with a PD‐L1 tumor proportion score (TPS) ≥50% and those with PD‐L1 TPS ≥1%.
5
,
6
Recently, phase III trials showed that combination therapy of ICIs and platinum‐based chemotherapy as first‐line treatment in NSCLC patients has a higher objective response rate (ORR) and offers longer progression‐free survival (PFS) and OS than chemotherapy alone, regardless of the PD‐L1 TPS.
7
,
8
,
9
However, the clinical benefits remain limited to a subset of patients, and the predictive factors for response and prognosis in patients treated with ICIs are still unclear.
Additionally, ICIs can induce various immune‐related adverse events (irAEs). In phase III trials, irAEs developed in 20%–30% of patients.
3
,
5
In the clinical setting, irAEs developed more frequently than those in the phase III trials, with 30%–60% of patients affected.
10
,
11
,
12
Nevertheless, knowledge of the frequency, risk factors, and management of irAEs in the clinical setting is insufficient. In particular, ICI‐related pneumonitis (ICI pneumonitis) accounts for 35% of anti‐PD‐1 inhibitor‐ and anti‐PD‐L1 inhibitor‐related deaths.
13
Therefore, it is the most serious and life‐threatening irAE, as stated in the American Thoracic Society research statement published in 2019.
14
In this statement, because patients with pre‐existing respiratory diseases were excluded in clinical trials, it is unknown whether such patients are associated with an increased risk for ICI pneumonitis.
Therefore, we retrospectively reviewed the clinical data of NSCLC patients treated with ICI monotherapy and aimed to identify predictive factors for response, prognosis, irAEs, and ICI pneumonitis in the clinical setting of these patients with or without pre‐existing respiratory diseases and those with idiopathic interstitial pneumonias (IIPs).
Methods
Subjects
From 1 January 2016 to 31 March 2019, 180 patients with unresectable advanced or recurrent NSCLC were treated with ICI monotherapy including nivolumab, pembrolizumab, and atezolizumab at our institution. The diagnosis of lung cancer was based on pathology or cytology findings. The clinical stage was established according to the eighth edition of the TNM classification. Information concerning tumorous characteristics including epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) rearrangement, c‐ros oncogene 1 (ROS‐1) rearrangement, BRAF V600E mutation, and PD‐L1 TPS was collected. The PD‐L1 TPS was assessed by means of the PD‐L1 immunohistochemistry 22C3 pharmDx assay. ICIs were administered until disease progression, intolerable toxicity, or patient refusal occurred. Pre‐existing respiratory diseases were diagnosed according to clinical features and high‐resolution computed tomography of the chest.
Study design
We retrospectively investigated patients' background, ORR, OS, and development and management of irAEs, including ICI pneumonitis. We also investigated the predictive factors for ORR, OS, irAEs, and ICI pneumonitis. Clinical data were collected from medical records. Baseline clinical parameters were obtained within one month of the initial diagnosis. Pre‐existing respiratory diseases were divided into IIPs with or without pulmonary emphysema (PE), radiation‐induced pulmonary fibrosis with or without PE, PE without interstitial lung diseases (ILDs), and others. Radiographic patterns of IIPs were classified according to the international multidisciplinary classification of the IIPs and clinical practice guideline for the diagnosis of idiopathic pulmonary fibrosis.
15
,
16
Pulmonary emphysema was defined as focal areas or regions of low attenuation, usually without visible walls on chest CT.
17
ORR was assessed according to the Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1.
18
OS was measured from first administration of the ICIs to death. The data cutoff date was 31 August 2019. The irAEs were assessed using the National Cancer Institute Common Terminology Criteria for Adverse Events (CTCAE) version 4.0. Radiographic patterns of ICI pneumonitis were classified into nonspecific interstitial pneumonia (NSIP) pattern, cryptogenic organizing pneumonia (COP) pattern, acute interstitial pneumonia/acute respiratory distress syndrome (AIP/ARDS) pattern, and hypersensitivity pneumonitis (HP) pattern.
19
The NSIP pattern is ground‐glass opacities (GGOs) and reticular opacities predominantly in peripheral and lower lung distribution, traction bronchiectasis and lower lobe volume loss. The COP pattern is multifocal bilateral parenchymal consolidations, GGOs and reticular opacities with peripheral and lower lung distribution. The HP pattern is diffuse GGOs, centrilobular nodularities, and air trapping. The AIP/ARDS pattern is diffuse or multifocal GGOs or consolidations predominantly in dependent lung regions, lung volume loss and traction bronchiectasis.
This study was conducted in accordance with the Declaration of Helsinki and was approved by the institutional review board of Saitama Cardiovascular and Respiratory Center.
Statistical analysis
Categorical data are summarized by frequency and percent, and continuous data are reported as the median and range. The Kaplan‐Meier method was used to estimate OS. Univariate and multivariate analyses were performed using a logistic regression model to determine predictors for ORR and a Cox proportional‐hazards model to determine predictors for OS, irAEs, and ICI pneumonitis. All statistical analyses were performed with EZR version 1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 3.4.3).
20
Results
Patient characteristics
In total, 180 patients with advanced NSCLC underwent ICI monotherapy (Table 1). The median patient age was 68.5 (range, 40–83) years, 77.8% of the patients were male, 84.4% were smokers, 90.6% had an Eastern Cooperative Oncology Group performance status (ECOG PS) of 0 or 1, 33.9% had no pre‐existing respiratory diseases, 11.1% had IIPs, 11.7% had radiation‐induced pulmonary fibrosis, 41.1% had PE, 55.6% had adenocarcinoma, 78.9% were at stage IV, and 22.8% had brain metastasis. A total of 13 patients used immunosuppressants, and three patients had autoimmune diseases. A total of 21 patients had an EGFR mutation, none had ALK fusion, three patients had ROS1 fusion, and two patients had a BRAF mutation. The percentages of patients with PD‐L1 TPS <1%, 1%–49%, and ≥50% were 13.9%, 18.3%, and 32.8%, respectively. Among the patients, 11.1% had received molecular targeted therapy, 28.9% had received radiation therapy, and 18.3% were treated with ICIs as first‐line therapy. Of the 99 patients with PE, 74 did not have ILDs including IIPs or radiation‐induced pulmonary fibrosis. The median follow‐up period from initiation of ICIs was 299.5 (range: 9–1314) days, and the median number of treatment cycle of ICIs was four (range: 1–70). Patients treated with pembrolizumab had a higher frequency of PD‐L1 TPS ≥50% compared to those treated with nivolumab or atezolizumab. Most patients treated with atezolizumab had PD‐L1 TPS <1%. In addition, about half of the patients treated with pembrolizumab had received it as first‐line therapy.
Table 1 Characteristics of patients treated with immune checkpoint inhibitors (ICIs)
ICI All (n = 180) Nivolumab (n = 99) Pembrolizumab (n = 70) Atezolizumab (n = 11)
Age at ICI initiation 68.5 (40–83) 68.0 (40–83) 70.0 (44–83) 65.0 (49–80)
Sex, male 140 (77.8) 79 (79.8) 55 (78.6) 6 (54.5)
Smoker 152 (84.4) 84 (84.8) 59 (84.3) 9 (81.8)
ECOG PS 0 or 1 163 (90.6) 89 (89.9) 64 (91.4) 10 (90.9)
Pre‐existing respiratory disease
PE 99 (55.0) 57 (57.6) 38 (54.3) 4 (36.4)
RIPF 21 (11.7) 15 (15.2) 4 (5.7) 2 (18.2)
IIPs 20 (11.1) 12 (12.1) 8 (11.4) 0 (0.0)
UIP pattern 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
Probable UIP pattern 6 (3.3) 4 (4.0) 2 (2.9) 0 (0.0)
Indeterminate for UIP pattern 9 (5.0) 5 (5.1) 4 (5.7) 0 (0.0)
NSIP pattern 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
Asthma 8 (4.4) 3 (3.0) 5 (7.1) 0 (0.0)
Old tuberculosis 3 (1.7) 1 (1.0) 2 (2.9) 0 (0.0)
MAC infection 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Bronchiectasis 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Silicosis 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
Autoimmune disease
Chronic thyroiditis 2 (1.1) 0 (0.0) 1 (1.4) 1 (9.1)
PBC 1 (0.6) 1 (1.0) 0 (0.0) 0 (0.0)
Use of corticosteroid or immunosuppressant 13 (7.2) 9 (9.1) 4 (5.7) 0 (0.0)
Histological type
Adenocarcinoma 100 (55.6) 54 (54.5) 37 (52.9) 9 (81.8)
Squamous cell carcinoma 47 (26.1) 28 (28.3) 19 (27.1) 0 (0.0)
Pleomorphic carcinoma 4 (2.2) 1 (1.0) 3 (4.3) 0 (0.0)
Adenosquamous carcinoma 2 (1.1) 2 (2.0) 0 (0.0) 0 (0.0)
LCNEC 1 (0.6) 0 (0.0) 1 (1.4) 0 (0.0)
NOS 26 (14.4) 14 (14.1) 10 (14.3) 2 (18.2)
EGFR mutation
Exon 19 deletion 11 (6.1) 6 (6.1) 4 (5.7) 1 (9.1)
L858R 7 (3.9) 4 (4.0) 3 (4.3) 0 (0.0)
Minor mutation 3 (1.7) 3 (3.0) 0 (0.0) 0 (0.0)
− 130 (72.2) 64 (64.6) 56 (80.0) 10 (90.9)
NA 29 (16.1) 22 (22.2) 7 (10.0) 0 (0.0)
ALK rearrangement
+ 0 (0.0) 0 (0.0) 0 (0.0) 0 (0.0)
− 139 (77.2) 70 (70.7) 59 (84.3) 10 (90.9)
NA 41 (22.8) 29 (29.3) 11 (15.7) 1 (9.1)
ROS‐1 rearrangement
+ 3 (1.7) 0 (0.0) 3 (4.3) 0 (0.0)
− 79 (43.9) 32 (32.3) 38 (54.3) 9 (81.8)
NA 98 (54.4) 67 (67.7) 29 (41.4) 2 (18.2)
BRAF V600E mutation
+ 2 (1.1) 1 (1.0) 1 (1.4) 0 (0.0)
− 31 (17.2) 15 (15.2) 11 (15.7) 5 (45.5)
NA 147 (81.7) 83 (83.8) 58 (82.9) 6 (54.5)
PD‐L1 TPS
<1% 25 (13.9) 15 (15.2) 2 (2.9) 8 (72.7)
1–49% 43 (23.9) 17 (17.2) 13 (32.9) 3 (27.3)
≥50% 49 (27.2) 4 (4.0) 45 (64.3) 0 (0.0)
NA 63 (35.0) 63 (63.6) 0 (0.0) 0 (0.0)
Stage
III 38 (21.1) 21 (21.2) 15 (21.4) 2 (18.2)
IV 142 (78.9) 78 (78.8) 55 (78.6) 9 (81.8)
Brain metastasis 41 (22.8) 21 (21.2) 15 (21.4) 5 (45.5)
Prior treatment for brain metastasis 33 (18.3) 17 (17.2) 12 (17.1) 4 (36.4)
Prior molecular targeted therapy 20 (11.1) 12 (12.1) 7 (10.0) 1 (9.1)
EGFR‐TKI 18 (10.0) 11 (11.1) 6 (8.6) 1 (9.1)
Prior radiotherapy 52 (28.9) 33 (33.3) 13 (32.9) 6 (54.4)
Prior thoracic radiotherapy 33 (18.3) 22 (22.2) 7 (10.0) 4 (36.4)
Line of ICI therapy
First‐line 33 (18.3) 0 (0.0) 33 (47.1) 0 (0.0)
Second‐line 66 (36.7) 37 (37.4) 26 (37.1) 3 (27.3)
≥Third‐line 81 (45.0) 62 (62.6) 11 (15.7) 8 (72.7)
Number of ICI therapies 4 (1–70) 3 (1–70) 5.5 (1–33) 4 (1–11)
Follow‐up period (days) 299.5 (9–1314) 242 (9–1314) 362 (11–856) 233 (62–456)
Data are presented as n, median (range) or n (%).
ALK, anaplastic lymphoma kinase; ECOG PS, Eastern Cooperative Oncology Group performance status; EGFR, epidermal growth factor receptor; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; LCNEC, large‐cell neuroendocrine carcinoma; MAC, Mycobacterium avium complex; NA, not available; NOS, not otherwise specified; NSIP, nonspecific interstitial pneumonia; PBC, primary biliary cirrhosis; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; ROS‐1, c‐ros oncogene 1; TKI, tyrosine kinase inhibitor; TPS, tumor proportion score; UIP, usual interstitial pneumonia.
IrAEs profile
Of the 180 patients treated with ICIs, 121 (67.2%) developed adverse events, and the most common of these other than irAEs were drug‐related fever and bacterial pneumonia (Table 2). IrAEs were observed in 85 (47.2%) patients, including 27 (15.0%) with ICI pneumonitis, 24 (13.3%) with rash, 23 (12.8%) with thyroid dysfunction, 20 (11.1%) with diarrhea or colitis, 13 (7.2%) with hepatitis, five (2.8%) with nephritis, four (2.2%) with arthritis, and three (1.7%) with isolated adrenocorticotropic hormone deficiency. A total of 21 (11.7%) patients experienced irAEs of grade 3 or higher in which ICI pneumonitis was the most frequent adverse event. Systemic corticosteroids were administered to 36 (42.4%) patients. Among the 34 patients requiring discontinuation of ICIs, seven (20.6%) underwent retreatment with ICIs and two experienced recurrence of irAEs. Most patients who develop side effects develop them within one year, especially within 90 days (Fig 1). In patients treated with nivolumab, pembrolizumab, and atezolizumab, 45 (45.5%), 38 (54.3%), and two (18.2%) had irAEs, and 14 (14.1%), 12 (17.1%), and 1 (9.1%) had ICI pneumonitis, respectively.
Table 2 Adverse events including immune‐related adverse events (irAEs)
Events Any grade Grade ≥3 Corticosteroid treatment Retreatment with ICIs irAEs after retreatment
Any AEs including irAEs 121 (67.2) 24 (13.3)
Drug‐related fever 26 (14.4) 1 (0.6)
Pneumonia 12 (6.7) 10 (5.6)
Asthma 4 (2.2) 0 (0.0)
Allergic rhinitis 3 (1.7) 0 (0.0)
Infusion reaction 1 (0.6) 0 (0.0)
LTBI 1 (0.6) 0 (0.0)
Pyothorax 1 (0.6) 1 (0.6)
Choledocholithic cholangitis 1 (0.6) 1 (0.6)
Any irAEs 85 (47.2) 21 (11.7) 36 (42.4) 7 (20.6) 2 (28.6)
ICI pneumonitis 27 (15.0) 10 (5.6) 20 (74.1) 1 (5.6) 0 (0.0)
Rash 24 (13.3) 2 (1.1) 4 (16.7) 1 (50.0) 1 (100.0)
Thyroid dysfunction 23 (12.8) 0 (0.0) 0 (0.0) 1 (20.0) 0 (0.0)
Colitis or diarrhea 20 (11.1) 2 (1.1) 6 (30.0) 3 (60.0) 1 (33.3)
Hepatitis 13 (7.2) 3 (1.7) 2 (15.4) 0 (0.0) NA
Nephritis 5 (2.8) 0 (0.0) 1 (20.0) NA NA
Arthritis 4 (2.2) 0 (0.0) 1 (25.0) 1 (100.0) 0 (0.0)
Isolated ACTH deficiency 3 (1.7) 3 (1.7) 0 (0.0) NA NA
Myocarditis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Uveitis 1 (0.6) 0 (0.0) 0 (0.0) NA NA
Eosinophilic fasciitis 1 (0.6) 1 (0.6) 1 (100.0) 0 (0.0) NA
Data are presented as n, median (range) or n (%).
ACTH, adrenocorticotropic hormone; AEs, adverse events; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LTBI, latent tuberculosis infection; NA, not available.
Figure 1 Kaplan‐Meier curves showing irAE free survival and irAE free survival rate at 30 days, 60 days, 90 days, 120 days, 150 days, 180 days and 365 days. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAE, immune‐related adverse event; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Predictive factors of antitumor response to ICIs
Of the 180 patients treated with ICIs, complete response was achieved in four patients (2.2%) and partial response in 44 (24.4%). Stable disease was present in 51 (28.3%) patients, and progressive disease occurred in 81 (45.0%). The overall ORR was 26.7%. The ORR of patients treated with nivolumab, pembrolizumab, and atezolizumab were 19.2%, 40.0%, and 9.1%, respectively. The ORR of patients with no pre‐existing respiratory disease, IIPs, radiation‐induced pulmonary fibrosis, and PE were 19.7%, 35.0%, 19.0%, and 31.1%, respectively. Univariate analysis indicated that type of ICIs, PD‐L1, line of ICI therapy, eosinophil count, lymphocyte count, lactate dehydrogenase (LDH) level, neutrophil‐to‐lymphocyte ratio (NLR), eosinophil count after treatment with ICIs, and irAEs were factors associated with antitumor response to ICIs (Table S1). In a multivariate logistic regression model, only LDH level and irAEs were significantly associated with antitumor response to ICIs (Table 3).
Table 3 Multivariate analyses of objective response rate and prognostic factors of all‐cause mortality in patients treated with immune checkpoint inhibitors (ICIs)
Analyses of objective response rate
n
ORR (%) OR (95% CI)
P‐value
PD‐L1 TPS <1% 25 12.0 Reference
1–49% 43 16.3 1.270 (0.229–7. 300) 0.785
≥50% 49 51.0 5.140 (0.836–31.600) 0.077
NA 63 20.6 2.200 (0.403–12.000) 0.363
ICIs Nivolumab 99 19.2 Reference
Atezolizumab 11 9.1 0.917 (0.074–11.300) 0.946
Pembrolizumab 70 40.0 1.850 (0.495–6.950) 0.360
Line of ICI therapy First‐line 33 48.5 0.876 (0.205–3.74) 0.858
Second‐line 66 19.7 Reference
≥Third‐line 81 23.5 1.960 (0.725–5.320) 0.184
Eosinophils (/μL) <500 158 22.8 Reference
≥500 22 54.5 2.190 (0.618–7.750) 0.225
Lymphocytes (/μL) <1500 103 20.4 Reference
≥1500 77 35.1 1.310 (0.545–3.150) 0.547
LDH (U/L) ≥230 68 16.2 Reference
<230 112 33.0 3.270 (1.340–8.020) 0.009
NLR ≥5 51 15.7 Reference
<5 129 31.0 2.940 (0.969–8.910) 0.057
Eosinophils after starting ICIs (/μL) <500 123 18.7 Reference
≥500 57 43.9 1.990 (0800–4.960) 0.139
irAEs None 95 15.8 Reference
Present 85 38.8 2.460 (1.070–5.650) 0.034
Analyses of prognostic factors
n
OS(days) HR (95% CI)
P‐value
ECOG PS 0–1 163 468 Reference
2–3 17 123 3.499 (1.756–6.969) < 0.001
PD‐L1 TPS ≥50% 49 NR Reference
1–49% 43 444 1.778 (0.713–4.435) 0.217
<1% 25 272 1.980 (0.685–5.720) 0.207
NA 63 315 1.183 (0.430–3.253) 0.745
Stage III 38 NR Reference
IV 142 367 1.867 (1.025–3.400) 0.041
ICIs Pembrolizumab 70 NR Reference
Nivolumab 99 296 2.493 (1.123–5.536) 0.025
Atezolizumab 11 307 2.803 (0.938–8.371) 0.065
Line of ICI therapy First‐line 33 NR Reference
Second‐line 66 289 1.134 (0.414–3.105) 0.807
≥Third‐line 81 385 0.692 (0.243–1.968) 0.490
WBC (/μL) <9000 146 467 Reference
≥9000 34 359 1.876 (0.985–3.570) 0.056
Monocytes (/μL) <600 116 592 Reference
≥600 64 296 1.170 (0.680–2.014) 0.570
Lymphocytes (/μL) ≥1500 77 592 Reference
<1500 103 296 1.313 (0.748–2.303) 0.343
LDH (U/L) <230 112 604 Reference
≥230 68 315 1.370 (0.888–2.112) 0.154
NLR <5 129 493 Reference
≥5 51 281 0.848 (0.446–1.614) 0.615
LMR ≥3 83 744 Reference
<3 97 281 1.782 (0.985–3.222) 0.056
PLR <300 139 472 Reference
≥300 41 226 1.711 (0.966–3.030) 0.066
Eosinophils after starting ICIs (/μL) ≥500 57 744 Reference
<500 123 322 1.191 (0.711–1.997) 0.507
irAEs Present 85 670 Reference
None 95 303 1.637 (1.041–2.573) 0.033
CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; ICIs, immune checkpoint inhibitors; irAEs, immune‐related adverse events; LDH, lactate dehydrogenase; LMR, lymphocyte‐to‐monocyte ratio; NA, not available; NLR, neutrophil‐to‐lymphocyte ratio; OR, odds ratio; ORR, objective response rate; PD‐L1, programmed cell death ligand‐1; PLR, platelet‐to‐lymphocyte ratio; TPS, tumor proportion score; WBC, white blood cell.
Prognostic factors of all‐cause mortality in patients treated with ICIs
The median OS was 444 days (95% confidence interval [CI]: 315–561) in all patients treated with ICIs (Fig 2). Univariate analysis indicated that ECOG PS, stage, type of ICI, PD‐L1, line of ICI therapy, white blood cell (WBC) count, monocyte count, lymphocyte count, LDH level, NLR, lymphocyte‐to‐monocyte ratio, platelet‐to‐lymphocyte ratio (PLR), eosinophil count after treatment with ICIs, and irAEs were prognostic factors (Table S2). In a multivariate Cox proportional hazard model, ECOG PS, type of ICI, stage IV, and irAEs were independent prognostic factors of all‐cause mortality (Table 3). Kaplan‐Meier curves for OS stratified by pre‐existing respiratory diseases, including IIPs, revealed no significant differences in patient prognosis between the various diseases (Fig 2a). Patients with IIPs of NSIP pattern tended to have a longer OS and patients with IIPs of UIP pattern tended to have a shorter OS (Fig 2b). However, the number of patients in each group was very small and there was no significant difference in prognosis. Other respiratory diseases included bronchial asthma in three and stable pulmonary tuberculosis in one. There were only four cases, two with PD‐L1 ≥50% and one with unknown PD‐L1, which may be due to the longest survival in this study. On the other hand, stratified by type of ICI revealed that patients treated with pembrolizumab had significantly longer median OS than those treated with nivolumab or atezolizumab (Fig 2c).
Figure 2 Kaplan‐Meier curves showing (a) surOS stratified by pre‐existing respiratory diseases; (b) OS stratified by radiographic pattern of IIPs; and (c) OS stratified by type of ICI in non‐small cell lung cancer patients treated with immune checkpoint inhibitors. The log‐rank test of the difference between survival curves of patients with and without pre‐existing respiratory disease was not significant. On the other hand, the log‐rank test revealed a significant survival benefit in patients treated with pembrolizumab compared to those treated with nivolumab or atezolizumab. CI, confidence interval; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; NR, not reached; NSIP, nonspecific interstitial pneumonia; PE, pulmonary emphysema; UIP, usual interstitial pneumonia.
Risk factors for irAEs
Univariate analysis indicated that age, WBC count, and lymphocyte count were risk factors for irAEs (Table S3). In a multivariate Cox proportional hazard model, only age and lymphocyte count were risk factors for irAEs (Table 4).
Table 4 Univariate and multivariate analyses of immune‐related adverse events (irAEs) and pneumonitis
Analyses of irAEs
n
irAEs (%) HR (95% CI)
P‐value
Age ≥75 42 31.0 Reference
<75 138 52.2 2.109 (1.167–3.813) 0.013
WBC (/μL) <9000 146 43.8 Reference
≥9000 34 61.8 1.649 (0.991–2.743) 0.054
Lymphocytes (/μL) <1500 103 37.9 Reference
≥1500 77 59.7 1.553 (1.001–2.409) 0.049
Analyses of pneumonitis
n
Pneumonitis (%) HR (95% CI)
P‐value
Pre‐existing respiratory disease None 61 6.6 Reference
IIPs 20 35.0 4.350 (1.225–15.440) 0.023
RIPF 21 19.0 3.096 (0.735–13.040) 0.124
PE without ILD 74 16.2 2.088 (0.645–6.760) 0.219
Others 4 0.0 <0.001 (0.000–Inf) 0.998
PD‐L1 TPS <1% 49 24.0 3.897 (0.911–16.670) 0.067
1–49% 43 3.0 Reference
≥50% 25 23.7 2.488 (0.660–9.380) 0.178
NA 63 9.5 1.480 (0.352–6.222) 0.593
WBC (/μL) <9000 146 12.3 Reference
≥9000 34 26.5 1.263 (0.492–3.243) 0.627
Eosinophils (/μL) <500 158 12.7 Reference
≥500 22 31.8 1.853 (0.705–4.873) 0.211
Monocytes (/μL) <600 116 8.6 Reference
≥600 64 26.6 2.080 (0.875–4.941) 0.097
Albumin (g/dL) ≥4 50 6.0 Reference
<4 126 19.0 2.090 (0.588–7.420) 0.254
NA 4 0.0 <0.001 (0.000–Inf) 0.998
CRP (mg/dL) <1 96 7.3 Reference
≥1 84 23.8 1.711 (0.645–4.537) 0.281
CI, confidence interval; CRP, C‐reactive protein; HR, hazard ratio; ICIs, immune checkpoint inhibitors; IIPs, idiopathic interstitial pneumonias; ILD, interstitial lung disease; irAEs, immune‐related adverse events; NA. not available; PD‐L1, programmed cell death ligand‐1; PE, pulmonary emphysema; RIPF, radiation‐induced pulmonary fibrosis; TPS, tumor proportion score; WBC, white blood cell.
Risk factors for ICI pneumonitis
Univariate analysis indicated that age, IIPs, PD‐L1, WBC count, eosinophil count, monocyte count, and albumin and C‐reactive protein (CRP) levels were risk factors for ICI pneumonitis (Table S4). In a multivariate Cox proportional hazard model, however, IIPs were the only risk factor for ICI pneumonitis (Table 4).
Characteristics of ICI pneumonitis
Of the 27 patients with ICI pneumonitis, the most common radiographic pattern was the COP pattern (16 patients; Fig 3a) followed by NSIP pattern (four patients; Fig 3b), HP pattern (three patients; Fig 3c), and AIP/ARDS pattern (three patients; Fig 3d). Time to onset of ICI pneumonitis with AIP/ARDS pattern ranged from five to 17 days and tended to be shorter than that of ICI pneumonitis with other radiographic patterns (Fig 4). Among the three patients who developed ICI pneumonitis with AIP/ARDS pattern, all three had respiratory diseases other than lung cancer (two with pulmonary emphysema and one with IIP), all three were at grade 3 severity at the onset of ICI pneumonitis, and all three died. All of the patients with ICI pneumonitis of grade 2 or higher were treated with corticosteroids, whereas all of the patients with ICI pneumonitis of grade 1 were observed without treatment.
Figure 3 Radiographic pattern of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis. (a) COP pattern; (b) NSIP pattern; (c) HP pattern; and (d) AIP/ARDS pattern. COP, cryptogenic organizing pneumonia; NSIP, nonspecific interstitial pneumonia; HP, hypersensitivity pneumonitis; AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome.
Figure 4 Radiographic pattern, grade, treatment, and outcome of immune checkpoint inhibitor (ICI)‐related pneumonitis (ICI pneumonitis). Data are presented as number of patients or range of time in days to onset of ICI pneumonitis. AIP/ARDS, acute interstitial pneumonia/acute respiratory distress syndrome; COP, cryptogenic organizing pneumonia; HP, hypersensitivity pneumonitis; mPSL, methylprednisolone; NSIP, nonspecific interstitial pneumonia; PSL, prednisolone.
Discussion
In this study, we revealed predictive factors for clinical outcome and irAEs in patients with advanced NSCLC treated with ICI monotherapy in a clinical setting. Predictive factors for clinical response were LDH level, and irAEs. Predictive factors for prognosis were ECOG PS, stage, type of ICI, and irAEs. Pembrolizumab had the highest frequency of irAEs and the best tumor response and prognosis. About half of the patients experienced irAEs, the risk factors for which were age and lymphocyte count. The most frequent irAE was ICI pneumonitis, and all three deaths were due to ICI pneumonitis with an AIP/ARDS radiographic pattern. Although IIPs were a significant risk factor for ICI pneumonitis, there were no significant differences in the ORR and OS between patients with IIPs and those without respiratory diseases.
Previously, it was reported that several factors predict the response and prognosis in patients treated with ICIs. In phase III trials, PD‐L1 expression was associated with OS in NSCLC patients treated with ICIs.
2
,
3
Tamiya et al. showed that ECOG PS ≥2, liver metastasis, and lung metastasis were predictive of poor PFS in NSCLC patients treated with nivolumab.
21
Additionally, several studies reported that irAEs were associated with clinical response and prognosis. Sato et al.
10
and Toi et al.
22
respectively investigated 38 and 70 NSCLC patients treated with nivolumab and reported that patients with irAEs had significantly higher ORR than those without irAEs (63.6 vs. 7.4% and 57 vs. 12%, respectively). Haratani et al.
23
investigated 134 NSCLC patients treated with nivolumab and reported that the patients with irAEs had significantly longer median OS than those without irAEs (not reached vs. 11.1 months). Similarly, Ricciuti et al.
24
studied 195 NSCLC patients treated with nivolumab and reported that the patients with irAEs experienced significantly longer median OS than those without irAEs (17.8 vs. 4.0 months), and patients who developed ≥2 irAEs had significantly longer median OS than those with one or no irAEs (26.8 vs. 11.9 vs. 4.0 months). The present study also revealed that irAEs were associated with both ORR and OS in NSCLC patients treated with ICIs. In contrast, Ksienski et al.
25
studied 271 patients treated with nivolumab or pembrolizumab and showed that treatment interruption due to irAEs was associated with a lower median OS than was continuous treatment (8.27 vs. 14.54 months). Therefore, appropriate assessment and management of irAEs is necessary.
Several studies have shown risk factors of irAEs. Diehl et al.
11
reported that baseline lymphocyte and eosinophil counts were associated with irAEs in solid tumor patients treated with ICIs. A pooled analysis including NSCLC patients from four trials of ICIs showed that patients aged ≥75 years had a lower incidence of grade 3 or 4 adverse events than patients aged <65 years (23 vs. 47%).
26
However, because a pooled analysis including NSCLC patients from three trials for pembrolizumab showed that there were no differences in the incidence of irAEs between patients aged <75 and ≥75 years (24.8 vs. 25.0%),
27
it remains controversial whether age is related to the incidence of irAEs.
In the present study, most of the patients who developed ICI pneumonitis or liver injury after ICI therapy discontinued ICIs permanently. According to the American Society of Clinical Oncology clinical practice guideline, if patients develop irAEs, ICI therapy is continued with close monitoring for grade 1 irAEs, is held for grade 2 or 3 irAEs until they improve to grade 1 or less, and is permanently discontinued for grade 4 irAEs except endocrinopathies.
28
Patients with grade 3 or 4 ICI pneumonitis and liver injury were required to permanently discontinue ICI therapy. Mouri et al.
29
reported the clinical differences between patients who discontinued ICIs and those who retreated after occurrences of irAEs. They found that patients who discontinued ICIs tended to more frequently have ICI pneumonitis, thyroid dysfunction, and liver injury than those retreated from therapy.
Although several clinical trials revealed that 2.5% to 5% of patients developed ICI pneumonitis,
14
its incidence was higher in the clinical setting than in the clinical trials, and 5.4% to 16.9% of patients experienced ICI pneumonitis.
10
,
11
,
30
Tone et al.
31
reported that patients with ICI pneumonitis of grade 3 or higher were associated with shorter median OS than those with ICI pneumonitis of grade 2 or lower or no ICI pneumonitis. A retrospective study reported that radiographic patterns were associated with grades of ICI pneumonitis, with the AIP/ARDS pattern associated with the highest grade, followed by the COP pattern, and the NSIP and HP patterns associated with lower grades.
32
Several studies have reported risk factors of ICI pneumonitis. Cui et al.
33
revealed that prior radiotherapy and combination therapy, defined as treatment with anti‐PD‐1 antibody and chemotherapy, targeted therapy, or anticytotoxic T‐lymphocyte‐associated antigen‐4 antibody, were significantly associated with ICI pneumonitis in a multivariable logistic regression model. Oshima et al.
34
analyzed the Food and Drug Administration Adverse Event Reporting System database and investigated the association between pneumonitis and the combination of nivolumab and EGFR‐tyrosine kinase inhibitor (TKI). They reported that 18 of the 70 patients who were treated with the combination developed pneumonitis (25.7%), with the order of treatment in 15 patients identified as EGFR‐TKI after nivolumab administration. A systematic review and meta‐analysis showed that the incidence of ICI pneumonitis in NSCLC was higher than that in melanoma.
35
Additionally, a retrospective study showed the incidence in NSCLC of the adenocarcinoma histological pattern to be lower than that in NSCLC of the squamous histological pattern.
36
Several studies showed the efficacy and safety of ICIs in patients with pre‐existing ILD or interstitial lung abnormalities, which are defined as areas of increased lung density on lung computed tomography in individuals with no known ILD.
30
Kanai et al.
37
investigated 216 NSCLC patients who had received nivolumab and reported that the incidence of ICI pneumonitis was significantly higher in patients with pre‐existing ILD than in patients without ILD (31 vs. 12%). There were no significant differences in the ORR (27 vs.13%) and median PFS (2.7 vs. 2.9 months). Nakanishi et al.
30
studied 83 NSCLC patients who had received nivolumab or pembrolizumab and found that the patients with ICI pneumonitis had a significantly higher frequency of interstitial lung abnormalities than those without ICI pneumonitis (42.9 vs. 10.1%).There were no significant differences in the response to the ICIs. Fujimoto et al.
38
studied the efficacy and safety of nivolumab for NSCLC patients with mild IIPs. They reported that two of the 18 patients (11.1%) with IIPs developed ICI pneumonitis. The ORR was 39%, median PFS was 7.4 months, and median OS was 15.6 months. Similar to the previous studies, the incidence of ICI pneumonitis in the present study was significantly higher in patients with pre‐existing IIPs than in those without pre‐existing respiratory diseases (35.0 vs. 6.6%), and the ORR in the patients with IIPs was 35.0%. In addition, patients with IIPs tended to have a longer OS, although the difference was not significant. In this study, patients treated with atezolizumab had the poorest ORR and OS, and none of the patients with IIP received atezolizumab. Furthermore, although IIPs was a risk factor for the development of ICI pneumonitis in this study, two‐thirds of ICI‐pneumonitis patients were Grade 1–2, with a fatality rate of only 10%, and patients with irAEs had better OS than those without irAEs. These findings may have contributed to the present study.
This study has several limitations. First, because it was retrospective, some patient characteristics were not available. Second, it was performed at a single hospital, and only Japanese patients were treated. Third, the sample size was small. Finally, diagnoses of ICI pneumonitis were largely based on clinical course and CT findings. Only a small percentage of patients underwent bronchoalveolar lavage to exclude pneumonia. However, pneumonitis was not resolved by antimicrobial drugs.
In summary, the incidence of irAEs might be a useful predictor of clinical response and prognosis in NSCLC patients treated with ICIs, and we believe that appropriate management of irAEs can lead to clinical benefit. Because all three patient deaths were due to ICI pneumonitis, we consider ICI pneumonitis to be the most important irAE, and radiological pattern classification was useful for predicting the prognosis of ICI pneumonitis. Pre‐existing IIPs were a risk factor for ICI pneumonitis; however, this study showed that ICI therapy can be offered to patients with pre‐existing respiratory diseases with the expectation of the same degree of response as that in patients without pre‐existing respiratory diseases.
Disclosure
The authors declare there are no conflicts of interest.
Supporting information
Table S1 Univariate and multivariate analyses of objective response rate.
Table S2 Univariate and multivariate analyses of prognostic factors of all‐cause mortality in patients treated with ICIs.
Table S3 Univariate and multivariate analyses of irAEs.
Table S4 Univariate and multivariate analyses of ICI pneumonitis.
Click here for additional data file. | Fatal | ReactionOutcome | CC BY | 33201587 | 18,564,141 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Basal cell carcinoma'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Cytomegalovirus infection reactivation'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Drug ineffective'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Eczema'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Gastroenteritis'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Genitourinary tract infection'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Injection site reaction'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Oral papilloma'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Oropharyngeal candidiasis'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Psoriasis'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Rash'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Respiratory tract infection'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Therapeutic response decreased'. | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | GOLIMUMAB, MESALAMINE, METHOTREXATE | DrugsGivenReaction | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
What was the administration route of drug 'GOLIMUMAB'? | Two-year effectiveness and safety of golimumab in ulcerative colitis: An IG-IBD study.
Few data exist regarding the long-term effectiveness of golimumab in ulcerative colitis. No data have been reported on real-world continuous clinical response.
This study aimed to describe the long-term outcomes in a large cohort of patients on golimumab who had ulcerative colitis.
Consecutive patients with active ulcerative colitis, started on golimumab, were enrolled and prospectively followed up. The primary end point was to evaluate the long-term persistence on golimumab therapy.
A total of 173 patients with ulcerative colitis were studied. Of these, 79.2% were steroid dependent, and 46.3% were naïve to anti-tumour necrosis factor alpha agents. The median duration of golimumab therapy was 52 weeks (range: 4-142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively. Biological-naïve status (odds ratio [OR] = 3.02, 95% confidence interval [CI]: 1.44-6.29; p = 0.003) and being able to discontinue steroids at Week 8 (OR = 3.32, 95% CI: 1.34-8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.08-8.02; p = 0.036) were associated with longer persistence on therapy. At Week 54, 65/124 (52.4%) postinduction responders were in continuous clinical response. A continuous clinical response was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (p < 0.01). Overall, 40 (23.1%) patients were in clinical remission at the last follow-up visit. Twenty-six adverse events were recorded, leading to golimumab withdrawal in 9.2% of patients.
Biological-naïve status and not requiring steroids at Weeks 8 and 14 seem to be associated with a longer persistence on golimumab therapy in ulcerative colitis.
1 INTRODUCTION
Ulcerative colitis (UC) is a chronic inflammatory disease involving the colon, characterised by a relapsing/remitting course and requiring lifelong medical therapies. Biological drugs and, more recently, Janus kinase inhibitors such as tofacitinib are the best medical option for patients with moderate‐to‐severe disease with an inadequate response or intolerance to conventional therapies (5‐amynosalicilates, steroids and/or thiopurines). 1 golimumab, a fully human IgG1 kappa monoclonal antibody, subcutaneously administered, has now been used in clinical practice for more than five years for the treatment of adult subjects with UC. 2 , 3 The efficacy of golimumab for the induction and maintenance of clinical remission in biological‐naïve UC patients has been studied in two completed clinical trials: PURSUIT induction and PURSUIT maintenance. 4 , 5 In the second trial, a continuous clinical response (CCR) through Week 54, that is, maintenance of a clinical response through Week 54 among golimumab‐induction responders, was adopted as the primary end point, and this achieved in 47.0% of patients receiving 50 mg golimumab and in 49.7% of receiving 100 mg golimumab compared to 31.2% receiving placebo. 5 Long‐term open‐label follow‐up confirmed a good profile of effectiveness up to 4 years, more evident among patients with CCR at 54 weeks. 6 , 7 To date, few long‐term real‐life data have been reported, showing highly variable persistence on golimumab therapy in some cohorts, and particularly reduced in patients pluri‐exposed to anti‐tumour necrosis factor alpha (TNF‐α) drugs and treated with the fixed dose of 50 mg during maintenance therapy. 8 , 9 , 10 , 11 , 12
The aims of this study were to investigate the mid‐ and long‐term outcomes of patients with UC treated with golimumab in real life and to explore potential predictors for these outcomes.
2 METHODS
We performed an observational retrospective/prospective study in which consecutive patients who started golimumab therapy between May 2014 and December 2015 at 29 Italian centres, affiliated with the Italian Group for the study of Inflammatory Bowel disease (IG‐IBD), were enrolled. All patients had a prospectively designed standardised follow‐up until December 2017.
In Italy, to guarantee the prescribing appropriateness, the Italian Medicine Agency (Agenzia Italiana del Farmaco [AIFA]) has instituted a computerised database system for several drugs, including golimumab, accessible to physicians and mandatory to finalise the prescription both at the beginning of and during maintenance treatment. Therefore, accessibility criteria and follow‐up visits scheduled every 8 weeks, requiring a clinical assessment through partial Mayo score (PMS), 13 are standardised for all patients on treatment with golimumab. Accordingly, we adopted a prospectively planned follow‐up protocol, with a shared common database mirroring the AIFA registry, to enrol patients and to follow them up until December 2017.
According to the current European‐approved golimumab label, 2 all patients received golimumab induction with 200 and 100 mg at Weeks 0 and 2, respectively, followed by 50 or 100 mg every 4 weeks, depending on their weight (>80 or <80 kg). Patients were not allowed to increase the dose in case of partial response after the induction or loss of response.
The collected baseline data included: sex, age, weight, height, body mass index, duration of UC, extension of UC according to the Montreal classification, 14 clinical and endoscopic activity, previous therapies (both conventional and biological), the date of the first golimumab dose and concomitant therapies. Baseline and follow‐up clinical and endoscopic activities were determined according to PMS and endoscopic subscore, respectively. 13 Concomitant medications, new prescriptions during follow‐up, the tapering of steroids and timing of treatment discontinuation were left to the investigators' evaluation.
The primary end point of our study was to evaluate the long‐term persistence on golimumab therapy due to sustained clinical benefit.
Secondary analyses looked for (a) proportion of patients achieving clinical remission at Week 54; (b) CCR through Week 54 among patients with a clinical response after induction; (c) rate of surgery for medical refractory UC; (d) effectiveness of treatment in sparing steroids among patients taking steroids at baseline; and (e) proportion of patients achieving endoscopic remission.
A clinical response was defined as a reduction in the PMS of at least two points and a decrease of at least 30% from the baseline score, with a decrease of at least one point on the rectal bleeding subscale or an absolute rectal bleeding score of 1 or 0. Clinical remission was defined as a PMS of two or lower and no subscore higher than one. We adopted the same definition of CCR through Week 54 previously reported, even though the interval between each clinical assessment was set every 8 weeks. 5 Endoscopic examinations were mandatory at Week 54, but could be anticipated according to clinical judgement. Endoscopic remission was defined as an endoscopic Mayo subscore of 0 or 1. For patients undergoing two or more endoscopic assessments during the study, the last one was considered for the evaluation of endoscopic remission.
Reasons for golimumab discontinuation were categorised as: primary failure, defined as the absence of a clinical response at Week 8; secondary failure, defined as a relapse of clinical symptoms during maintenance treatment requiring physicians' interventions; and others, including intolerance or adverse events, lost to follow‐up and pregnancy.
All adverse events that occurred from the beginning of golimumab treatment to the date of withdrawal or last follow‐up visit on therapy were recorded and categorised as adverse events of interest (AEI) if requiring medical intervention/hospitalisation and/or treatment discontinuation (temporary or permanent).
2.1 Statistical analysis
Data were described using means with standard deviation and medians with range for continuous data and percentages for discrete data. Categorical variables were compared using the χ2 test (or Fisher exact test). Cumulative probabilities of persistence on golimumab therapy and CCR through Week 54 were estimated by the Kaplan–Meier method. Binary logistics regression was used to estimate the association between each predictor and persistence on golimumab therapy. Variables that tested significant at binary regression (p < 0.2) were then included in a multivariate logistic regression analysis. Steroid use was updated at each available time point. Results are shown as odds ratios (ORs) and 95% confidence intervals (CIs). A p < 0.05 indicated statistical significance. All analyses were performed with IBM SPSS Statistics for Windows v24.0 (IBM Corp).
2.2 Ethics approval
The protocol was approved by the ethics committee of the coordinator centre (Fondazione Policlinico Universitario A. Gemelli IRCCS‐Universita Cattolica del Sacro Cuore, Roma, Italy, protocol 1462, 26 January 2017) and of all participating centres. The study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Written informed consent was obtained from each patient included in the study.
3 RESULTS
3.1 Patient population
A total of 173 patients were included. Baseline patients' characteristics are summarised in Table 1. According to AIFA eligibility criteria, all patients had moderate to severe active disease, and all of them had showed an inadequate response or had a contraindication to steroids. In particular, 137 (79.2%) patients were steroid dependent, and 27 (15.6%) were refractory according to IG‐IBD definitions. 15 The remaining nine patients had contraindications to steroid therapy. At baseline, 60 (34.7%) patients were on concomitant steroid therapy; 52 (30.1%) and 36 (20.8%) were taking oral steroids at Weeks 8 and 14, respectively. A total of 131 (75.7%) patients weighed less than 80 kg and thus received 50 mg every 4 weeks as a maintenance dose; the remaining 42 (24.2%) weighed more than 80 kg and thus received 100 mg every 4 weeks. A total of 111 (64.2%) patients had been previously exposed to thiopurines, and 92 (53.2%) patients had been previously exposed to at least one anti‐TNF agent: 52 (30.1%) to infliximab, six (3.5%) to adalimumab and 34 (19.7%) to both.
TABLE 1 Baseline patient characteristics
Characteristic Value (N = 173)
Male, n (%) 94 (54.3)
Age (years), median (range) 45.7 (18.0–71.1)
Weight (kg), M ± SD 68.6 ± 14.8
>80 kg, n (%) 40 (23)
BMI (kg/m2), M ± SD 23.5 ± 3.88
Duration of disease (years), median (range) 6.50 (0–58.8)
Disease extent, n (%)
E1 6 (3.5)
E2 62 (35.8)
E3 105 (60.7)
Clinical severity at baseline PMS, n (%)
Moderate 89 (51.4)
Severe 84 (48.6)
Endoscopic score at baseline, n (%)
Mayo 2 75 (43.4)
Mayo 3 98 (56.6)
Previous exposure to anti‐TNF‐α, n (%) 92 (53.2)
Infliximab 52 (30.1)
Adalimumab 6 (3.5)
Both 34 (19.7)
Previous therapies, n (%)
Steroids 164 (94.7)
Thiopurine 111 (64.2)
Cyclosporine 3 (1.7)
Methotrexate 9 (5.2)
Steroid dependence, n (%) 137 (79.2)
Steroid refractoriness, n (%) 27 (15.6)
Concomitant therapies, n (%)
Steroids 60 (34.7)
Thiopurine 17 (9.8)
5‐ASA 107 (61.8)
Methotrexate 3 (1.7)
Abbreviations: 5‐ASA, 5‐aminosalicylic acid; BMI, body mass index; PMS, partial Mayo score (5–7 = moderate, >7 = severe); SD, standard deviation; TNF‐α, tumour necrosis factor alpha.
3.2 Persistency on golimumab therapy
The median time on golimumab treatment was 52 weeks (range: 4–142 weeks). The cumulative probability of maintaining golimumab treatment was 47.3% and 22.5% at 54 and 108 weeks, respectively (Figure 1). Overall, 126 (72.8%) patients withdrew from golimumab therapy after a median of 31.5 weeks (range: 4–126 weeks). Reasons for discontinuation were primary failure in 51 (40.5%) patients, secondary failure in 51 (40.5%) patients and other causes in 24 (19.1%) patients. Among the 102 patients who withdrew from treatment due to failure, 65 (63.7%) were anti‐TNF‐α experienced compared to 37 (36.3%) who were naïve (p = 0.007; Figure 2). Multivariate regression analysis showed that patients who were anti‐TNF‐α experienced were more likely to withdraw from golimumab therapy compared to patients who were anti‐TNF‐α naive (OR = 3.02, 95% CI: 1.44–6.29; p = 0.003). Moreover, not requiring steroids at Week 8 (OR = 3.32, 95% CI: 1.34–8.30; p = 0.010) and Week 14 (OR = 2.94, 95% CI: 1.088.02; p = 0.036) was associated with higher golimumab persistence. Conversely, male sex seemed to be protective from golimumab withdrawal (OR = 0.44, 95% CI: 0.21–0.94; p = 0.035; Table 2).
FIGURE 1 Cumulative probability of maintaining golimumab treatment
FIGURE 2 Cumulative probability of maintaining golimumab treatment. Patients split between those who were anti‐tumour necrosis factor (TNF) alpha naïve and those who were anti‐TNF alpha experienced
TABLE 2 Results of binary logistic regression for persistence on golimumab therapy in 173 UC patients
Variable Univariate, OR (CI), p Multivariate, OR (CI), p
Sex (male vs. female) OR = 0.52 (CI: 0.26–1.04), p = 0.061 OR = 0.44 (CI: 0.21–0.94), p = 0.035
Age (<45 vs. >45 years) OR = 0.62 (CI: 0.32–1.22), p = 0.166 OR = 1.32 (CI: 0.64–2.75), p = 0.453
Weight (<80 vs. ˃80 kg) OR = 1.08 (CI: 0.49–2.40), p = 0.846 –
Clinical activity at baseline (moderate vs. severe) OR = 0.88 (CI: 0.44–1.73), p = 0.701 –
Endoscopic activity at baseline (Mayo 2 vs. Mayo 3) OR = 0.53 (CI: 0.29–1.06), p = 0.072 OR = 1.63 (CI: 0.79–3.35), p = 0.188
Previous anti‐TNF‐α (exposed vs. naïve) OR = 2.60 (CI: 1.30–5.19), p = 0.006 OR = 3.02 (CI: 1.45–6.30), p = 0.003
BMI (<25 vs. >25) OR = 1.02 (CI: 0.47–2.19), p = 0.970 –
Disease extension (E1–E2 vs. E3) OR = 1.45 (CI: 0.72–2.89), p = 0.295 –
Steroids at Week 8 (yes vs. no) OR = 2.45 (CI: 1.22–8.73), p = 0.006 OR = 3.33 (CI: 1.34–8.29), p = 0.010
Steroids at Week 14 (yes vs. no) OR = 2.14 (CI: 1.08–7.65), p = 0.048 OR = 2.94 (CI: 1.08–8.02), p = 0.036
Abbreviations: BMI, body mass index; CI, confidence interval; OR, odds ratio; TNF‐α, tumour necrosis factor alpha; UC, ulcerative colitis.
3.3 Secondary outcomes
Among 124 patients in clinical response after induction, 65 (52.4%) maintained CCR through Week 54. Clinical remission at Week 54 was recorded in 40 (23.1%) patients. Among the 83 patients still on therapy after 1 year, CCR through Week 54 was associated with a lower likelihood of golimumab discontinuation throughout the subsequent year of therapy (23% with CCR vs. 61% without CCR; p < 0.01). No patients required colectomy after achieving CCR at week 54 compared to six patients not in CCR at Week 54 (p < 0.05).
Twenty‐two (12.7%) patients underwent total colectomy due to medical refractoriness after a median time of 28 weeks (range: 11–92 weeks) from golimumab initiation. Of these, 20 (90.9%) were anti‐TNF‐α experienced. Sixty (34.7%) patients were taking steroids at baseline: 36 (60%) were able to withdraw corticosteroids within 30 weeks. Among the remaining 24 patients, 21 (87.5%) withdrew from golimumab therapy during follow‐up.
At least one follow‐up endoscopy was performed in 119 (68.8%) patients after a median of 54 weeks (range: 8–122 weeks) from starting golimumab. Endoscopic remission was reported in 44/119 (36.9%) patients.
3.4 Golimumab safety
Twenty‐six AEI were reported by 21 (12.1%) patients. The most frequent AEI were infections (eight patients, 4.6%). Four patients had respiratory infections, one patient had acute gastroenteritis and one patient had genitourinary infection. Two patients experienced opportunistic infections: one experienced cytomegalovirus reactivation, and another was diagnosed with oropharyngeal candidiasis. The last two patients were on concomitant steroid therapy. Six (3.4%) patients developed skin manifestations (two psoriasis and four eczematous dermatitis). Four patients showed allergic reactions: one reaction at the injection site, and three diffuse skin rashes. One patient was diagnosed with oral condyloma, and one with basal‐cell carcinoma. Sixteen patients discontinued golimumab due to an AEI: five infections (three respiratory, one genitourinary and one candidiasis), six skin manifestation, four allergic reactions and one basal‐cell carcinoma.
4 DISCUSSION
This study focused on the long‐term clinical effectiveness and safety of a large cohort of 173 patients with moderate to severe active UC treated with golimumab. Most of our patients (60.7%) had extensive colitis, and more than a half (53.2%) had already been exposed to at least one anti‐TNF‐α agent. In our cohort, the median follow‐up on golimumab therapy was 52 weeks (range: 4–142 weeks), and the cumulative probability of maintaining golimumab treatment due to sustained clinical benefit was 47.3% and 22.5% at 54 and 108 weeks, respectively. These figures are different from other real‐world experiences, showing around up to 60% of persistence at Week 54. 8 , 11 However, the higher frequency of golimumab discontinuation in our study could be partially explained by the impossibility of escalating to 100 mg early in patients with a primary nonresponse or partial response during the maintenance phase. Most of our patients (75.7%) were in fact maintained with golimumab 50 mg because of their weight (<80 kg). We recorded a primary failure rate of up to 40.5% and 30% golimumab withdrawal within the first 14 weeks.
A post hoc analysis of the PURSUIT trial showed that up to 28.1% of Week 6 nonresponders who were escalated early to golimumab 100 mg achieved a clinical response at Week 14. Moreover, after 1 year, these late responders achieved similar clinical and endoscopic outcomes compared to early responders. Pharmacokinetic data showed that early Week 6 nonresponders had half the golimumab serum concentrations compared to early Week 6 responders. 16 Indeed, in their recent work, Magro et al. 17 found that Week 6 golimumab serum levels were positively correlated with clinical, endoscopic and histological remission, thus reinforcing the idea that early dose escalation could reduce the rates of primary nonresponse.
In our cohort, naive patients were more likely to maintain golimumab therapy because of a sustained clinical benefit compared to anti‐TNF‐α exposed patients. It should be noted that in about 37% of patients who were anti‐TNF‐α experienced, golimumab was used as a third‐line treatment after failure of infliximab and adalimumab. This situation has already been shown to be associated with a worse outcome compared to first‐ or second‐line utilisation. 8 Therefore, the use of golimumab should be advised at most after the failure of first‐line TNF‐α therapy. Therapeutic drug monitoring could help physicians to determine the most suitable therapeutic option in case of a loss of response to anti‐TNF‐α drugs, including switching within the class for patients with a high titre of neutralising anti‐drug antibodies or, conversely, out of class for patients with a ‘pharmacodynamics escape’ (trough levels within the therapeutic range with negative anti‐drug antibodies). 18
Most patients (79.2%) included in our study were steroid dependent. For such patients, golimumab was expected to provide a clinical improvement by exerting a steroid‐sparing effect as well. Among those who were taking steroids at baseline, the inability to discontinue them after 8 and 14 weeks of golimumab therapy was indeed associated with a higher rate of treatment discontinuation. Accordingly, we might suggest that in clinical practice, patients on golimumab therapy who still need steroids after 2–3 months or, similarly, require an early reintroduction should be revaluated for a therapeutic change.
CCR through Week 54 was observed in 65 (52.4%) patients comparable to those reported in the clinical trial. 5 Achieving CCR was associated with a higher rate of long‐term persistence on golimumab therapy. Moreover, none of the CCR patients underwent colectomy in the subsequent year.
The outcome of CCR, introduced for the first time in the PURSUIT study, also represents a potential goal for the treatment of UC patients in clinical practice, since it is based on the concept of tight monitoring of patients and of targeting continuous disease control. 19 Even though the evidence supporting that uncontrolled inflammation causes structural bowel damages are limited in comparison with Crohn's disease, 20 UC shows features of a progressive disease, including the proximal extension and the developing of structuring or functional disorders. 21 , 22
Finally, the overall safety profile of golimumab was confirmed to be good, consistent with those reported in other real‐life experiences and of other anti‐TNF‐alpha drugs. 8 , 12 , 16 No new safety concerns about golimumab emerged during our two years of follow‐up.
Our study has some limitations: as described above, including the impossibility of adapting the dose in patients with a partial or lack of response, but also a lack of data on inflammation markers (e.g., C‐reactive protein, faecal calprotectin). Conversely, the strengths of our study are the follow‐up of up to 2 years (median 52 weeks, range: 4–142 weeks), predefined standardised intervals between each clinical visit and homogeneous assessments of clinical and endoscopic activities. Moreover, we reported, for the first time to our knowledge, data on CCR in the real‐life setting and its correlation with a more favourable long‐term outcome.
In conclusion, golimumab may be considered as an effective and safe treatment option in UC patients, with higher rate of retention in therapy for biological‐naive patients and for those who are able to discontinue steroids early. CCR could potentially represent a target to pursue in clinical practice in order to improve disease control.
CONFLICT OF INTERESTS
The authors declare the following conflicts of interest: Daniela Pugliese received speaker fees from AbbVie, MSD, Takeda, Janssen and Pfeizer. Giuseppe Privitera received consultancies fees from Alphasigma. Mariangela Allocca received consulting fees from Nikkiso Europe and lecture fees from Janssen, Abbvie and Pfizer. Maria Cappello served as an advisory board member for AbbVie, MSD and Takeda Pharmaceuticals, and received lecture grants from AbbVie, MSD, Chiesi and Takeda Pharmaceuticals. Marco Daperno received lectures, board and/or congress fees from Abbvie, Pfizer, Takeda, Mundipharma, Janssen, MS&D, SOFAR, Ferring and Chiesi. Maria Di Girolamo received speaker fees from Abbvie. Fernando Rizzello acted as consultant for Janssen, Abbvie, Takeda, MSD and Amgen, and participated in a speaker's bureau sponsored by Abbvie, Janssen, Takeda, Ferring, MSD, Sofar and Chiesi. Alessandro Armuzzi received consulting and/or advisory board fees from AbbVie, Allergan, Amgen, Biogen, Bristol‐Myers Squibb, Celgene, Celltrion, Ferring, Janssen, Lilly, MSD, Mylan, Pfizer, Samsung Bioepis, Sandoz and Takeda; lecture and/or speaker bureau fees from AbbVie, Amgen, Biogen, Ferring, Janssen, MSD, Mitsubishi‐Tanabe, Nikkiso, Pfizer, Sandoz, Samsung Bioepis and Takeda; and research grants from MSD, Pfizer and Takeda. All the other authors have no conflict of interest to declare.
ACKNOWLEDGMENTS
Ennio Sarli provided statistical consulting. | Subcutaneous | DrugAdministrationRoute | CC BY-NC-ND | 33203342 | 18,572,703 | 2021-02 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Diabetes mellitus inadequate control'. | Alopecia areata universalis under treatment with sitagliptin : Possible immunological effect of dipeptidyl peptidase-4 inhibitors?
A 64-year old man developed alopecia universalis after one month of treatment with metformin and sitagliptin, a dipeptidyl peptidase‑4 (DPP-4) inhibitor. Diabetes treatment was changed to another genericum of sitagliptin and dapagliflozin. Following our recommendation, sitagliptin was interrupted and monotherapy with dapagliflozin was continued. After 6 weeks, sitagliptin was reassumed due to unsatisfactory diabetes control. Alopecia did not improve. We suspect a connection between DPP‑4 inhibition and development of alopecia due to its immunological potential. We assume that the treatment interruption might have been too short to induce regrowth of hair. DPP‑4 may result in both inhibition and activation of the immune system.
Anamnese
Wir berichten über den Fall eines 64-jährigen Patienten mit Erstdiagnose eines Diabetes mellitus Typ 2. Die Initialtherapie erfolgte mit Metformin 1000 mg und dem Dipeptidylpeptidase-4(DPP-4)-Inhibitor Sitagliptin 50 mg täglich. Nach 1 Monat entwickelte der Patient eine Alopecia universalis, die Maximalvariante einer Alopecia areata, mit vollständigem Verlust der gesamten Körper- und Gesichtsbehaarung. Andere Medikamente wurden nach Angaben des Patienten mit Ausnahme von Candesartan im vorangegangenen halben Jahr nicht eingenommen. Vorangegangene Alopezieepisoden wurden verneint, ebenso eine Atopie in der Eigen- oder Familienanamnese. Nach 1 Monat wurde die Diabetestherapie auf 10 mg Dapagliflozin und erneut 50 mg Sitagliptin (anderer Handelsname) jeweils täglich umgestellt. Zwei Monate später stellte sich der Patient in unserer Klinik vor.
Befund
Es zeigte sich ein vollständiger narbenloser Verlust der Körperbehaarung. Exemplarisch ist das Gesicht in Abb. 1 dargestellt. Schilddrüsenparameter, -autoantikörper sowie ANA(antinukleäre Antikörper)-Titer waren normwertig bzw. negativ.
Diagnose
Alopecia universalis, am ehesten als unerwünschte Arzneimittelwirkung (UAW) auf Sitagliptin.
Therapie und Verlauf
Aufgrund vorheriger Fallberichte [3, 11] vermuteten wir einen Zusammenhang zwischen der Neueinnahme von Sitagliptin und der Alopezie. Nach unserer Empfehlung erfolgte seitens des Diabetologen eine Umstellung auf eine Monotherapie mit Dapagliflozin. Nach 6 Wochen wurde die Therapie mit Sitagliptin aufgrund eines unzureichend eingestellten Diabetes erneut eingeleitet.
Beobachtung
Im 6‑wöchigen, therapiefreien Intervall blieb eine Besserung der Alopecia aus. Bis dato ist diese bestehend.
Diskussion
Die Alopecia universalis wird wegen eines Verlustes des immunologischen Privilegs des Haarfollikels mit konsekutiver autoimmunologischer Inflammation als Autoimmunerkrankung angesehen [10]. Während des Anagens, der Haarwachstumsphase, umzingeln T‑Helfer-Zellen, zytotoxische T‑Zellen, natürliche Killer(NK)- und plasmazytoide dendritische Zellen den unteren Teil des Follikels [1, 13]. In der Vergangenheit wurden genomische Regionen identifiziert, die mit der Alopecia areata assoziiert sind. In diesen Regionen werden unter anderem das zytotoxische T‑Lymphozyten-assoziierte Antigen 4 (CTLA-4), Interleukin(IL)-2, IL-21, IL-2-Rezeptor A und Eos (bekannt als IKZF4) kodiert [10]. Assoziationen bestehen auch für Gene des Haarfollikels (PRDX5 und STX17) und das ULBP(„cytomegalovirus UL16-binding protein“)-Gen-Cluster. ULBP wiederum aktiviert den NK-Rezeptor-Liganden NKG2D und induziert dadurch möglicherweise eine autoimmunologische Reaktion.
Eine Dysregulation der Immunogenität des Haarfollikels kann grundsätzlich durch bestimmte Zytokine regulatorische Vorgänge des Immunsystems beeinflussen oder einen proinflammatorischen Zustand mit konsekutiver Immunantwort begünstigen.
Sitagliptin, ein DPP-4-Inhibitor, und Metformin, ein Biguanid, sind etablierte Therapieoptionen des Diabetes mellitus Typ 2. Die Naranjo-Wahrscheinlichkeitsskala, ein Punktesystem für die Wahrscheinlichkeit einer UAW, erreicht in unserem Fall für Sitagliptin einen Punktwert von 5/13 und in Bezug auf Metformin 2/13. Es besteht daher eine stärkere Assoziation zwischen der Alopezie und Sitagliptin als kausalem Faktor.
Unsere Literaturrecherche ergab 2 Kasuistiken mit möglicher Assoziation zwischen der Einnahme von Sitagliptin bzw. Metformin und dem Auftreten einer Alopezie. Der erste Fall handelt von einer Patientin, die eine akute Alopezie in Form von Ausdünnen des Haupthaares 3 Monate nach Einnahme von Metformin (Dosierung unbekannt) und einem weiteren Monat mit Dosissteigerung auf 1000 mg 2‑mal täglich entwickelt hatte [3]. Sechs Monate nach Absetzen der Medikation war die Alopezie vollständig regredient. Der zweite Fall handelt von einem Patienten, der initial unter 4‑monatiger Therapie mit Metformin keine UAW zeigte. Vier Monate später, nach Beginn einer kombinierten Therapie mit 50 mg Sitagliptin und 850 mg Metformin, zeigten sich eine Alopezie der Augenbrauen, schnell fortschreitendes Grauwerden des Haupthaars, der Brustbehaarung sowie ein verlangsamtes Bartwachstum [11]. Unter Monotherapie mit Metformin (2850 mg täglich) war die Alopezie 3 Monate später reversibel. Beide Fälle wurden von nichtdermatologischen Fachärzten publiziert und betreut, sodass die klinisch dermatologische Befundbeschreibung nicht eindeutig ist.
Aufgrund der zahlreichen Fixkombinationen aus Gliptinen und Metformin ist eine eindeutige kausale Zuordnung zu einem Auslöser in unserem Fall nicht sicher möglich. Denkbar wäre auch eine Beeinflussung durch Kombination beider Präparate. Aufgrund des autoimmunologischen Interaktionspotenzials vermuten wir jedoch Sitagliptin als mitverantwortlichen Auslöser der Alopezie. Pathophysiologisch kann die Inhibition von DPP‑4 durch Sitagliptin die regulatorische Funktion von DPP‑4 auf T‑Zellen (hier als T‑Zell-Aktivator), stimulierten NK-Zellen, aktivierten B‑Zellen, dendritischen Zellen, Monozyten und Makrophagen stören [5]. DPP‑4 spielt als Korezeptor für CD8+-Zellen eine wichtige Rolle bei der erworbenen Immunantwort, der Gedächtnis-T-Zell-Generierung und -Emigration sowie während der Immunseneszenz [5]. Eine Verbindung zu immunologischen Erkrankungen wie Psoriasis, rheumatoide Arthritis, Lupus erythematodes, Sjögren-Syndrom oder dem allergischen Asthma wird beschrieben [5].
Gliptine zeigen teilweise auch ein protektives und immunsuppressives Potenzial. So vermittelt Sitagliptin einen immunsuppressiven Effekt durch Inhibition der T‑Zell-Rezeptor-Signaltransduktion und Proliferation von CD4+- und CD8+-Zellen [4].
In einem Tiermodell zeigte Linagliptin eine protektive Wirkung bezüglich einer Alopezie [2]. In diesem Modell resultierte die Alopezie aus einem aktivierten Wnt-Signalweg [8], der die Homöostase von Stammzellen negativ beeinflusst. Linagliptin scheint diesen Signalweg zu antagonisieren. Vermutet wird eine Inhibition über die Wnt/β-Catenin-Kaskade, die immunologisch zu einer Aktivierung von dendritischen Zellen mit IL-10-Sekretion und Th1- und Th17-Hemmung führt [12].
Im Gegensatz hierzu werden DPP-4-Inhibitoren ebenfalls mit dem Auftreten eines bullösen Pemphigoids (BP) in Verbindung gebracht [6, 7]. Es wird eine Inhibition von DPP‑8 und DPP‑9 durch Vildagliptin wegen einer geringen Selektivität innerhalb der DPPs vermutet [7]. Dies bewirkt eine konsekutive Aktivierung des Caspase-1-Signalwegs, die zu einem erhöhten Risiko der Entstehung eines BP beisteuert. Andererseits kann DPP‑4 proinflammatorische Zytokine wie TNF (Tumor-Nekrose-Faktor), IL-1β, IL-22, IL-17, IL-23 durch die Peptidaseaktivität deaktivieren [9]. Eine Inhibition kann in einer erhöhten Verfügbarkeit dieser Zytokine resultieren und folglich zu einer Aktivierung einer chronisch inflammatorischen Antwort und/oder Verschlimmerung eines autoimmunologischen Prozesses führen.
Der exakte Einfluss auf Autoimmunologie und immunologische Prozesse ist nicht hinreichend geklärt, da sowohl Aktivierung als auch Inhibition möglich scheinen.
Zusammenfassend vermuten wir in diesem Fall Sitagliptin als auslösenden Faktor für die Alopezie, weil DPP‑4 einen wichtigen Stellenwert hinsichtlich des Immunsystems und der Autoimmunität besitzt. Das therapiefreie Intervall von 6 Wochen vor erneuter Gabe von Sitagliptin erscheint zu kurz, um einen Besserungseffekt zu erzielen. In Fällen, in denen eine immunologische Erkrankung entsteht oder sich verschlimmert, sollte eine Verbindung zu DPP-4-Inhibitoren erwogen werden.
Fazit für die Praxis
Neben einer Assoziation zum bullösen Pemphigoid könnten ebenfalls andere autoimmunologische Erkrankungen wie eine Alopezie mit DPP-4(Dipeptidylpeptidase-4)-Inhibitoren in Verbindung stehen.
Bei Neuauftreten oder Verschlimmerung von vorbestehenden autoimmunologischen Erkrankungen sollte eine Assoziation zu DPP-4-Inhibitoren berücksichtigt werden.
Der genaue Einfluss (Hemmung oder Aktivierung) von DPP‑4 auf das Immunsystem ist nicht hinreichend geklärt, sodass weitere Untersuchungen sinnvoll erscheinen.
Funding
Open Access funding enabled and organized by Projekt DEAL.
Einhaltung ethischer Richtlinien
Interessenkonflikt
J. Kohlmann, R.A. Ferrer, A. Markovic, M. Illes und M. Kunz geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autoren keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien. Für Bildmaterial oder anderweitige Angaben innerhalb des Manuskripts, über die Patienten zu identifizieren sind, liegt von ihnen und/oder ihren gesetzlichen Vertretern eine schriftliche Einwilligung vor. | CANDESARTAN, DAPAGLIFLOZIN, METFORMIN HYDROCHLORIDE, SITAGLIPTIN | DrugsGivenReaction | CC BY | 33205256 | 18,647,979 | 2021-07 |
What was the administration route of drug 'CANDESARTAN'? | Alopecia areata universalis under treatment with sitagliptin : Possible immunological effect of dipeptidyl peptidase-4 inhibitors?
A 64-year old man developed alopecia universalis after one month of treatment with metformin and sitagliptin, a dipeptidyl peptidase‑4 (DPP-4) inhibitor. Diabetes treatment was changed to another genericum of sitagliptin and dapagliflozin. Following our recommendation, sitagliptin was interrupted and monotherapy with dapagliflozin was continued. After 6 weeks, sitagliptin was reassumed due to unsatisfactory diabetes control. Alopecia did not improve. We suspect a connection between DPP‑4 inhibition and development of alopecia due to its immunological potential. We assume that the treatment interruption might have been too short to induce regrowth of hair. DPP‑4 may result in both inhibition and activation of the immune system.
Anamnese
Wir berichten über den Fall eines 64-jährigen Patienten mit Erstdiagnose eines Diabetes mellitus Typ 2. Die Initialtherapie erfolgte mit Metformin 1000 mg und dem Dipeptidylpeptidase-4(DPP-4)-Inhibitor Sitagliptin 50 mg täglich. Nach 1 Monat entwickelte der Patient eine Alopecia universalis, die Maximalvariante einer Alopecia areata, mit vollständigem Verlust der gesamten Körper- und Gesichtsbehaarung. Andere Medikamente wurden nach Angaben des Patienten mit Ausnahme von Candesartan im vorangegangenen halben Jahr nicht eingenommen. Vorangegangene Alopezieepisoden wurden verneint, ebenso eine Atopie in der Eigen- oder Familienanamnese. Nach 1 Monat wurde die Diabetestherapie auf 10 mg Dapagliflozin und erneut 50 mg Sitagliptin (anderer Handelsname) jeweils täglich umgestellt. Zwei Monate später stellte sich der Patient in unserer Klinik vor.
Befund
Es zeigte sich ein vollständiger narbenloser Verlust der Körperbehaarung. Exemplarisch ist das Gesicht in Abb. 1 dargestellt. Schilddrüsenparameter, -autoantikörper sowie ANA(antinukleäre Antikörper)-Titer waren normwertig bzw. negativ.
Diagnose
Alopecia universalis, am ehesten als unerwünschte Arzneimittelwirkung (UAW) auf Sitagliptin.
Therapie und Verlauf
Aufgrund vorheriger Fallberichte [3, 11] vermuteten wir einen Zusammenhang zwischen der Neueinnahme von Sitagliptin und der Alopezie. Nach unserer Empfehlung erfolgte seitens des Diabetologen eine Umstellung auf eine Monotherapie mit Dapagliflozin. Nach 6 Wochen wurde die Therapie mit Sitagliptin aufgrund eines unzureichend eingestellten Diabetes erneut eingeleitet.
Beobachtung
Im 6‑wöchigen, therapiefreien Intervall blieb eine Besserung der Alopecia aus. Bis dato ist diese bestehend.
Diskussion
Die Alopecia universalis wird wegen eines Verlustes des immunologischen Privilegs des Haarfollikels mit konsekutiver autoimmunologischer Inflammation als Autoimmunerkrankung angesehen [10]. Während des Anagens, der Haarwachstumsphase, umzingeln T‑Helfer-Zellen, zytotoxische T‑Zellen, natürliche Killer(NK)- und plasmazytoide dendritische Zellen den unteren Teil des Follikels [1, 13]. In der Vergangenheit wurden genomische Regionen identifiziert, die mit der Alopecia areata assoziiert sind. In diesen Regionen werden unter anderem das zytotoxische T‑Lymphozyten-assoziierte Antigen 4 (CTLA-4), Interleukin(IL)-2, IL-21, IL-2-Rezeptor A und Eos (bekannt als IKZF4) kodiert [10]. Assoziationen bestehen auch für Gene des Haarfollikels (PRDX5 und STX17) und das ULBP(„cytomegalovirus UL16-binding protein“)-Gen-Cluster. ULBP wiederum aktiviert den NK-Rezeptor-Liganden NKG2D und induziert dadurch möglicherweise eine autoimmunologische Reaktion.
Eine Dysregulation der Immunogenität des Haarfollikels kann grundsätzlich durch bestimmte Zytokine regulatorische Vorgänge des Immunsystems beeinflussen oder einen proinflammatorischen Zustand mit konsekutiver Immunantwort begünstigen.
Sitagliptin, ein DPP-4-Inhibitor, und Metformin, ein Biguanid, sind etablierte Therapieoptionen des Diabetes mellitus Typ 2. Die Naranjo-Wahrscheinlichkeitsskala, ein Punktesystem für die Wahrscheinlichkeit einer UAW, erreicht in unserem Fall für Sitagliptin einen Punktwert von 5/13 und in Bezug auf Metformin 2/13. Es besteht daher eine stärkere Assoziation zwischen der Alopezie und Sitagliptin als kausalem Faktor.
Unsere Literaturrecherche ergab 2 Kasuistiken mit möglicher Assoziation zwischen der Einnahme von Sitagliptin bzw. Metformin und dem Auftreten einer Alopezie. Der erste Fall handelt von einer Patientin, die eine akute Alopezie in Form von Ausdünnen des Haupthaares 3 Monate nach Einnahme von Metformin (Dosierung unbekannt) und einem weiteren Monat mit Dosissteigerung auf 1000 mg 2‑mal täglich entwickelt hatte [3]. Sechs Monate nach Absetzen der Medikation war die Alopezie vollständig regredient. Der zweite Fall handelt von einem Patienten, der initial unter 4‑monatiger Therapie mit Metformin keine UAW zeigte. Vier Monate später, nach Beginn einer kombinierten Therapie mit 50 mg Sitagliptin und 850 mg Metformin, zeigten sich eine Alopezie der Augenbrauen, schnell fortschreitendes Grauwerden des Haupthaars, der Brustbehaarung sowie ein verlangsamtes Bartwachstum [11]. Unter Monotherapie mit Metformin (2850 mg täglich) war die Alopezie 3 Monate später reversibel. Beide Fälle wurden von nichtdermatologischen Fachärzten publiziert und betreut, sodass die klinisch dermatologische Befundbeschreibung nicht eindeutig ist.
Aufgrund der zahlreichen Fixkombinationen aus Gliptinen und Metformin ist eine eindeutige kausale Zuordnung zu einem Auslöser in unserem Fall nicht sicher möglich. Denkbar wäre auch eine Beeinflussung durch Kombination beider Präparate. Aufgrund des autoimmunologischen Interaktionspotenzials vermuten wir jedoch Sitagliptin als mitverantwortlichen Auslöser der Alopezie. Pathophysiologisch kann die Inhibition von DPP‑4 durch Sitagliptin die regulatorische Funktion von DPP‑4 auf T‑Zellen (hier als T‑Zell-Aktivator), stimulierten NK-Zellen, aktivierten B‑Zellen, dendritischen Zellen, Monozyten und Makrophagen stören [5]. DPP‑4 spielt als Korezeptor für CD8+-Zellen eine wichtige Rolle bei der erworbenen Immunantwort, der Gedächtnis-T-Zell-Generierung und -Emigration sowie während der Immunseneszenz [5]. Eine Verbindung zu immunologischen Erkrankungen wie Psoriasis, rheumatoide Arthritis, Lupus erythematodes, Sjögren-Syndrom oder dem allergischen Asthma wird beschrieben [5].
Gliptine zeigen teilweise auch ein protektives und immunsuppressives Potenzial. So vermittelt Sitagliptin einen immunsuppressiven Effekt durch Inhibition der T‑Zell-Rezeptor-Signaltransduktion und Proliferation von CD4+- und CD8+-Zellen [4].
In einem Tiermodell zeigte Linagliptin eine protektive Wirkung bezüglich einer Alopezie [2]. In diesem Modell resultierte die Alopezie aus einem aktivierten Wnt-Signalweg [8], der die Homöostase von Stammzellen negativ beeinflusst. Linagliptin scheint diesen Signalweg zu antagonisieren. Vermutet wird eine Inhibition über die Wnt/β-Catenin-Kaskade, die immunologisch zu einer Aktivierung von dendritischen Zellen mit IL-10-Sekretion und Th1- und Th17-Hemmung führt [12].
Im Gegensatz hierzu werden DPP-4-Inhibitoren ebenfalls mit dem Auftreten eines bullösen Pemphigoids (BP) in Verbindung gebracht [6, 7]. Es wird eine Inhibition von DPP‑8 und DPP‑9 durch Vildagliptin wegen einer geringen Selektivität innerhalb der DPPs vermutet [7]. Dies bewirkt eine konsekutive Aktivierung des Caspase-1-Signalwegs, die zu einem erhöhten Risiko der Entstehung eines BP beisteuert. Andererseits kann DPP‑4 proinflammatorische Zytokine wie TNF (Tumor-Nekrose-Faktor), IL-1β, IL-22, IL-17, IL-23 durch die Peptidaseaktivität deaktivieren [9]. Eine Inhibition kann in einer erhöhten Verfügbarkeit dieser Zytokine resultieren und folglich zu einer Aktivierung einer chronisch inflammatorischen Antwort und/oder Verschlimmerung eines autoimmunologischen Prozesses führen.
Der exakte Einfluss auf Autoimmunologie und immunologische Prozesse ist nicht hinreichend geklärt, da sowohl Aktivierung als auch Inhibition möglich scheinen.
Zusammenfassend vermuten wir in diesem Fall Sitagliptin als auslösenden Faktor für die Alopezie, weil DPP‑4 einen wichtigen Stellenwert hinsichtlich des Immunsystems und der Autoimmunität besitzt. Das therapiefreie Intervall von 6 Wochen vor erneuter Gabe von Sitagliptin erscheint zu kurz, um einen Besserungseffekt zu erzielen. In Fällen, in denen eine immunologische Erkrankung entsteht oder sich verschlimmert, sollte eine Verbindung zu DPP-4-Inhibitoren erwogen werden.
Fazit für die Praxis
Neben einer Assoziation zum bullösen Pemphigoid könnten ebenfalls andere autoimmunologische Erkrankungen wie eine Alopezie mit DPP-4(Dipeptidylpeptidase-4)-Inhibitoren in Verbindung stehen.
Bei Neuauftreten oder Verschlimmerung von vorbestehenden autoimmunologischen Erkrankungen sollte eine Assoziation zu DPP-4-Inhibitoren berücksichtigt werden.
Der genaue Einfluss (Hemmung oder Aktivierung) von DPP‑4 auf das Immunsystem ist nicht hinreichend geklärt, sodass weitere Untersuchungen sinnvoll erscheinen.
Funding
Open Access funding enabled and organized by Projekt DEAL.
Einhaltung ethischer Richtlinien
Interessenkonflikt
J. Kohlmann, R.A. Ferrer, A. Markovic, M. Illes und M. Kunz geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autoren keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien. Für Bildmaterial oder anderweitige Angaben innerhalb des Manuskripts, über die Patienten zu identifizieren sind, liegt von ihnen und/oder ihren gesetzlichen Vertretern eine schriftliche Einwilligung vor. | Oral | DrugAdministrationRoute | CC BY | 33205256 | 18,647,979 | 2021-07 |
What was the administration route of drug 'DAPAGLIFLOZIN'? | Alopecia areata universalis under treatment with sitagliptin : Possible immunological effect of dipeptidyl peptidase-4 inhibitors?
A 64-year old man developed alopecia universalis after one month of treatment with metformin and sitagliptin, a dipeptidyl peptidase‑4 (DPP-4) inhibitor. Diabetes treatment was changed to another genericum of sitagliptin and dapagliflozin. Following our recommendation, sitagliptin was interrupted and monotherapy with dapagliflozin was continued. After 6 weeks, sitagliptin was reassumed due to unsatisfactory diabetes control. Alopecia did not improve. We suspect a connection between DPP‑4 inhibition and development of alopecia due to its immunological potential. We assume that the treatment interruption might have been too short to induce regrowth of hair. DPP‑4 may result in both inhibition and activation of the immune system.
Anamnese
Wir berichten über den Fall eines 64-jährigen Patienten mit Erstdiagnose eines Diabetes mellitus Typ 2. Die Initialtherapie erfolgte mit Metformin 1000 mg und dem Dipeptidylpeptidase-4(DPP-4)-Inhibitor Sitagliptin 50 mg täglich. Nach 1 Monat entwickelte der Patient eine Alopecia universalis, die Maximalvariante einer Alopecia areata, mit vollständigem Verlust der gesamten Körper- und Gesichtsbehaarung. Andere Medikamente wurden nach Angaben des Patienten mit Ausnahme von Candesartan im vorangegangenen halben Jahr nicht eingenommen. Vorangegangene Alopezieepisoden wurden verneint, ebenso eine Atopie in der Eigen- oder Familienanamnese. Nach 1 Monat wurde die Diabetestherapie auf 10 mg Dapagliflozin und erneut 50 mg Sitagliptin (anderer Handelsname) jeweils täglich umgestellt. Zwei Monate später stellte sich der Patient in unserer Klinik vor.
Befund
Es zeigte sich ein vollständiger narbenloser Verlust der Körperbehaarung. Exemplarisch ist das Gesicht in Abb. 1 dargestellt. Schilddrüsenparameter, -autoantikörper sowie ANA(antinukleäre Antikörper)-Titer waren normwertig bzw. negativ.
Diagnose
Alopecia universalis, am ehesten als unerwünschte Arzneimittelwirkung (UAW) auf Sitagliptin.
Therapie und Verlauf
Aufgrund vorheriger Fallberichte [3, 11] vermuteten wir einen Zusammenhang zwischen der Neueinnahme von Sitagliptin und der Alopezie. Nach unserer Empfehlung erfolgte seitens des Diabetologen eine Umstellung auf eine Monotherapie mit Dapagliflozin. Nach 6 Wochen wurde die Therapie mit Sitagliptin aufgrund eines unzureichend eingestellten Diabetes erneut eingeleitet.
Beobachtung
Im 6‑wöchigen, therapiefreien Intervall blieb eine Besserung der Alopecia aus. Bis dato ist diese bestehend.
Diskussion
Die Alopecia universalis wird wegen eines Verlustes des immunologischen Privilegs des Haarfollikels mit konsekutiver autoimmunologischer Inflammation als Autoimmunerkrankung angesehen [10]. Während des Anagens, der Haarwachstumsphase, umzingeln T‑Helfer-Zellen, zytotoxische T‑Zellen, natürliche Killer(NK)- und plasmazytoide dendritische Zellen den unteren Teil des Follikels [1, 13]. In der Vergangenheit wurden genomische Regionen identifiziert, die mit der Alopecia areata assoziiert sind. In diesen Regionen werden unter anderem das zytotoxische T‑Lymphozyten-assoziierte Antigen 4 (CTLA-4), Interleukin(IL)-2, IL-21, IL-2-Rezeptor A und Eos (bekannt als IKZF4) kodiert [10]. Assoziationen bestehen auch für Gene des Haarfollikels (PRDX5 und STX17) und das ULBP(„cytomegalovirus UL16-binding protein“)-Gen-Cluster. ULBP wiederum aktiviert den NK-Rezeptor-Liganden NKG2D und induziert dadurch möglicherweise eine autoimmunologische Reaktion.
Eine Dysregulation der Immunogenität des Haarfollikels kann grundsätzlich durch bestimmte Zytokine regulatorische Vorgänge des Immunsystems beeinflussen oder einen proinflammatorischen Zustand mit konsekutiver Immunantwort begünstigen.
Sitagliptin, ein DPP-4-Inhibitor, und Metformin, ein Biguanid, sind etablierte Therapieoptionen des Diabetes mellitus Typ 2. Die Naranjo-Wahrscheinlichkeitsskala, ein Punktesystem für die Wahrscheinlichkeit einer UAW, erreicht in unserem Fall für Sitagliptin einen Punktwert von 5/13 und in Bezug auf Metformin 2/13. Es besteht daher eine stärkere Assoziation zwischen der Alopezie und Sitagliptin als kausalem Faktor.
Unsere Literaturrecherche ergab 2 Kasuistiken mit möglicher Assoziation zwischen der Einnahme von Sitagliptin bzw. Metformin und dem Auftreten einer Alopezie. Der erste Fall handelt von einer Patientin, die eine akute Alopezie in Form von Ausdünnen des Haupthaares 3 Monate nach Einnahme von Metformin (Dosierung unbekannt) und einem weiteren Monat mit Dosissteigerung auf 1000 mg 2‑mal täglich entwickelt hatte [3]. Sechs Monate nach Absetzen der Medikation war die Alopezie vollständig regredient. Der zweite Fall handelt von einem Patienten, der initial unter 4‑monatiger Therapie mit Metformin keine UAW zeigte. Vier Monate später, nach Beginn einer kombinierten Therapie mit 50 mg Sitagliptin und 850 mg Metformin, zeigten sich eine Alopezie der Augenbrauen, schnell fortschreitendes Grauwerden des Haupthaars, der Brustbehaarung sowie ein verlangsamtes Bartwachstum [11]. Unter Monotherapie mit Metformin (2850 mg täglich) war die Alopezie 3 Monate später reversibel. Beide Fälle wurden von nichtdermatologischen Fachärzten publiziert und betreut, sodass die klinisch dermatologische Befundbeschreibung nicht eindeutig ist.
Aufgrund der zahlreichen Fixkombinationen aus Gliptinen und Metformin ist eine eindeutige kausale Zuordnung zu einem Auslöser in unserem Fall nicht sicher möglich. Denkbar wäre auch eine Beeinflussung durch Kombination beider Präparate. Aufgrund des autoimmunologischen Interaktionspotenzials vermuten wir jedoch Sitagliptin als mitverantwortlichen Auslöser der Alopezie. Pathophysiologisch kann die Inhibition von DPP‑4 durch Sitagliptin die regulatorische Funktion von DPP‑4 auf T‑Zellen (hier als T‑Zell-Aktivator), stimulierten NK-Zellen, aktivierten B‑Zellen, dendritischen Zellen, Monozyten und Makrophagen stören [5]. DPP‑4 spielt als Korezeptor für CD8+-Zellen eine wichtige Rolle bei der erworbenen Immunantwort, der Gedächtnis-T-Zell-Generierung und -Emigration sowie während der Immunseneszenz [5]. Eine Verbindung zu immunologischen Erkrankungen wie Psoriasis, rheumatoide Arthritis, Lupus erythematodes, Sjögren-Syndrom oder dem allergischen Asthma wird beschrieben [5].
Gliptine zeigen teilweise auch ein protektives und immunsuppressives Potenzial. So vermittelt Sitagliptin einen immunsuppressiven Effekt durch Inhibition der T‑Zell-Rezeptor-Signaltransduktion und Proliferation von CD4+- und CD8+-Zellen [4].
In einem Tiermodell zeigte Linagliptin eine protektive Wirkung bezüglich einer Alopezie [2]. In diesem Modell resultierte die Alopezie aus einem aktivierten Wnt-Signalweg [8], der die Homöostase von Stammzellen negativ beeinflusst. Linagliptin scheint diesen Signalweg zu antagonisieren. Vermutet wird eine Inhibition über die Wnt/β-Catenin-Kaskade, die immunologisch zu einer Aktivierung von dendritischen Zellen mit IL-10-Sekretion und Th1- und Th17-Hemmung führt [12].
Im Gegensatz hierzu werden DPP-4-Inhibitoren ebenfalls mit dem Auftreten eines bullösen Pemphigoids (BP) in Verbindung gebracht [6, 7]. Es wird eine Inhibition von DPP‑8 und DPP‑9 durch Vildagliptin wegen einer geringen Selektivität innerhalb der DPPs vermutet [7]. Dies bewirkt eine konsekutive Aktivierung des Caspase-1-Signalwegs, die zu einem erhöhten Risiko der Entstehung eines BP beisteuert. Andererseits kann DPP‑4 proinflammatorische Zytokine wie TNF (Tumor-Nekrose-Faktor), IL-1β, IL-22, IL-17, IL-23 durch die Peptidaseaktivität deaktivieren [9]. Eine Inhibition kann in einer erhöhten Verfügbarkeit dieser Zytokine resultieren und folglich zu einer Aktivierung einer chronisch inflammatorischen Antwort und/oder Verschlimmerung eines autoimmunologischen Prozesses führen.
Der exakte Einfluss auf Autoimmunologie und immunologische Prozesse ist nicht hinreichend geklärt, da sowohl Aktivierung als auch Inhibition möglich scheinen.
Zusammenfassend vermuten wir in diesem Fall Sitagliptin als auslösenden Faktor für die Alopezie, weil DPP‑4 einen wichtigen Stellenwert hinsichtlich des Immunsystems und der Autoimmunität besitzt. Das therapiefreie Intervall von 6 Wochen vor erneuter Gabe von Sitagliptin erscheint zu kurz, um einen Besserungseffekt zu erzielen. In Fällen, in denen eine immunologische Erkrankung entsteht oder sich verschlimmert, sollte eine Verbindung zu DPP-4-Inhibitoren erwogen werden.
Fazit für die Praxis
Neben einer Assoziation zum bullösen Pemphigoid könnten ebenfalls andere autoimmunologische Erkrankungen wie eine Alopezie mit DPP-4(Dipeptidylpeptidase-4)-Inhibitoren in Verbindung stehen.
Bei Neuauftreten oder Verschlimmerung von vorbestehenden autoimmunologischen Erkrankungen sollte eine Assoziation zu DPP-4-Inhibitoren berücksichtigt werden.
Der genaue Einfluss (Hemmung oder Aktivierung) von DPP‑4 auf das Immunsystem ist nicht hinreichend geklärt, sodass weitere Untersuchungen sinnvoll erscheinen.
Funding
Open Access funding enabled and organized by Projekt DEAL.
Einhaltung ethischer Richtlinien
Interessenkonflikt
J. Kohlmann, R.A. Ferrer, A. Markovic, M. Illes und M. Kunz geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autoren keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien. Für Bildmaterial oder anderweitige Angaben innerhalb des Manuskripts, über die Patienten zu identifizieren sind, liegt von ihnen und/oder ihren gesetzlichen Vertretern eine schriftliche Einwilligung vor. | Oral | DrugAdministrationRoute | CC BY | 33205256 | 18,647,979 | 2021-07 |
What was the outcome of reaction 'Alopecia universalis'? | Alopecia areata universalis under treatment with sitagliptin : Possible immunological effect of dipeptidyl peptidase-4 inhibitors?
A 64-year old man developed alopecia universalis after one month of treatment with metformin and sitagliptin, a dipeptidyl peptidase‑4 (DPP-4) inhibitor. Diabetes treatment was changed to another genericum of sitagliptin and dapagliflozin. Following our recommendation, sitagliptin was interrupted and monotherapy with dapagliflozin was continued. After 6 weeks, sitagliptin was reassumed due to unsatisfactory diabetes control. Alopecia did not improve. We suspect a connection between DPP‑4 inhibition and development of alopecia due to its immunological potential. We assume that the treatment interruption might have been too short to induce regrowth of hair. DPP‑4 may result in both inhibition and activation of the immune system.
Anamnese
Wir berichten über den Fall eines 64-jährigen Patienten mit Erstdiagnose eines Diabetes mellitus Typ 2. Die Initialtherapie erfolgte mit Metformin 1000 mg und dem Dipeptidylpeptidase-4(DPP-4)-Inhibitor Sitagliptin 50 mg täglich. Nach 1 Monat entwickelte der Patient eine Alopecia universalis, die Maximalvariante einer Alopecia areata, mit vollständigem Verlust der gesamten Körper- und Gesichtsbehaarung. Andere Medikamente wurden nach Angaben des Patienten mit Ausnahme von Candesartan im vorangegangenen halben Jahr nicht eingenommen. Vorangegangene Alopezieepisoden wurden verneint, ebenso eine Atopie in der Eigen- oder Familienanamnese. Nach 1 Monat wurde die Diabetestherapie auf 10 mg Dapagliflozin und erneut 50 mg Sitagliptin (anderer Handelsname) jeweils täglich umgestellt. Zwei Monate später stellte sich der Patient in unserer Klinik vor.
Befund
Es zeigte sich ein vollständiger narbenloser Verlust der Körperbehaarung. Exemplarisch ist das Gesicht in Abb. 1 dargestellt. Schilddrüsenparameter, -autoantikörper sowie ANA(antinukleäre Antikörper)-Titer waren normwertig bzw. negativ.
Diagnose
Alopecia universalis, am ehesten als unerwünschte Arzneimittelwirkung (UAW) auf Sitagliptin.
Therapie und Verlauf
Aufgrund vorheriger Fallberichte [3, 11] vermuteten wir einen Zusammenhang zwischen der Neueinnahme von Sitagliptin und der Alopezie. Nach unserer Empfehlung erfolgte seitens des Diabetologen eine Umstellung auf eine Monotherapie mit Dapagliflozin. Nach 6 Wochen wurde die Therapie mit Sitagliptin aufgrund eines unzureichend eingestellten Diabetes erneut eingeleitet.
Beobachtung
Im 6‑wöchigen, therapiefreien Intervall blieb eine Besserung der Alopecia aus. Bis dato ist diese bestehend.
Diskussion
Die Alopecia universalis wird wegen eines Verlustes des immunologischen Privilegs des Haarfollikels mit konsekutiver autoimmunologischer Inflammation als Autoimmunerkrankung angesehen [10]. Während des Anagens, der Haarwachstumsphase, umzingeln T‑Helfer-Zellen, zytotoxische T‑Zellen, natürliche Killer(NK)- und plasmazytoide dendritische Zellen den unteren Teil des Follikels [1, 13]. In der Vergangenheit wurden genomische Regionen identifiziert, die mit der Alopecia areata assoziiert sind. In diesen Regionen werden unter anderem das zytotoxische T‑Lymphozyten-assoziierte Antigen 4 (CTLA-4), Interleukin(IL)-2, IL-21, IL-2-Rezeptor A und Eos (bekannt als IKZF4) kodiert [10]. Assoziationen bestehen auch für Gene des Haarfollikels (PRDX5 und STX17) und das ULBP(„cytomegalovirus UL16-binding protein“)-Gen-Cluster. ULBP wiederum aktiviert den NK-Rezeptor-Liganden NKG2D und induziert dadurch möglicherweise eine autoimmunologische Reaktion.
Eine Dysregulation der Immunogenität des Haarfollikels kann grundsätzlich durch bestimmte Zytokine regulatorische Vorgänge des Immunsystems beeinflussen oder einen proinflammatorischen Zustand mit konsekutiver Immunantwort begünstigen.
Sitagliptin, ein DPP-4-Inhibitor, und Metformin, ein Biguanid, sind etablierte Therapieoptionen des Diabetes mellitus Typ 2. Die Naranjo-Wahrscheinlichkeitsskala, ein Punktesystem für die Wahrscheinlichkeit einer UAW, erreicht in unserem Fall für Sitagliptin einen Punktwert von 5/13 und in Bezug auf Metformin 2/13. Es besteht daher eine stärkere Assoziation zwischen der Alopezie und Sitagliptin als kausalem Faktor.
Unsere Literaturrecherche ergab 2 Kasuistiken mit möglicher Assoziation zwischen der Einnahme von Sitagliptin bzw. Metformin und dem Auftreten einer Alopezie. Der erste Fall handelt von einer Patientin, die eine akute Alopezie in Form von Ausdünnen des Haupthaares 3 Monate nach Einnahme von Metformin (Dosierung unbekannt) und einem weiteren Monat mit Dosissteigerung auf 1000 mg 2‑mal täglich entwickelt hatte [3]. Sechs Monate nach Absetzen der Medikation war die Alopezie vollständig regredient. Der zweite Fall handelt von einem Patienten, der initial unter 4‑monatiger Therapie mit Metformin keine UAW zeigte. Vier Monate später, nach Beginn einer kombinierten Therapie mit 50 mg Sitagliptin und 850 mg Metformin, zeigten sich eine Alopezie der Augenbrauen, schnell fortschreitendes Grauwerden des Haupthaars, der Brustbehaarung sowie ein verlangsamtes Bartwachstum [11]. Unter Monotherapie mit Metformin (2850 mg täglich) war die Alopezie 3 Monate später reversibel. Beide Fälle wurden von nichtdermatologischen Fachärzten publiziert und betreut, sodass die klinisch dermatologische Befundbeschreibung nicht eindeutig ist.
Aufgrund der zahlreichen Fixkombinationen aus Gliptinen und Metformin ist eine eindeutige kausale Zuordnung zu einem Auslöser in unserem Fall nicht sicher möglich. Denkbar wäre auch eine Beeinflussung durch Kombination beider Präparate. Aufgrund des autoimmunologischen Interaktionspotenzials vermuten wir jedoch Sitagliptin als mitverantwortlichen Auslöser der Alopezie. Pathophysiologisch kann die Inhibition von DPP‑4 durch Sitagliptin die regulatorische Funktion von DPP‑4 auf T‑Zellen (hier als T‑Zell-Aktivator), stimulierten NK-Zellen, aktivierten B‑Zellen, dendritischen Zellen, Monozyten und Makrophagen stören [5]. DPP‑4 spielt als Korezeptor für CD8+-Zellen eine wichtige Rolle bei der erworbenen Immunantwort, der Gedächtnis-T-Zell-Generierung und -Emigration sowie während der Immunseneszenz [5]. Eine Verbindung zu immunologischen Erkrankungen wie Psoriasis, rheumatoide Arthritis, Lupus erythematodes, Sjögren-Syndrom oder dem allergischen Asthma wird beschrieben [5].
Gliptine zeigen teilweise auch ein protektives und immunsuppressives Potenzial. So vermittelt Sitagliptin einen immunsuppressiven Effekt durch Inhibition der T‑Zell-Rezeptor-Signaltransduktion und Proliferation von CD4+- und CD8+-Zellen [4].
In einem Tiermodell zeigte Linagliptin eine protektive Wirkung bezüglich einer Alopezie [2]. In diesem Modell resultierte die Alopezie aus einem aktivierten Wnt-Signalweg [8], der die Homöostase von Stammzellen negativ beeinflusst. Linagliptin scheint diesen Signalweg zu antagonisieren. Vermutet wird eine Inhibition über die Wnt/β-Catenin-Kaskade, die immunologisch zu einer Aktivierung von dendritischen Zellen mit IL-10-Sekretion und Th1- und Th17-Hemmung führt [12].
Im Gegensatz hierzu werden DPP-4-Inhibitoren ebenfalls mit dem Auftreten eines bullösen Pemphigoids (BP) in Verbindung gebracht [6, 7]. Es wird eine Inhibition von DPP‑8 und DPP‑9 durch Vildagliptin wegen einer geringen Selektivität innerhalb der DPPs vermutet [7]. Dies bewirkt eine konsekutive Aktivierung des Caspase-1-Signalwegs, die zu einem erhöhten Risiko der Entstehung eines BP beisteuert. Andererseits kann DPP‑4 proinflammatorische Zytokine wie TNF (Tumor-Nekrose-Faktor), IL-1β, IL-22, IL-17, IL-23 durch die Peptidaseaktivität deaktivieren [9]. Eine Inhibition kann in einer erhöhten Verfügbarkeit dieser Zytokine resultieren und folglich zu einer Aktivierung einer chronisch inflammatorischen Antwort und/oder Verschlimmerung eines autoimmunologischen Prozesses führen.
Der exakte Einfluss auf Autoimmunologie und immunologische Prozesse ist nicht hinreichend geklärt, da sowohl Aktivierung als auch Inhibition möglich scheinen.
Zusammenfassend vermuten wir in diesem Fall Sitagliptin als auslösenden Faktor für die Alopezie, weil DPP‑4 einen wichtigen Stellenwert hinsichtlich des Immunsystems und der Autoimmunität besitzt. Das therapiefreie Intervall von 6 Wochen vor erneuter Gabe von Sitagliptin erscheint zu kurz, um einen Besserungseffekt zu erzielen. In Fällen, in denen eine immunologische Erkrankung entsteht oder sich verschlimmert, sollte eine Verbindung zu DPP-4-Inhibitoren erwogen werden.
Fazit für die Praxis
Neben einer Assoziation zum bullösen Pemphigoid könnten ebenfalls andere autoimmunologische Erkrankungen wie eine Alopezie mit DPP-4(Dipeptidylpeptidase-4)-Inhibitoren in Verbindung stehen.
Bei Neuauftreten oder Verschlimmerung von vorbestehenden autoimmunologischen Erkrankungen sollte eine Assoziation zu DPP-4-Inhibitoren berücksichtigt werden.
Der genaue Einfluss (Hemmung oder Aktivierung) von DPP‑4 auf das Immunsystem ist nicht hinreichend geklärt, sodass weitere Untersuchungen sinnvoll erscheinen.
Funding
Open Access funding enabled and organized by Projekt DEAL.
Einhaltung ethischer Richtlinien
Interessenkonflikt
J. Kohlmann, R.A. Ferrer, A. Markovic, M. Illes und M. Kunz geben an, dass kein Interessenkonflikt besteht.
Für diesen Beitrag wurden von den Autoren keine Studien an Menschen oder Tieren durchgeführt. Für die aufgeführten Studien gelten die jeweils dort angegebenen ethischen Richtlinien. Für Bildmaterial oder anderweitige Angaben innerhalb des Manuskripts, über die Patienten zu identifizieren sind, liegt von ihnen und/oder ihren gesetzlichen Vertretern eine schriftliche Einwilligung vor. | Not recovered | ReactionOutcome | CC BY | 33205256 | 18,657,448 | 2021-07 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Dry gangrene'. | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | ATORVASTATIN, HEPARIN SODIUM, INSULIN NOS, UNSPECIFIED INGREDIENT | DrugsGivenReaction | CC BY | 33205744 | 19,418,771 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Pulmonary embolism'. | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | ATORVASTATIN, HEPARIN SODIUM, INSULIN NOS, UNSPECIFIED INGREDIENT | DrugsGivenReaction | CC BY | 33205744 | 19,418,771 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Pulseless electrical activity'. | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | ATORVASTATIN, HEPARIN SODIUM, INSULIN NOS, UNSPECIFIED INGREDIENT | DrugsGivenReaction | CC BY | 33205744 | 19,418,771 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Treatment failure'. | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | ATORVASTATIN, HEPARIN SODIUM, INSULIN NOS, UNSPECIFIED INGREDIENT | DrugsGivenReaction | CC BY | 33205744 | 19,418,771 | 2021-01 |
What was the administration route of drug 'INSULIN NOS'? | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | Intravenous (not otherwise specified) | DrugAdministrationRoute | CC BY | 33205744 | 19,397,730 | 2021-01 |
What was the outcome of reaction 'Dry gangrene'? | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | Not recovered | ReactionOutcome | CC BY | 33205744 | 19,418,771 | 2021-01 |
What was the outcome of reaction 'Pulmonary embolism'? | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | Fatal | ReactionOutcome | CC BY | 33205744 | 19,418,771 | 2021-01 |
What was the outcome of reaction 'Pulseless electrical activity'? | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | Fatal | ReactionOutcome | CC BY | 33205744 | 19,418,771 | 2021-01 |
What was the outcome of reaction 'Treatment failure'? | Case Report: Multiple Strokes and Digital Ischemia in a Young COVID-19 Patient.
COVID-19 is an infectious disease caused by SARS-CoV-2. This enveloped RNA coronavirus primarily has tropism for the respiratory tract. However, it has also been shown to have various extrapulmonary manifestations such as pulmonary embolism, ischemic strokes, deep venous thrombosis, or arterial thrombosis. We present a case of a 34-year-old woman who had severe COVID-19 infection with no respiratory symptoms and developed strokes in multiple vascular territories and digital ischemia due to thrombosis formation in the brachial circulation of her arm despite receiving therapeutic anticoagulation.
INTRODUCTION
The first case of SARS-CoV-2 was reported in China. Common symptoms include fever, cough, muscle pain, and fatigue, and nearly 80% of patients have normal or decreased white blood cell counts, with many presenting with lymphocytopenia.1 One of the complications of this virus is its induction of hypercoagulability which is not completely understood. From the Virchow’s triad, it can be suspected that a hypercoagulable state results from fluctuation or manipulation of circulating prothrombotic factors such as Lupus anticoagulant (LA) and fibrinogen. Endothelial injury by the virus has also been reported as a trigger for the hypercoagulable state. Our patient had severe COVID-19 infection with strokes in multiple vascular territories and suffered from digital ischemia due to clot formation in brachial circulation of her arm.
CASE REVIEW
We present a case of a 34-year-old Hispanic woman who presented with fatigue, loss of appetite, and malaise in addition to headache and tingling sensation in her fingers. At the time of presentation, she did not have any complaints regarding subjective fevers at home, shortness of breath, chest pain, cough, loss of smell, abdominal pain, nausea, vomiting, or any changes in bowel habits. Her past medical history was notable for insulin-dependent diabetes mellitus type I and hyperlipidemia, with no history of atrial fibrillation or other prothrombotic diseases. Her home medications included both short- and long-acting insulin along with atorvastatin 80 mg. She was not on any oral contraceptive medications. Initial laboratory values and vital signs in the emergency department were pertinent for a heart rate of 121 beats/minute, blood oxygen saturation of 98% on room air, glucose of 515 mg/dL, anion gap of 13, presence of beta-hydroxybutyrate, C-reactive protein of 17.5 mg/L, fibrinogen of 886 mg/dL (213–536 mg/dL), D-dimer of 2.54 mcg/mL (0–0.5 mcg/mL), and ferritin of 170.6 ng/mL (5–148 ng/mL).1–3 A portable chest X-ray (CXR) on admission revealed a normal cardiac silhouette with no signs of focal consolidation or effusion (Figure 1). She was found to have diabetic ketoacidosis, and workup for possible triggers revealed an infectious etiology. Her COVID-19 reverse transcriptase–PCR test (RT-PCR) nasal swab specimen resulted positive. Her CXR did not show pneumonia, and urinalysis was negative for infection. She did not receive any treatment for coronavirus infection such as remdesivir, hydroxychloroquine, glucocorticoids, or plasma because of lack of hypoxemia and symptoms including cough, dyspnea, loss of smell, and other viral symptoms. She was started on intravenous insulin infusion and placed on low molecular weight heparin (LMWH) for deep venous thrombosis prophylaxis regimen during the first few hours of admission. Her mental status continued to decline with decorticate posturing of the left upper extremity. Magnetic resonance imaging of the brain showed acute ischemic strokes in the right middle cerebral artery (MCA) and anterior cerebral artery (ACA) vascular territory with the development of cerebral edema (Figure 2). She did not receive tissue plasminogen activator because of being outside the treatment window. However, the patient’s mental status continued to worsen with medical management, warranting emergent surgical right frontotemporal decompressive hemicraniectomy.
Figure 1. A portable chest X-ray shows a normal cardiac silhouette. There is no focal consolidation or effusion. Costophrenic angles are present. Trachea is midline.
Figure 2. Magnetic resonance imaging (MRI). (A) shows confluent diffusion-weighted imaging hyperintensities of the right frontal temporal lobe, right temporal parietal, and frontal lobe with panel (B) MRI showing corresponding fluid-attenuated inversion recovery hyperintensities.
Multiple strokes in a young female prompted further hypercoagulable workup which revealed positive LA screen with elevated dilute Russell viper venom time (DRVVT) at 63.9 seconds (36.1–50.8 seconds), DRVVT screen/confirm ratio 1.31 (0.97–1.22) with normal levels of beta-2 glycoprotein I Ab IgM, IgG, and IgA. Multiple days after surgery, her course was complicated by the slow development of right-hand swelling, for which the arterial duplex showed radial artery occlusion and monophasic flow in the ulnar artery. Computed tomographic angiography of the right upper extremity showed poor to non-opacification of the radial, ulnar, and palmar arch vessels and irregularities of the brachial arteries, suggesting mural thrombus. Despite therapeutic anticoagulation with unfractionated heparin (UH), her right-hand swelling worsened to dry gangrene of right digits (Figure 3). Initially, there was no acute surgical intervention performed for dry gangrene and eventually was managed with therapeutic dosing of LMWH. She was scheduled for eventual amputation of the ischemic fingers but later suffered a pulseless electrical activity arrest (PEA) and was not able to be resuscitated despite cardiopulmonary resuscitation. We suspect the leading differential for her PEA is pulmonary embolism, given her hypercoagulable state.
Figure 3. Photograph of the volar hand exhibiting dry gangrenous digits from digital ischemia. This figure appears in color at www.ajtmh.org.
DISCUSSION
The most salient feature of this case includes the predominant symptom burden of SARS-CoV-2 infection manifesting as thrombosis in an otherwise healthy young woman with no past medical history of coagulopathy. As per Oxley et al.,2 which reported large-vessel stroke as a presenting feature in the young population, three of the five cases mentioned the involvement of the MCA as a major vascular territory. This is further supported by the data published in the global COVID-19 stroke registry where it reported MCA as the most frequently affected vascular territory.3 In addition to the ischemic stroke presentation, the top presenting complaints were dyspnea/hypoxia.4 Our patient’s stroke involved both the MCA and ACA. Another salient feature is despite having tropism for the lungs, she had no respiratory symptoms as evidenced by imaging and clinical presentation.
As per Merkler et al.4 which compared risk of ischemic stroke in patients with COVID-19 versus patients with influenza, it mentions the association of the virus with vigorous inflammatory response accompanied by coagulopathy, with elevated D-dimer levels and frequent presence of antiphospholipid antibodies. Our patient also had elevated D-dimer levels and LA antibody screenings, indicating a prothrombotic state. We acknowledge that these antibodies can transiently rise in any critical illness, but multiple case reports now have documented that in addition to elevated D-dimer levels, antiphospholipid antibodies have also been elevated in COVID-19 patients.5 However, given the ultimate demise of the patient, repeat serologic confirmation of LA laboratory markers at 12 weeks could not be retested to confirm or refute an underlying antiphospholipid syndrome.
Several autopsy studies showed that COVID-19 affects arterial and venous blood vessels of various sizes and capillary beds but does not symmetrically involve all tissues.6 One of the theories that appear to have been playing a role is the selectivity of coronavirus toward angiotensin-converting enzyme 2 (ACE2) receptors.7 On binding to cell surface ACE2 receptors, it inhibits the vaso-protective functions of ACE, causing pro-inflammatory and prothrombotic states. In addition, the systemic increase of pro-inflammatory cytokines such as interleukin-6 is also believed to be a major contributor of inducing hypercoagulable state in COVID-19. Furthermore, COVID-19 compromises the integrity of endothelial monolayer by causing endothelial cell death through its lytic replication,7 thus exposing the thrombogenic basement membrane and in turn leading to the activation of coagulation cascade.7
Acknowledging that SARS-CoV-2 leads to a hypercoagulable state, the role of anticoagulation is important. As per Klok et al.,8 the incidence of thrombotic complications in critically ill patients in intensive care unit is 31%. The American Society of Hematology recommends LMWH as venous thromboembolism prophylaxis agent over UH. Literature has varied in terms of dosing, selection, and timing of therapeutic agents. Some recommend using therapeutic LMWH if there is a rise of the inflammatory markers and D-dimer on days 7–14.9 Others recommend the use of scoring calculators such as sepsis-induced coagulopathy published by the International Society of Thrombosis and Hemostasis, which uses platelet count, international normalized ratio, and sequential organ failure assessment scores to risk-stratify and guide anticoagulation strategies.10 Our patient received both prophylactic and therapeutic doses of anticoagulation during the hospital course and developed digital ischemia despite receiving therapeutic anticoagulation.
In conclusion, we report a COVID-19 patient who did not develop pneumonia and had no history of hypercoagulable condition developing multiple arterial thrombosis involving the neurovascular and peripheral vascular system with both leading to grave sequelae, despite receiving anticoagulation. Our case reveals that more research needs to be undertaken to understand the cascade of events leading to a prothrombotic state and the role of anticoagulation regimen for prevention and treatment of these highly susceptible thrombotic events.
Acknowledgments:
We would like to thank the wonderful staff who shared with us their insight about this particular case. Publication charges for this article were waived due to the ongoing pandemic of COVID-19. | Fatal | ReactionOutcome | CC BY | 33205744 | 19,418,771 | 2021-01 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Dizziness'. | Vernakalant for Rapid Cardioversion of Recent-Onset Atrial Fibrillation: Results from the SPECTRUM Study.
Rapid restoration of sinus rhythm using pharmacological cardioversion is commonly indicated in patients with symptomatic recent-onset atrial fibrillation (AF). The objectives of this large, international, multicenter observational study were to determine the safety and effectiveness of intravenous (IV) vernakalant for conversion of AF to sinus rhythm in daily practice.
Consenting patients with symptomatic recent-onset AF (< 7 days) treated with IV vernakalant were enrolled and followed up to 24 h after the last infusion or until discharge, in order to determine the incidence of predefined serious adverse events (SAEs) and other observed SAEs and evaluate the conversion rate within the first 90 min. Overall, 2009 treatment episodes in 1778 patients were analyzed. The age of patients was 62.3 ± 13.0 years (mean ± standard deviation). Median AF duration before treatment was 11.1 h (IQR 5.4-27.0 h). A total of 28 SAEs occurred in 26 patients including 19 predefined SAEs, i.e., sinus arrest (n = 4, 0.2%), significant bradycardia (n = 11, 0.5%), significant hypotension (n = 2, 0.1%), and atrial flutter with 1:1 conduction (n = 2, 0.1%). There were no cases of sustained ventricular arrhythmias or deaths. All patients who experienced SAEs recovered fully (n = 25) or with sequelae (n = 1). Conversion rate to sinus rhythm was 70.2%, within a median of 12 min (IQR 8.0-28.0 min).
This large multicenter, international observational study confirms the good safety profile and the high effectiveness of vernakalant for the rapid cardioversion of recent-onset AF in daily hospital practice.
Introduction and Purpose of the Study
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, with an estimated 33.5 million people affected worldwide [1]. One in four adults over 55 years of age in Europe and the USA develop AF, with greater prevalence in older populations [1, 2]. Patients with AF are at increased risk of stroke and heart failure [3, 4]. A significant number of patients with recent-onset AF seen in the emergency departments (EDs) undergo commonly in Europe pharmacological cardioversion.
Vernakalant is a partial atrial-selective antiarrhythmic agent by its action through IKur and IKACh channel inhibition [5]. However, it has a modest effect on the ventricle via Ina and IKr channels resulting in a limited effect on ventricular repolarization (QT interval) [5]. Vernakalant is contra-indicated in patients with prolonged QT interval.
Intravenous vernakalant has been approved by the European Medicine Agency [2010] for the rapid conversion of recent-onset AF [6]. To date, a number of studies have shown vernakalant to be well tolerated and effective for cardioversion of AF [7–18].
The FDA (Food and Drug Administration agency) decided in 2008 and in December 2019 not to approve to market vernakalant in the USA for safety concerns. In 2010, the EMA requested a post-authorization safety study to better define the risk benefit ratio in routine clinical practice. The objectives of SPECTRUM (Surveillance of Pharmacologic thErapy for Cardioversion in aTrial fibrillation Registry Using IV treatMent) (NCT01370629 and EUPAS2078) study were to assess the rates of adverse events and to estimate the effectiveness of the drug in a large cohort of patients with recent-onset AF.
Methods
Definitions
Recent-onset AF was defined as symptomatic episode within 7 days that will be undergoing cardioversion taking into account that about 70% of patients with symptomatic AF < 72 h were reported to convert spontaneously [19]. Beyond 7 days, AF is likely to persist and the chances of pharmacological cardioversion to be successful become low. Hypertension was reported when documented on the medical record or the patient report. Coronary artery disease (CAD) was diagnosed when the patient had a documented history of CAD and/or a history of coronary revascularization.
Patients and Procedures
Adult patients (≥ 18 years) with recent-onset AF occurring between September 1, 2011 and April 11, 2018 who received vernakalant for cardioversion were eligible for inclusion in this international, multicenter, observational, post-authorization study. Fifty-five hospitals in Austria, Denmark, Germany, Spain, Sweden, and Finland participated in the study, 53 of which enrolled patients. While administration of vernakalant was at the discretion of the treating physician, consecutively treated patients were enrolled and reasons for non-participation were documented. A preinfusion checklist and healthcare provider educational card were implemented during the study period to assist in identifying patients for treatment consistent with the approved indications and contraindications.
Patients were required to give informed consent for participation in the study and could be enrolled more than once if they presented on multiple occasions for AF episodes. Patients who had participated in an investigational drug/device clinical trial within 30 days prior to enrollment were not eligible. In order to enhance enrollment and reach the EMA required target of 2000 episodes, a protocol amendment was made in September 2016, which permitted retrospective inclusion of patients who had received vernakalant between April 2013 and the end of the study, provided that they fulfilled the established eligibility criteria. For prospectively enrolled patients, data were collected from both medical records and supplemental standardized data collection forms. For retrospectively enrolled patients, only medical records were available. The study period comprised a baseline assessment and up to 24-h follow-up after completion of the last infusion or until discharge. This study was mandated and approved by the European Committee for Medicinal Products for Human Use. The study protocol was approved by the appropriate local research ethics committees for all participating centers, and the study was conducted in accordance with applicable national and local regulations/guidelines, accepted standards for Good Clinical Practice, Guidelines for Good Pharmacoepidemiology Practices, and the Declaration of Helsinki [20].
Study Objectives and Endpoints
The primary objectives of the study was to estimate the incidence of clinically predefined serious adverse events (SAEs), i.e., significant hypotension (systolic blood pressure < 90 mmHg or requiring vasopressors); sustained (> 30 s) ventricular arrhythmias, Torsade de Pointes (>10 s) or ventricular fibrillation, atrial flutter with 1:1 conduction, bradycardia requiring temporary electrical pacing, or sinus arrest (> 3 s). Definition of these predefined SAEs was based on events from previous controlled studies on IV vernakalant [7, 8, 11, 12] and from the reported adverse events (AEs) on other antiarrhythmic agents. Secondary objectives included the rates of all other SAEs. Each SAE was reviewed and adjudicated by an independent expert Safety Review Committee (SRC). This study had also the objective to determine the conversion rate to sinus rhythm in a large population of patients outside the setting of controlled clinical trials.
The duration of the index AF episode was calculated as the time between the patient-reported time of symptom onset and the start of the first vernakalant infusion. Successful cardioversion was defined as conversion to sinus rhythm within 90 min of the start of vernakalant infusion. Conversion rate was calculated in all patients, as well as in an effectiveness population excluding all treatment episodes in which patients received another therapy for cardioversion within 90 min of the start of vernakalant administration (e.g., electrical or pharmacological cardioversion). Vernakalant is recommended to be administered in a step-dose fashion. Each treatment episode can comprise up to two infusions, separated by a 15-min observation period. The recommended doses for the first and second infusions are 3.0 mg/kg and 2.0 mg/kg, respectively, each administered over 10 min. For patients above 113 kg, vernakalant has a fixed initial dose of 339 mg. If conversion to sinus rhythm does not occur within 15 min after the end of the initial infusion, a second 10-min infusion of 226 mg may be administered.
Statistics and Analyses
A target sample size of 2000 vernakalant IV treatment episodes was chosen to allow adequate statistical precision, as expressed by a two-sided 95% confidence limit. Enrollment per site was capped at 10% of the total study population and 40% per country to minimize any potential bias in practice patterns. Categorical variable frequency, along with 95% confidence intervals (CIs), was determined for the summed treatment episodes. Continuous variables were summarized using descriptive statistics. Data were analyzed based on enrollment method (prospective vs retrospective) and reported as stratified and unstratified CIs. All analyses were performed using Statistical Analysis System v9.2, or later, software.
Results
Study Population
A total of 1778 patients who presented with 2009 treatment episodes were included: 1580 episodes were in prospectively enrolled patients and 429 in retrospectively enrolled patients (Table 1). The majority of patients were treated in the ED for 1289 (64.1%) AF episodes and 563 (28.0%) AF episodes in the coronary or intensive care units, with the remainder 157 (7.8%) episodes being treated in other hospital settings. As seen in Fig. 1, the main reason for non-inclusion in the study was lack of informed consent. In 1905 (94.7%) AF episodes, vernakalant was administered to non-surgery patients, and in 104 (5.2%) to post-cardiac surgery patients. The later are among the prospectively included patients. The mean age of the overall patient population at time of treatment was 62.3 ± 13.0 years (mean ± standard deviation [SD]), ranging from 18.0 to 94.0 years, and 1222 (60.8%) episodes occurred in men (Table 1). At baseline, systolic blood pressure (BP) was 132.5 ± 19.5 mmHg and heart rate (HR) was 112.9 ± 25.5/min (mean ± SD). The median duration of AF episode prior to treatment was 11.1 (5.4–27.0) hours (median [interquartile range, IQR]). In 88.9% of episodes, the patients were treated within 48 h of the onset of symptoms, and in 72.5% within 24 h. Duration of AF before treatment in 104 post-cardiac surgery patients was shorter than in the overall population, with 3.6 h (range 0.8–15.4) (median [IQR]). Baseline demographics and characteristics were similar between patients enrolled prospectively and retrospectively. Total length of ED stay was 7.5 (5.0–13.5) hours (median [IQR]). Only 167 (13.0%) of patients initially managed in the ED were in hospital for 24 h or longer. The number of vernakalant infusions was available in 1990 patients. Of these, 1201 (60.4%) received one vernakalant infusion and 789 (39.6) received a total of 2 infusions.Table 1 Clinical characteristics of patients
Total Prospective Retrospective
No. of patients 2009 1580 429
Age (years) mean ± SD 62.3 ± 13.0 61.9 ± 13.5 63.6 ± 11.2
Range (years) 18.0–94 18–93 30–94
Male, n (%) 1222 (60.8) 998 (63.2) 224 (52.2)
Body weight (kg) mean ± SD 84.1 ± 16.5 84.3 ± 16.5 83.1 (16.9)
Range (kg) 45.0–189.0 45.0–189.0 45.0–165.0
Body mass index (kg/m2) 27.8 ± 4.9 27.7 ± 4.8 28.2 ± 5.1
Associated conditions, n (%)
Hypertension 1103 (54.9) 884 (55.9) 219 (51.0)
Coronary artery disease 118 (5.9) 82 (5.2) 36 (8.4)
Cardiomyopathy 33 (1.6) 31 (2.0) 2 (0.5%)
Heart failure (history) 63 (3.1) 59 (3.7) 4 (0.9)
Diabetes 199 (9.9) 165 (10.4) 34 (7.9)
Stroke (history) 91 (4.5) 68 (4.3) 23 (5.4)
Pacemaker/ICD 36 (1.8) 24 (1.5) 12 (2.8)
Type of AF episode
First detected 477 (23.7) 393 (24.9) 84 (19.6)
Previous history of AF 1458 (72.6) 1115 (70.6) 343 (80.0)
Onset unknown/not assessed 5 (0.2) 3 (0.2) 2 (0.5)
Post-surgery 69 (3.4) 69 (4.4) 0 (0.0)
Symptoms on admission, n (%)
Palpitations, irregular heart beat 1749 (87.1) 1337 (84.6) 412 (96.0)
Dyspnea or shortness of breath 352 (17.5) 306 (19.4) 46 (10.7)
Dizziness, light-headedness 320 (15.9) 251 (15.9) 69 (16.1)
Chest pain 271 (13.5) 220 (13.9) 51 (11.9)
Syncope, near syncope 61 (3.0) 53 (3.4) 8 (1.9)
Duration of the index episode
Less than 24 h, n (%) 1438 (72.5) 1107 (70.2) 331 (81.5)
24–48 h, n (%) 347 (17.5) 288 (18.3) 59 (14.5)
More than 48 h 199 (10.0) 183 (11.6) 16 (3.9)
Mean duration ± SD (h) 23.2 ± 44.9 24.9 ± 45.8 16.8 ± 40.6
Median (IQR 25–75) (h) 11.1 (5.44–27.03) 11.9 (5.8–29.7) 8.2 (4.8–18.3)
Antiarrhythmic agents, n (%)
Betablockers 1055 (52.5) 800 (50.6) 255 (59.4)
Calcium channels blockers 22 (1.1) 20 (1.3) 2 (0.5)
Class I agents* 85 (4.2) 71 (4.5) 14 (3.3)
Class III agents* 98 (4.9) 89 (5.6) 9 (2.1)
Digitalis glycosides 22 (1.1) 18 (1.1) 4 (0.9)
*Using the Vaughan-Williams classification
Fig. 1 Study flow chart. Flow chart showing patient enrollment in the SPECTRUM study. The term patient here refers to individual treatment episodes (asterisk). Owing to lack of informed consent (n = 500) (dagger). Other reasons included patient enrollment in an investigational drug trial in the past 30 days, spontaneous conversion to sinus rhythm, ejection fraction 30–35%, electrical cardioversion preferred, missing information regarding start of atrial fibrillation, inclusion criteria not met, other, or no reason provided or known. Source data could not be verified to confirm that vernakalant IV was administered (double dagger). Spontaneous conversion to sinus rhythm before vernakalant IV administration (section sign). Patient decision and lack of follow-up after cardioversion in one case each (double vertical line). IV intravenous
Predefined Serious Adverse Events and Other Adverse Events
No deaths were recorded in our study. Nineteen predefined SAEs were reported during or after 17 treatment episodes (cumulative incidence 0.8%; CI 0.5–1.4%) (Table 2). Eighteen of the 19 events occurred within 2 h from the start of infusion. The remaining event was an episode of atrial flutter with 1:1 conduction which occurred 3.1 h after drug infusion and was terminated by electrical shock. Symptomatic bradycardia was the most common event occurring in 11 (0.5%; CI 0.4–1.2%) episodes (Table 2). Conversion to sinus rhythm occurred in 10 of these cases. A pause described as sinus arrest preceding the restoration of sinus rhythm occurred in 4 patients. In 2 patients, sinus arrest was associated with sinus bradycardia. In all bradycardia and sinus arrest cases, the vernakalant infusion was immediately discontinued. One of these 4 sinus arrests occurred in a 66-year-old man, sportive cyclist with no history of heart disease, admitted for a first episode of AF with a mean ventricular response of 95 beats/min. He received 300 mg orally of flecainide which failed to restore sinus rhythm. The treating physician decided 4 h later, to administer IV vernakalant. At the end of the infusion, a pause of 6 s, with a brief dizziness, occurred and resolved spontaneously, followed by a normal sinus rhythm with a HR of 47 beats/min which was patient usual HR and a BP of 120/85 mmHg. This event was considered a SAE although there was probably an interaction between oral flecainide still active and vernakalant in this event. One of the bradycardia events occurred in a retrospectively enrolled 69-year-old woman on bisoprolol with a history of hypertension and CAD, who developed 8 min after the second infusion of vernakalant a sinus bradycardia which rapidly resolved with IV atropine. Two bradycardia episodes occurred in post-cardiac surgery patients requiring temporary electrical pacing through the electrodes left in place by the surgeon. Both patients converted to sinus rhythm. None of the non-surgery patients required temporary electrical pacing. Significant hypotension occurred on two (0.1%; CI < 0.1–0.4%) occasions, associated with sinus bradycardia in both instances. Both events resolved with intravenous atropine and fluid. There were two cases of atrial flutter with 1:1 ventricular conduction terminated with electrical shock whereas no cases of sustained ventricular tachycardia (VT), ventricular fibrillation, or Torsade de Pointes were observed. In addition to the predefined SAEs, there were 9 other SAEs, one of which occurred in a retrospectively enrolled patient (Table 2). They included two instances of hypotension not requiring vasopressor agents, 2 non-sustained VT which deserve special attention. The first non-sustained VT occurred in a 48-year-old man with asthma admitted with fever, palpitations, dyspnea, and first episode of AF with a ventricular rate of 144 bpm. During vernakalant infusion, 5 beats of non-sustained VT was observed. Among the tests done, coronary angiography was reported as normal. The same run of 5 beats of non-sustained VT was observed 20 h after infusion (next day) making the causal effect of vernakalant unlikely. The other event occurred in a 57-year-old patient with a 6-year history of recurrent symptomatic AF and arterial hypertension with left ventricular hypertrophy. He was admitted with palpitations, irregular heartbeats, and dizziness. He was on dronedarone, and ECG showed AF with a ventricular rate of 135 bpm. During infusion of vernakalant, he had 6 s of non-sustained VT observed on the monitor and was given 5 mg of bisoprolol which reduced the heart rate to 120 beats/min and relieved patient symptoms. The Safety Review Committee considered that in the first case, the wide QRS complexes were due to aberrant conduction during rapid AF (Ashman phenomenon). Among the non-predefined SAEs, one supraventricular tachycardia (120 beats/min) and a single report each of angina pectoris, pericardial effusion, transient visual disturbance, and vernakalant overdose (Table 2). A total of 188 non-serious AEs were reported, the most common of which were dysgeusia (n = 35) and sneezing (n = 27). All patients with vernakalant-related AEs recovered without sequelae. All but 6 of the 28 SAEs were considered by the investigators and the SRC to be related to vernakalant administration.Table 2 Adverse events in 2009 episodes during treatment and observation periods
Event type Number of events Incidence (95% CI) Considered drug-related, n (%)
All SAEs 28 1.3% (0.8–1.9) 22 (78.6)
Predefined SAEs 19 0.8% (0.5–1.4) 18 (94.7)
Significant hypotension 2 0.1% (< 0.1–0.4) 2 (100.0)
Bradycardiaα 11 0.5% (0.3–10) 10 (93.3)
Sinus arrest (> 3 s)β 4 0.2% (< 0.1–0.4) 4 (100.0)
Atrial flutter with 1: 1 AV conduction 2 0.1% (0.1–0.4) 2 (100.0)
Ventricular tachycardia γ 0 0 0 (0.0)
Other than predefined SAEs 9 0.45% 5 (55.6)
Hypotension 2 0.1% 1 (50.0)
Supraventricular tachycardiaδ 1 < 0.1% 1 (100.0)
Non-sustained ventricular tachycardiaε 2 < 0.1% 1 (50.0)
Angina pectoris 1 (< 0.1) < 0.1% 0 (0.0)
Pericardial effusion 1 (< 0.1) < 0.1% 0 (0.0)
Visual disturbance 1 (< 0.1) < 0.1% 0 (0.0)
Vernakalant overdoseζ 1 (< 0.1) < 0.1% 1 (100.0)
αNine cases of sinus bradycardia and 2 reported as significant bradycardia
βOne patient had both sinus arrest followed by sinus bradycardia
γOne event reclassified as atrial flutter with 1:1 conduction
δAtrial arrhythmia other than atrial flutter
εSee text, exceeding 5% of the weight-based dosing recommendation. In this case, the administered dose was 51% in excess of the recommended dose
Rates of Conversion to Sinus Rhythm
Overall, conversion to sinus rhythm at any time following vernakalant infusion occurred in 1448 out of 2009 (72.1%) treatment episodes. Successful cardioversion was recorded in 70.2% (CI 68.1–72.2%) of the 1936 episodes of the effectiveness population excluding those in which either electrical cardioversion (n = 68) or an additional intravenous Class I/III antiarrhythmic drug (n = 6) was given within 90 min of infusion initiation. The rate of cardioversion was similar between the 1107 of 1580 (70.1%) episodes included prospectively and the 297 of 421 (70.5%) episodes of retrospectively enrolled patients. Successful cardioversion of AF was reported in 68 of 104 (65.4%) of treatment episodes in the post-cardiac surgery patients. Time to cardioversion was recorded in 1413 of 1448 episodes with successful conversion to sinus rhythm. The median time to conversion was 12.0 (8.0–28.0) minutes (median [IQR]) Fig. 2). One thousand one hundred eight of 1413 (78.4%) successful cardioversions were treated with only one drug infusion. The percentage of successful cardioversion was 70.1% in the prospective patients and 70.5% in the retrospective patients. The median hospital stay time in those treated in the ED was 7.5 h allowing patient discharges when their condition was clinically stable.Fig. 2 Time to conversion to sinus rhythm. Time to conversion to sinus rhythm with vernakalant IV in the effectiveness analysis population (N = 1936). Time to conversion was not recorded in 29 treatment episodes in which patients converted to sinus rhythm; these episodes are not displayed on the graph but are taken into account for the proportion calculation. IV intravenous
Anticoagulation
About a quarter of patients presenting with recent-onset AF at baseline were on vitamin K antagonists or direct oral anticoagulants. Investigators respected current guidelines [3] on anticoagulation both peri-procedurally and after hospital discharge.
Discussion
The SPECTRUM study included a large real-world patient population of 1778 patients with 2009 recent-onset AF episodes in whom pharmacological cardioversion was performed with vernakalant. About 70% of patients were cardioverted within 12 min from onset of infusion and 11 h from the AF onset. Our findings confirm the safety and efficacy of vernakalant reported in previous studies [7–18, 21–25] and extend their consistency to routine hospital use in large populations. To our knowledge, the present study provides the largest series of patients with recent-onset AF undergoing pharmacological cardioversion with a specific antiarrhythmic agent. The safety was the main objective of this study. We found the incidence of both predefined and other SAEs to be lower than expected. There were no death and no sustained ventricular arrhythmia. Overall, 28 SAEs (1.3%) were recorded. The majority of patients were AF treated in ED and intensive care units.
Pharmacological cardioversion is frequently indicated as part of a rhythm control strategy or as a tool to control patient symptoms and avoid hospitalization in clinically stable condition [25, 26]. It is often preferred to electrical cardioversion in patients with hemodynamically stable condition as it does not require general anesthesia or sedation. Among agents currently available for rapid termination of recent-onset AF, vernakalant represents an option [3]. However, there has been to our knowledge, no large study exploring the safety of vernakalant in daily practice.
There is no universal definition for recent-onset AF. In current literature, the duration limits of AF episodes range from < 24 [27] to < 48 h and even < 7 days [28, 29]. The prevalence of recent-onset AF among all AF subsets varies from 11% when restricted to the first detected episode (new onset) [30] to 26% [31]. The characteristics of patients were similar to those of other AF cohorts [31, 32].
As with electrical cardioversion, pharmacological cardioversion can be associated with post-cardioversion bradyarrhythmias, often unmasking pre-existing sinus node dysfunction or atrioventricular conduction abnormalities and can result in ventricular escape rhythms or prolonged ventricular pauses. Of interest, these pauses were first reported by Lown [33], following electrical cardioversion as the possible reflect of sinus dysfunction. Another possible mechanism for these sinus arrests is right atrial stunning [34]. The cumulative incidences of bradycardia (0.5%), sinus arrest (0.2%), and hypotension (0.1%) observed in this study were also low. The incidence of atrial flutter with 1:1 conduction was lower than that reported with oral Class Ic antiarrhythmics, such as flecainide or propafenone. The “pill in the pocket” approach requires initiation of therapy in hospital to verify its safety [35]. No cases of Torsade de Pointes or sustained VT was observed, which is in line with the low risk of ventricular proarrhythmia associated with vernakalant owing to its electrophysiological properties [5]. This contrasts with the reported incidence of Torsade de Pointes [24, 36, 37] in patients with AF/atrial flutter of 4.3% with intravenous ibutilide in the report of Kowey et al. including 1.7% of which required cardioversion [36]. Of note, all but one of the predefined SAEs in this study occurred within 2 h of the start of infusion. As aforementioned, the remaining patient had atrial flutter with 1:1 ventricular conduction which occurred 3.1 h following infusion initiation, indicating that close cardiac monitoring should be available during and after treatment in some patients.
Conversion to sinus rhythm with vernakalant was rapid (median time of 12.0 min) similar to what was previously reported [9–18]. The conversion median time of ibutilide was significantly longer than that of vernakalant (26 min versus 10 min, P = 0.01) in a randomized comparison [18]. Furthermore, in this particularly large real-world study, the median duration of AF episode was short (11.1 h) as there is important evidence, and relevant guidelines [3] suggesting that prompt cardioversion could be associated with benefits in terms of lower risk of thromboembolic events [4, 38, 39].
Although the baseline characteristics of the study population were consistent with AF population-based studies [31, 32] and clinical studies with vernakalant, the conversion rate was higher than that observed in recent review and meta-analysis (~ 50%) [22–25]. This seems likely to be due to patients being treated soon after symptom onset in European clinical practice. Other recent but smaller observational studies [13–15, 17, 18], which collectively included almost 1300 patients, have found similarly high conversion rates (65–86%) when vernakalant was administered soon after the onset of AF, particularly within the first 48 h [15, 17, 21]. Vernakalant has also been shown to induce a higher rate of cardioversion compared with flecainide (67% vs 46%) in a non-randomized cohort study [21]. Similarly, in randomized studies, vernakalant was more effective than amiodarone [12] (52% vs 5%; after 90 min) and ibutilide [18, 24] (69% vs 43%; within 90 min). The SPECTRUM results are consistent with previous reports that vernakalant is safe and effective for the rapid cardioversion of recent-onset AF and extends them to daily practice.
Owing to the rapid time to conversion with vernakalant, the median hospital stay time for those treated in the ED was 7.5 h. This is encouraging given that a study in France reported that hospitalization constitutes 60% of the cost of care for patients with AF [40].
Study Limitations
This multicenter international study was observational as the main objective was to determine the safety of vernakalant as used in daily hospital practice without interfering on the management of recent-onset AF by the treating physician. For these reasons, the adverse events were expected to be higher in an “uncontrolled” setting with no guidance on patient selection than those reported in controlled studies with strict protocols. In fact, SAEs were low in this study. Data collection for prospectively enrolled patients was comprehensive owing to use of both study-specific tools and medical records. However, for retrospectively enrolled patients, it was not possible to routinely collect all data of interest in a standardized manner. Nevertheless, baseline characteristics, medical histories, and SAEs in the retrospective cohort were similar to those in the prospective cohort, supporting the use of a retrospective analysis.
Conclusions
The results of this large multicenter study showed that vernakalant has a good safety profile and is effective in enabling rapid cardioversion in clinical practice. Moreover, the rates of serious complications were lower than those observed in early trials reflecting appropriate patient selection in clinical practice. In conclusion, vernakalant provides a rapid and effective means of pharmacological conversion in patients with recent-onset AF undergoing cardioversion undergoing cardioversion in daily hospital practice.
Appendix. List of Investigators
Austria: Michael Joannidis, Klemens Zotter Frank Hartig, Anton Sandhofer, Alois Süssenbacher, Bernd Eber, Elisabeth Lassnig, Ulrike Pfeifenberger, Michaela Steiner, Hans Domanovits, Alexander Simon, Alexander Spiel, Jan Niederdockl, Nikola Schuetz, Daniel Wehinger, Franz Xaver Roithinger, Isabella M. von Katzler, Katharina Bichler, Robert Schoenbauer, Lukas Fiedler, Michael Pfeffer, Markus Peck, Florian Benische, Michael Hackl, Susane Demschar, Astrid Ebner, Melanie Eder, Rainer Huditz, Arnulf Isak, Michael Moser, Georg Pinter, Thomas Singer, Claudia Waldhauser, Helmut Pürerfellner, Martin Martinek, Sandra Muellner, Andrea Ploechl, Tanja Koppler, Elisabeth Sigmund, Michael Derndorfer, Sabine Metz, Karin Streicher, Clemens Steinwender, Karim Saleh, Andreas Lueger, Petra Fladerer, Eiko Meister, Heinz Drexel, Alexandra Schuler, Susanne Waeger, Karl-Martin Ebner, Christine Heinzle, Arthur Mader, Peter Schwerzler, Berta Patsch, Abdurahman Said, Claudia Stoeckloecker, Daniela Zanolin, Jutta Bergler-Klein, Ljubica Mandic, Mariann Gyöngyösi, Neraida Cene, Zsuzsanna Szankai, Abelina Zimba.
Denmark: Henrik Nielsen, Bjerre Flemming, Michaelsen Michaelsen, Elisa Stokholm, Katja Holm, Charlotte Schmidt Skov, Pauline Gøgsig Johansen, Soren shjortshoj, Thomas Melchior, Ole Dyg Pedersen, Sanne Heinsvig, Inge Larsen, Vibeke Perret-Gentil, Thomas Wagner Nielsen, Axel Brandes, Marianne Jensen, Ida Rosenlund, Liv Gøtzsche, Heidi Munk Andersen.
Germany: Andreas Götte, Matthias Hammwohner, Britta Möehring, Jutta Schaertl, Daniel Steven, Iris Berg, Alexandra Kuehn, Hannes Reuter, Elena Terentieva, Christian Loges, Christine Lindner, Hendrik Bonnemeier, Christan Wulff, Thomas Demming, Svenja Gediehn, Johanna Parlitz, Wilhelm Haverkamp, Buehner Kathrin, Hubert Katja, Iacovella Ines, Bernhard Korbmacher, Marc Thone, Hannan Dalyanoglu, Da Un Chung, Naujoks Angela, Dirk Weismann, Björn Lengenfelder, Jan Becher, Klaus Meyer, Irina Turkin, Sebastian Maier, Marcus Koller, Alban Glaser, Lisa Gebele, Jale Goezuebueyuek, Ralph Hampe, Barbara Ruemmler, Hagen Schrötter, Manja Hubald, Cornelia Fritz, Martin Domhardt, Kathrin Haacke, Nicole Schmiedehausen, Ruth Strasser, Kristof Graf, Lidia Fischer, Roland Thieme, Karlheinz Seidl, Martin Kulzer, Monika Zackel, Gerian Grönefeld, Christina Paitazoglou, Simone Müller, ThoraBotschafter Britta Goldmann, Andrea Moeller, Sindy Bartel, Joern Schmitt, Damir Erkapic, Gabriele Hellwig-Bahavar, Ritvan Chasan, Christopher Gemein, Victoria Johnson, Christiane Kelm, Kay Weipert, Johannes Brachmann, Michael Held, Andrea Höhn, Ute Goebel, Andrea Linss, Swetlana Rube, Ahmed Saleh, Steffen Schnupp, Yeong-Hoon Choi, Vera Wolf, Andrea Plate, Anton Sabashnikov, Antje-Christin Deppe, Petra Krause.
Spain: Ignacio Fernandez Lozano, Manuel Sánchez, Francisco Hernández, Alfonso Martin Martinez, Pedro Vazquez, Esther Alvarez, Pascual Lopez, Raquel Torres, José Carbajosa Dalmau, Laura Parades, Nestor Hernandez, Inmaculada Jimenez Ruiz, Ana Maria Lopez, Luis López-Andujar, Alexandre Noguera, Francisco Roman Cerdan, Carmen del Arco Galan, Daniel Afonso, Raquel Caminero, Manual Lunquera, Monica Negro, Cristina Santiago, Nestor Villalba, José Manuel Garrido Castilla, Roberto Martinez Asenjo, Elena Mejia Martinez, José Luís Merino Llorens, Maria Jesus Diaz-Pintado, Jorge Alejandro Figueroa, Alberto Borobia Perez, Sergio Castrejon Castrejon, David Filgueiras Rama, Manuel Quintana Diaz, Maria Angelica Ribera Nunez, Miguel Angel Ramirez Marrero, Antonio Martin, José Miguel Ormaetxe Merodio, Mercedes Varona Peinador, Maria Fe Arcocha, Larraitz Gaztañaga, F. Xavier Palom Rico, Javier Jacob Rodriguez, Pascual Piñera Salmeron, Juan Cosin Sales, Isabel Navarro, Francisco Buendia Funetes.
Sweden: Henrik Wallentin, Kerstin Roos, Arash Mokhtari, Hans-Jörgen Nilsson, Peter Vasko, Terese Nyström, Martina Gustensson, Susanne Johansson, Inga Uggeldahl, Göran Andersson, Olle Bergström, Thomas Aronsson, Mehmet Hamid, Kerstin Giocondi, Deborah Svanerö, Qassim Awad, Tord Juhlin, Hjördis Jernhed, Stefan Berglund, Magnus Forsgren, Michael Guggi, Pär-Lennart Agren, Kristina Eriksson, Kristina Karlsson, Per Blomström, Alejandro Utreras, Caroline Lundgren, Maria Soderlund, Frederik Buijs, Solveig Östberg, Johan-Emil Bager, Ingrid Hendequist, Maria Just, Siv Heden, Liselott Lisjo, Chrichan Mansson, Helen Svanstrom, Mikael Dellborg, Helena Dellborg, Gorel Hultsberg Olsson, Linus Hansson.
Finland: Hannu Sulonen, Hanna Suurmunne, Anna Petrovskaja, Johanna Markkanen, Juha Hartikainen, Lari Kujanen, Antti Heikkola, Hanna Pohjantahti MaarooS.
Safety Review Committee
Lars Kober, MD, Samuel Lévy, MD (Chair), Cristina Varas-Lorenzo, MD, MSc, PhD, Manel Pladevall-Vila MD, MS.
The authors wish to thank Nathalie Dunkel and her team for their help in accessing the data of SPECTRUM and providing information and documents upon our request.
The authors would like to thank all of the investigators involved in the SPECTRUM study.
Funding
This study was funded by Correvio International Sàrl, Geneva, Switzerland. The authors received no funding for their participation in this manuscript.
Data Availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Compliance with Ethical Standards
Conflict of Interest
Professor Levy reports no conflicts of interest. Professor Hartikainen has been an investigator in studies sponsored by AstraZeneca, Biosense Webster, Boehringer Ingelheim, Correvio International Sàrl, Medtronic, and St. Jude Medical. Dr. Ritz is an employee of Correvio International Sàrl, Geneva, Switzerland. Dr. Juhlin has received speaker honoraria from Correvio International Sàrl. Professor Domanovits reports no conflicts of interest. Dr. Carbajosa-Dalmau has received speaker honoraria from Correvio International Sàrl.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. | FLECAINIDE, VERNAKALANT | DrugsGivenReaction | CC BY | 33206300 | 19,673,350 | 2021-04 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Drug ineffective'. | Vernakalant for Rapid Cardioversion of Recent-Onset Atrial Fibrillation: Results from the SPECTRUM Study.
Rapid restoration of sinus rhythm using pharmacological cardioversion is commonly indicated in patients with symptomatic recent-onset atrial fibrillation (AF). The objectives of this large, international, multicenter observational study were to determine the safety and effectiveness of intravenous (IV) vernakalant for conversion of AF to sinus rhythm in daily practice.
Consenting patients with symptomatic recent-onset AF (< 7 days) treated with IV vernakalant were enrolled and followed up to 24 h after the last infusion or until discharge, in order to determine the incidence of predefined serious adverse events (SAEs) and other observed SAEs and evaluate the conversion rate within the first 90 min. Overall, 2009 treatment episodes in 1778 patients were analyzed. The age of patients was 62.3 ± 13.0 years (mean ± standard deviation). Median AF duration before treatment was 11.1 h (IQR 5.4-27.0 h). A total of 28 SAEs occurred in 26 patients including 19 predefined SAEs, i.e., sinus arrest (n = 4, 0.2%), significant bradycardia (n = 11, 0.5%), significant hypotension (n = 2, 0.1%), and atrial flutter with 1:1 conduction (n = 2, 0.1%). There were no cases of sustained ventricular arrhythmias or deaths. All patients who experienced SAEs recovered fully (n = 25) or with sequelae (n = 1). Conversion rate to sinus rhythm was 70.2%, within a median of 12 min (IQR 8.0-28.0 min).
This large multicenter, international observational study confirms the good safety profile and the high effectiveness of vernakalant for the rapid cardioversion of recent-onset AF in daily hospital practice.
Introduction and Purpose of the Study
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, with an estimated 33.5 million people affected worldwide [1]. One in four adults over 55 years of age in Europe and the USA develop AF, with greater prevalence in older populations [1, 2]. Patients with AF are at increased risk of stroke and heart failure [3, 4]. A significant number of patients with recent-onset AF seen in the emergency departments (EDs) undergo commonly in Europe pharmacological cardioversion.
Vernakalant is a partial atrial-selective antiarrhythmic agent by its action through IKur and IKACh channel inhibition [5]. However, it has a modest effect on the ventricle via Ina and IKr channels resulting in a limited effect on ventricular repolarization (QT interval) [5]. Vernakalant is contra-indicated in patients with prolonged QT interval.
Intravenous vernakalant has been approved by the European Medicine Agency [2010] for the rapid conversion of recent-onset AF [6]. To date, a number of studies have shown vernakalant to be well tolerated and effective for cardioversion of AF [7–18].
The FDA (Food and Drug Administration agency) decided in 2008 and in December 2019 not to approve to market vernakalant in the USA for safety concerns. In 2010, the EMA requested a post-authorization safety study to better define the risk benefit ratio in routine clinical practice. The objectives of SPECTRUM (Surveillance of Pharmacologic thErapy for Cardioversion in aTrial fibrillation Registry Using IV treatMent) (NCT01370629 and EUPAS2078) study were to assess the rates of adverse events and to estimate the effectiveness of the drug in a large cohort of patients with recent-onset AF.
Methods
Definitions
Recent-onset AF was defined as symptomatic episode within 7 days that will be undergoing cardioversion taking into account that about 70% of patients with symptomatic AF < 72 h were reported to convert spontaneously [19]. Beyond 7 days, AF is likely to persist and the chances of pharmacological cardioversion to be successful become low. Hypertension was reported when documented on the medical record or the patient report. Coronary artery disease (CAD) was diagnosed when the patient had a documented history of CAD and/or a history of coronary revascularization.
Patients and Procedures
Adult patients (≥ 18 years) with recent-onset AF occurring between September 1, 2011 and April 11, 2018 who received vernakalant for cardioversion were eligible for inclusion in this international, multicenter, observational, post-authorization study. Fifty-five hospitals in Austria, Denmark, Germany, Spain, Sweden, and Finland participated in the study, 53 of which enrolled patients. While administration of vernakalant was at the discretion of the treating physician, consecutively treated patients were enrolled and reasons for non-participation were documented. A preinfusion checklist and healthcare provider educational card were implemented during the study period to assist in identifying patients for treatment consistent with the approved indications and contraindications.
Patients were required to give informed consent for participation in the study and could be enrolled more than once if they presented on multiple occasions for AF episodes. Patients who had participated in an investigational drug/device clinical trial within 30 days prior to enrollment were not eligible. In order to enhance enrollment and reach the EMA required target of 2000 episodes, a protocol amendment was made in September 2016, which permitted retrospective inclusion of patients who had received vernakalant between April 2013 and the end of the study, provided that they fulfilled the established eligibility criteria. For prospectively enrolled patients, data were collected from both medical records and supplemental standardized data collection forms. For retrospectively enrolled patients, only medical records were available. The study period comprised a baseline assessment and up to 24-h follow-up after completion of the last infusion or until discharge. This study was mandated and approved by the European Committee for Medicinal Products for Human Use. The study protocol was approved by the appropriate local research ethics committees for all participating centers, and the study was conducted in accordance with applicable national and local regulations/guidelines, accepted standards for Good Clinical Practice, Guidelines for Good Pharmacoepidemiology Practices, and the Declaration of Helsinki [20].
Study Objectives and Endpoints
The primary objectives of the study was to estimate the incidence of clinically predefined serious adverse events (SAEs), i.e., significant hypotension (systolic blood pressure < 90 mmHg or requiring vasopressors); sustained (> 30 s) ventricular arrhythmias, Torsade de Pointes (>10 s) or ventricular fibrillation, atrial flutter with 1:1 conduction, bradycardia requiring temporary electrical pacing, or sinus arrest (> 3 s). Definition of these predefined SAEs was based on events from previous controlled studies on IV vernakalant [7, 8, 11, 12] and from the reported adverse events (AEs) on other antiarrhythmic agents. Secondary objectives included the rates of all other SAEs. Each SAE was reviewed and adjudicated by an independent expert Safety Review Committee (SRC). This study had also the objective to determine the conversion rate to sinus rhythm in a large population of patients outside the setting of controlled clinical trials.
The duration of the index AF episode was calculated as the time between the patient-reported time of symptom onset and the start of the first vernakalant infusion. Successful cardioversion was defined as conversion to sinus rhythm within 90 min of the start of vernakalant infusion. Conversion rate was calculated in all patients, as well as in an effectiveness population excluding all treatment episodes in which patients received another therapy for cardioversion within 90 min of the start of vernakalant administration (e.g., electrical or pharmacological cardioversion). Vernakalant is recommended to be administered in a step-dose fashion. Each treatment episode can comprise up to two infusions, separated by a 15-min observation period. The recommended doses for the first and second infusions are 3.0 mg/kg and 2.0 mg/kg, respectively, each administered over 10 min. For patients above 113 kg, vernakalant has a fixed initial dose of 339 mg. If conversion to sinus rhythm does not occur within 15 min after the end of the initial infusion, a second 10-min infusion of 226 mg may be administered.
Statistics and Analyses
A target sample size of 2000 vernakalant IV treatment episodes was chosen to allow adequate statistical precision, as expressed by a two-sided 95% confidence limit. Enrollment per site was capped at 10% of the total study population and 40% per country to minimize any potential bias in practice patterns. Categorical variable frequency, along with 95% confidence intervals (CIs), was determined for the summed treatment episodes. Continuous variables were summarized using descriptive statistics. Data were analyzed based on enrollment method (prospective vs retrospective) and reported as stratified and unstratified CIs. All analyses were performed using Statistical Analysis System v9.2, or later, software.
Results
Study Population
A total of 1778 patients who presented with 2009 treatment episodes were included: 1580 episodes were in prospectively enrolled patients and 429 in retrospectively enrolled patients (Table 1). The majority of patients were treated in the ED for 1289 (64.1%) AF episodes and 563 (28.0%) AF episodes in the coronary or intensive care units, with the remainder 157 (7.8%) episodes being treated in other hospital settings. As seen in Fig. 1, the main reason for non-inclusion in the study was lack of informed consent. In 1905 (94.7%) AF episodes, vernakalant was administered to non-surgery patients, and in 104 (5.2%) to post-cardiac surgery patients. The later are among the prospectively included patients. The mean age of the overall patient population at time of treatment was 62.3 ± 13.0 years (mean ± standard deviation [SD]), ranging from 18.0 to 94.0 years, and 1222 (60.8%) episodes occurred in men (Table 1). At baseline, systolic blood pressure (BP) was 132.5 ± 19.5 mmHg and heart rate (HR) was 112.9 ± 25.5/min (mean ± SD). The median duration of AF episode prior to treatment was 11.1 (5.4–27.0) hours (median [interquartile range, IQR]). In 88.9% of episodes, the patients were treated within 48 h of the onset of symptoms, and in 72.5% within 24 h. Duration of AF before treatment in 104 post-cardiac surgery patients was shorter than in the overall population, with 3.6 h (range 0.8–15.4) (median [IQR]). Baseline demographics and characteristics were similar between patients enrolled prospectively and retrospectively. Total length of ED stay was 7.5 (5.0–13.5) hours (median [IQR]). Only 167 (13.0%) of patients initially managed in the ED were in hospital for 24 h or longer. The number of vernakalant infusions was available in 1990 patients. Of these, 1201 (60.4%) received one vernakalant infusion and 789 (39.6) received a total of 2 infusions.Table 1 Clinical characteristics of patients
Total Prospective Retrospective
No. of patients 2009 1580 429
Age (years) mean ± SD 62.3 ± 13.0 61.9 ± 13.5 63.6 ± 11.2
Range (years) 18.0–94 18–93 30–94
Male, n (%) 1222 (60.8) 998 (63.2) 224 (52.2)
Body weight (kg) mean ± SD 84.1 ± 16.5 84.3 ± 16.5 83.1 (16.9)
Range (kg) 45.0–189.0 45.0–189.0 45.0–165.0
Body mass index (kg/m2) 27.8 ± 4.9 27.7 ± 4.8 28.2 ± 5.1
Associated conditions, n (%)
Hypertension 1103 (54.9) 884 (55.9) 219 (51.0)
Coronary artery disease 118 (5.9) 82 (5.2) 36 (8.4)
Cardiomyopathy 33 (1.6) 31 (2.0) 2 (0.5%)
Heart failure (history) 63 (3.1) 59 (3.7) 4 (0.9)
Diabetes 199 (9.9) 165 (10.4) 34 (7.9)
Stroke (history) 91 (4.5) 68 (4.3) 23 (5.4)
Pacemaker/ICD 36 (1.8) 24 (1.5) 12 (2.8)
Type of AF episode
First detected 477 (23.7) 393 (24.9) 84 (19.6)
Previous history of AF 1458 (72.6) 1115 (70.6) 343 (80.0)
Onset unknown/not assessed 5 (0.2) 3 (0.2) 2 (0.5)
Post-surgery 69 (3.4) 69 (4.4) 0 (0.0)
Symptoms on admission, n (%)
Palpitations, irregular heart beat 1749 (87.1) 1337 (84.6) 412 (96.0)
Dyspnea or shortness of breath 352 (17.5) 306 (19.4) 46 (10.7)
Dizziness, light-headedness 320 (15.9) 251 (15.9) 69 (16.1)
Chest pain 271 (13.5) 220 (13.9) 51 (11.9)
Syncope, near syncope 61 (3.0) 53 (3.4) 8 (1.9)
Duration of the index episode
Less than 24 h, n (%) 1438 (72.5) 1107 (70.2) 331 (81.5)
24–48 h, n (%) 347 (17.5) 288 (18.3) 59 (14.5)
More than 48 h 199 (10.0) 183 (11.6) 16 (3.9)
Mean duration ± SD (h) 23.2 ± 44.9 24.9 ± 45.8 16.8 ± 40.6
Median (IQR 25–75) (h) 11.1 (5.44–27.03) 11.9 (5.8–29.7) 8.2 (4.8–18.3)
Antiarrhythmic agents, n (%)
Betablockers 1055 (52.5) 800 (50.6) 255 (59.4)
Calcium channels blockers 22 (1.1) 20 (1.3) 2 (0.5)
Class I agents* 85 (4.2) 71 (4.5) 14 (3.3)
Class III agents* 98 (4.9) 89 (5.6) 9 (2.1)
Digitalis glycosides 22 (1.1) 18 (1.1) 4 (0.9)
*Using the Vaughan-Williams classification
Fig. 1 Study flow chart. Flow chart showing patient enrollment in the SPECTRUM study. The term patient here refers to individual treatment episodes (asterisk). Owing to lack of informed consent (n = 500) (dagger). Other reasons included patient enrollment in an investigational drug trial in the past 30 days, spontaneous conversion to sinus rhythm, ejection fraction 30–35%, electrical cardioversion preferred, missing information regarding start of atrial fibrillation, inclusion criteria not met, other, or no reason provided or known. Source data could not be verified to confirm that vernakalant IV was administered (double dagger). Spontaneous conversion to sinus rhythm before vernakalant IV administration (section sign). Patient decision and lack of follow-up after cardioversion in one case each (double vertical line). IV intravenous
Predefined Serious Adverse Events and Other Adverse Events
No deaths were recorded in our study. Nineteen predefined SAEs were reported during or after 17 treatment episodes (cumulative incidence 0.8%; CI 0.5–1.4%) (Table 2). Eighteen of the 19 events occurred within 2 h from the start of infusion. The remaining event was an episode of atrial flutter with 1:1 conduction which occurred 3.1 h after drug infusion and was terminated by electrical shock. Symptomatic bradycardia was the most common event occurring in 11 (0.5%; CI 0.4–1.2%) episodes (Table 2). Conversion to sinus rhythm occurred in 10 of these cases. A pause described as sinus arrest preceding the restoration of sinus rhythm occurred in 4 patients. In 2 patients, sinus arrest was associated with sinus bradycardia. In all bradycardia and sinus arrest cases, the vernakalant infusion was immediately discontinued. One of these 4 sinus arrests occurred in a 66-year-old man, sportive cyclist with no history of heart disease, admitted for a first episode of AF with a mean ventricular response of 95 beats/min. He received 300 mg orally of flecainide which failed to restore sinus rhythm. The treating physician decided 4 h later, to administer IV vernakalant. At the end of the infusion, a pause of 6 s, with a brief dizziness, occurred and resolved spontaneously, followed by a normal sinus rhythm with a HR of 47 beats/min which was patient usual HR and a BP of 120/85 mmHg. This event was considered a SAE although there was probably an interaction between oral flecainide still active and vernakalant in this event. One of the bradycardia events occurred in a retrospectively enrolled 69-year-old woman on bisoprolol with a history of hypertension and CAD, who developed 8 min after the second infusion of vernakalant a sinus bradycardia which rapidly resolved with IV atropine. Two bradycardia episodes occurred in post-cardiac surgery patients requiring temporary electrical pacing through the electrodes left in place by the surgeon. Both patients converted to sinus rhythm. None of the non-surgery patients required temporary electrical pacing. Significant hypotension occurred on two (0.1%; CI < 0.1–0.4%) occasions, associated with sinus bradycardia in both instances. Both events resolved with intravenous atropine and fluid. There were two cases of atrial flutter with 1:1 ventricular conduction terminated with electrical shock whereas no cases of sustained ventricular tachycardia (VT), ventricular fibrillation, or Torsade de Pointes were observed. In addition to the predefined SAEs, there were 9 other SAEs, one of which occurred in a retrospectively enrolled patient (Table 2). They included two instances of hypotension not requiring vasopressor agents, 2 non-sustained VT which deserve special attention. The first non-sustained VT occurred in a 48-year-old man with asthma admitted with fever, palpitations, dyspnea, and first episode of AF with a ventricular rate of 144 bpm. During vernakalant infusion, 5 beats of non-sustained VT was observed. Among the tests done, coronary angiography was reported as normal. The same run of 5 beats of non-sustained VT was observed 20 h after infusion (next day) making the causal effect of vernakalant unlikely. The other event occurred in a 57-year-old patient with a 6-year history of recurrent symptomatic AF and arterial hypertension with left ventricular hypertrophy. He was admitted with palpitations, irregular heartbeats, and dizziness. He was on dronedarone, and ECG showed AF with a ventricular rate of 135 bpm. During infusion of vernakalant, he had 6 s of non-sustained VT observed on the monitor and was given 5 mg of bisoprolol which reduced the heart rate to 120 beats/min and relieved patient symptoms. The Safety Review Committee considered that in the first case, the wide QRS complexes were due to aberrant conduction during rapid AF (Ashman phenomenon). Among the non-predefined SAEs, one supraventricular tachycardia (120 beats/min) and a single report each of angina pectoris, pericardial effusion, transient visual disturbance, and vernakalant overdose (Table 2). A total of 188 non-serious AEs were reported, the most common of which were dysgeusia (n = 35) and sneezing (n = 27). All patients with vernakalant-related AEs recovered without sequelae. All but 6 of the 28 SAEs were considered by the investigators and the SRC to be related to vernakalant administration.Table 2 Adverse events in 2009 episodes during treatment and observation periods
Event type Number of events Incidence (95% CI) Considered drug-related, n (%)
All SAEs 28 1.3% (0.8–1.9) 22 (78.6)
Predefined SAEs 19 0.8% (0.5–1.4) 18 (94.7)
Significant hypotension 2 0.1% (< 0.1–0.4) 2 (100.0)
Bradycardiaα 11 0.5% (0.3–10) 10 (93.3)
Sinus arrest (> 3 s)β 4 0.2% (< 0.1–0.4) 4 (100.0)
Atrial flutter with 1: 1 AV conduction 2 0.1% (0.1–0.4) 2 (100.0)
Ventricular tachycardia γ 0 0 0 (0.0)
Other than predefined SAEs 9 0.45% 5 (55.6)
Hypotension 2 0.1% 1 (50.0)
Supraventricular tachycardiaδ 1 < 0.1% 1 (100.0)
Non-sustained ventricular tachycardiaε 2 < 0.1% 1 (50.0)
Angina pectoris 1 (< 0.1) < 0.1% 0 (0.0)
Pericardial effusion 1 (< 0.1) < 0.1% 0 (0.0)
Visual disturbance 1 (< 0.1) < 0.1% 0 (0.0)
Vernakalant overdoseζ 1 (< 0.1) < 0.1% 1 (100.0)
αNine cases of sinus bradycardia and 2 reported as significant bradycardia
βOne patient had both sinus arrest followed by sinus bradycardia
γOne event reclassified as atrial flutter with 1:1 conduction
δAtrial arrhythmia other than atrial flutter
εSee text, exceeding 5% of the weight-based dosing recommendation. In this case, the administered dose was 51% in excess of the recommended dose
Rates of Conversion to Sinus Rhythm
Overall, conversion to sinus rhythm at any time following vernakalant infusion occurred in 1448 out of 2009 (72.1%) treatment episodes. Successful cardioversion was recorded in 70.2% (CI 68.1–72.2%) of the 1936 episodes of the effectiveness population excluding those in which either electrical cardioversion (n = 68) or an additional intravenous Class I/III antiarrhythmic drug (n = 6) was given within 90 min of infusion initiation. The rate of cardioversion was similar between the 1107 of 1580 (70.1%) episodes included prospectively and the 297 of 421 (70.5%) episodes of retrospectively enrolled patients. Successful cardioversion of AF was reported in 68 of 104 (65.4%) of treatment episodes in the post-cardiac surgery patients. Time to cardioversion was recorded in 1413 of 1448 episodes with successful conversion to sinus rhythm. The median time to conversion was 12.0 (8.0–28.0) minutes (median [IQR]) Fig. 2). One thousand one hundred eight of 1413 (78.4%) successful cardioversions were treated with only one drug infusion. The percentage of successful cardioversion was 70.1% in the prospective patients and 70.5% in the retrospective patients. The median hospital stay time in those treated in the ED was 7.5 h allowing patient discharges when their condition was clinically stable.Fig. 2 Time to conversion to sinus rhythm. Time to conversion to sinus rhythm with vernakalant IV in the effectiveness analysis population (N = 1936). Time to conversion was not recorded in 29 treatment episodes in which patients converted to sinus rhythm; these episodes are not displayed on the graph but are taken into account for the proportion calculation. IV intravenous
Anticoagulation
About a quarter of patients presenting with recent-onset AF at baseline were on vitamin K antagonists or direct oral anticoagulants. Investigators respected current guidelines [3] on anticoagulation both peri-procedurally and after hospital discharge.
Discussion
The SPECTRUM study included a large real-world patient population of 1778 patients with 2009 recent-onset AF episodes in whom pharmacological cardioversion was performed with vernakalant. About 70% of patients were cardioverted within 12 min from onset of infusion and 11 h from the AF onset. Our findings confirm the safety and efficacy of vernakalant reported in previous studies [7–18, 21–25] and extend their consistency to routine hospital use in large populations. To our knowledge, the present study provides the largest series of patients with recent-onset AF undergoing pharmacological cardioversion with a specific antiarrhythmic agent. The safety was the main objective of this study. We found the incidence of both predefined and other SAEs to be lower than expected. There were no death and no sustained ventricular arrhythmia. Overall, 28 SAEs (1.3%) were recorded. The majority of patients were AF treated in ED and intensive care units.
Pharmacological cardioversion is frequently indicated as part of a rhythm control strategy or as a tool to control patient symptoms and avoid hospitalization in clinically stable condition [25, 26]. It is often preferred to electrical cardioversion in patients with hemodynamically stable condition as it does not require general anesthesia or sedation. Among agents currently available for rapid termination of recent-onset AF, vernakalant represents an option [3]. However, there has been to our knowledge, no large study exploring the safety of vernakalant in daily practice.
There is no universal definition for recent-onset AF. In current literature, the duration limits of AF episodes range from < 24 [27] to < 48 h and even < 7 days [28, 29]. The prevalence of recent-onset AF among all AF subsets varies from 11% when restricted to the first detected episode (new onset) [30] to 26% [31]. The characteristics of patients were similar to those of other AF cohorts [31, 32].
As with electrical cardioversion, pharmacological cardioversion can be associated with post-cardioversion bradyarrhythmias, often unmasking pre-existing sinus node dysfunction or atrioventricular conduction abnormalities and can result in ventricular escape rhythms or prolonged ventricular pauses. Of interest, these pauses were first reported by Lown [33], following electrical cardioversion as the possible reflect of sinus dysfunction. Another possible mechanism for these sinus arrests is right atrial stunning [34]. The cumulative incidences of bradycardia (0.5%), sinus arrest (0.2%), and hypotension (0.1%) observed in this study were also low. The incidence of atrial flutter with 1:1 conduction was lower than that reported with oral Class Ic antiarrhythmics, such as flecainide or propafenone. The “pill in the pocket” approach requires initiation of therapy in hospital to verify its safety [35]. No cases of Torsade de Pointes or sustained VT was observed, which is in line with the low risk of ventricular proarrhythmia associated with vernakalant owing to its electrophysiological properties [5]. This contrasts with the reported incidence of Torsade de Pointes [24, 36, 37] in patients with AF/atrial flutter of 4.3% with intravenous ibutilide in the report of Kowey et al. including 1.7% of which required cardioversion [36]. Of note, all but one of the predefined SAEs in this study occurred within 2 h of the start of infusion. As aforementioned, the remaining patient had atrial flutter with 1:1 ventricular conduction which occurred 3.1 h following infusion initiation, indicating that close cardiac monitoring should be available during and after treatment in some patients.
Conversion to sinus rhythm with vernakalant was rapid (median time of 12.0 min) similar to what was previously reported [9–18]. The conversion median time of ibutilide was significantly longer than that of vernakalant (26 min versus 10 min, P = 0.01) in a randomized comparison [18]. Furthermore, in this particularly large real-world study, the median duration of AF episode was short (11.1 h) as there is important evidence, and relevant guidelines [3] suggesting that prompt cardioversion could be associated with benefits in terms of lower risk of thromboembolic events [4, 38, 39].
Although the baseline characteristics of the study population were consistent with AF population-based studies [31, 32] and clinical studies with vernakalant, the conversion rate was higher than that observed in recent review and meta-analysis (~ 50%) [22–25]. This seems likely to be due to patients being treated soon after symptom onset in European clinical practice. Other recent but smaller observational studies [13–15, 17, 18], which collectively included almost 1300 patients, have found similarly high conversion rates (65–86%) when vernakalant was administered soon after the onset of AF, particularly within the first 48 h [15, 17, 21]. Vernakalant has also been shown to induce a higher rate of cardioversion compared with flecainide (67% vs 46%) in a non-randomized cohort study [21]. Similarly, in randomized studies, vernakalant was more effective than amiodarone [12] (52% vs 5%; after 90 min) and ibutilide [18, 24] (69% vs 43%; within 90 min). The SPECTRUM results are consistent with previous reports that vernakalant is safe and effective for the rapid cardioversion of recent-onset AF and extends them to daily practice.
Owing to the rapid time to conversion with vernakalant, the median hospital stay time for those treated in the ED was 7.5 h. This is encouraging given that a study in France reported that hospitalization constitutes 60% of the cost of care for patients with AF [40].
Study Limitations
This multicenter international study was observational as the main objective was to determine the safety of vernakalant as used in daily hospital practice without interfering on the management of recent-onset AF by the treating physician. For these reasons, the adverse events were expected to be higher in an “uncontrolled” setting with no guidance on patient selection than those reported in controlled studies with strict protocols. In fact, SAEs were low in this study. Data collection for prospectively enrolled patients was comprehensive owing to use of both study-specific tools and medical records. However, for retrospectively enrolled patients, it was not possible to routinely collect all data of interest in a standardized manner. Nevertheless, baseline characteristics, medical histories, and SAEs in the retrospective cohort were similar to those in the prospective cohort, supporting the use of a retrospective analysis.
Conclusions
The results of this large multicenter study showed that vernakalant has a good safety profile and is effective in enabling rapid cardioversion in clinical practice. Moreover, the rates of serious complications were lower than those observed in early trials reflecting appropriate patient selection in clinical practice. In conclusion, vernakalant provides a rapid and effective means of pharmacological conversion in patients with recent-onset AF undergoing cardioversion undergoing cardioversion in daily hospital practice.
Appendix. List of Investigators
Austria: Michael Joannidis, Klemens Zotter Frank Hartig, Anton Sandhofer, Alois Süssenbacher, Bernd Eber, Elisabeth Lassnig, Ulrike Pfeifenberger, Michaela Steiner, Hans Domanovits, Alexander Simon, Alexander Spiel, Jan Niederdockl, Nikola Schuetz, Daniel Wehinger, Franz Xaver Roithinger, Isabella M. von Katzler, Katharina Bichler, Robert Schoenbauer, Lukas Fiedler, Michael Pfeffer, Markus Peck, Florian Benische, Michael Hackl, Susane Demschar, Astrid Ebner, Melanie Eder, Rainer Huditz, Arnulf Isak, Michael Moser, Georg Pinter, Thomas Singer, Claudia Waldhauser, Helmut Pürerfellner, Martin Martinek, Sandra Muellner, Andrea Ploechl, Tanja Koppler, Elisabeth Sigmund, Michael Derndorfer, Sabine Metz, Karin Streicher, Clemens Steinwender, Karim Saleh, Andreas Lueger, Petra Fladerer, Eiko Meister, Heinz Drexel, Alexandra Schuler, Susanne Waeger, Karl-Martin Ebner, Christine Heinzle, Arthur Mader, Peter Schwerzler, Berta Patsch, Abdurahman Said, Claudia Stoeckloecker, Daniela Zanolin, Jutta Bergler-Klein, Ljubica Mandic, Mariann Gyöngyösi, Neraida Cene, Zsuzsanna Szankai, Abelina Zimba.
Denmark: Henrik Nielsen, Bjerre Flemming, Michaelsen Michaelsen, Elisa Stokholm, Katja Holm, Charlotte Schmidt Skov, Pauline Gøgsig Johansen, Soren shjortshoj, Thomas Melchior, Ole Dyg Pedersen, Sanne Heinsvig, Inge Larsen, Vibeke Perret-Gentil, Thomas Wagner Nielsen, Axel Brandes, Marianne Jensen, Ida Rosenlund, Liv Gøtzsche, Heidi Munk Andersen.
Germany: Andreas Götte, Matthias Hammwohner, Britta Möehring, Jutta Schaertl, Daniel Steven, Iris Berg, Alexandra Kuehn, Hannes Reuter, Elena Terentieva, Christian Loges, Christine Lindner, Hendrik Bonnemeier, Christan Wulff, Thomas Demming, Svenja Gediehn, Johanna Parlitz, Wilhelm Haverkamp, Buehner Kathrin, Hubert Katja, Iacovella Ines, Bernhard Korbmacher, Marc Thone, Hannan Dalyanoglu, Da Un Chung, Naujoks Angela, Dirk Weismann, Björn Lengenfelder, Jan Becher, Klaus Meyer, Irina Turkin, Sebastian Maier, Marcus Koller, Alban Glaser, Lisa Gebele, Jale Goezuebueyuek, Ralph Hampe, Barbara Ruemmler, Hagen Schrötter, Manja Hubald, Cornelia Fritz, Martin Domhardt, Kathrin Haacke, Nicole Schmiedehausen, Ruth Strasser, Kristof Graf, Lidia Fischer, Roland Thieme, Karlheinz Seidl, Martin Kulzer, Monika Zackel, Gerian Grönefeld, Christina Paitazoglou, Simone Müller, ThoraBotschafter Britta Goldmann, Andrea Moeller, Sindy Bartel, Joern Schmitt, Damir Erkapic, Gabriele Hellwig-Bahavar, Ritvan Chasan, Christopher Gemein, Victoria Johnson, Christiane Kelm, Kay Weipert, Johannes Brachmann, Michael Held, Andrea Höhn, Ute Goebel, Andrea Linss, Swetlana Rube, Ahmed Saleh, Steffen Schnupp, Yeong-Hoon Choi, Vera Wolf, Andrea Plate, Anton Sabashnikov, Antje-Christin Deppe, Petra Krause.
Spain: Ignacio Fernandez Lozano, Manuel Sánchez, Francisco Hernández, Alfonso Martin Martinez, Pedro Vazquez, Esther Alvarez, Pascual Lopez, Raquel Torres, José Carbajosa Dalmau, Laura Parades, Nestor Hernandez, Inmaculada Jimenez Ruiz, Ana Maria Lopez, Luis López-Andujar, Alexandre Noguera, Francisco Roman Cerdan, Carmen del Arco Galan, Daniel Afonso, Raquel Caminero, Manual Lunquera, Monica Negro, Cristina Santiago, Nestor Villalba, José Manuel Garrido Castilla, Roberto Martinez Asenjo, Elena Mejia Martinez, José Luís Merino Llorens, Maria Jesus Diaz-Pintado, Jorge Alejandro Figueroa, Alberto Borobia Perez, Sergio Castrejon Castrejon, David Filgueiras Rama, Manuel Quintana Diaz, Maria Angelica Ribera Nunez, Miguel Angel Ramirez Marrero, Antonio Martin, José Miguel Ormaetxe Merodio, Mercedes Varona Peinador, Maria Fe Arcocha, Larraitz Gaztañaga, F. Xavier Palom Rico, Javier Jacob Rodriguez, Pascual Piñera Salmeron, Juan Cosin Sales, Isabel Navarro, Francisco Buendia Funetes.
Sweden: Henrik Wallentin, Kerstin Roos, Arash Mokhtari, Hans-Jörgen Nilsson, Peter Vasko, Terese Nyström, Martina Gustensson, Susanne Johansson, Inga Uggeldahl, Göran Andersson, Olle Bergström, Thomas Aronsson, Mehmet Hamid, Kerstin Giocondi, Deborah Svanerö, Qassim Awad, Tord Juhlin, Hjördis Jernhed, Stefan Berglund, Magnus Forsgren, Michael Guggi, Pär-Lennart Agren, Kristina Eriksson, Kristina Karlsson, Per Blomström, Alejandro Utreras, Caroline Lundgren, Maria Soderlund, Frederik Buijs, Solveig Östberg, Johan-Emil Bager, Ingrid Hendequist, Maria Just, Siv Heden, Liselott Lisjo, Chrichan Mansson, Helen Svanstrom, Mikael Dellborg, Helena Dellborg, Gorel Hultsberg Olsson, Linus Hansson.
Finland: Hannu Sulonen, Hanna Suurmunne, Anna Petrovskaja, Johanna Markkanen, Juha Hartikainen, Lari Kujanen, Antti Heikkola, Hanna Pohjantahti MaarooS.
Safety Review Committee
Lars Kober, MD, Samuel Lévy, MD (Chair), Cristina Varas-Lorenzo, MD, MSc, PhD, Manel Pladevall-Vila MD, MS.
The authors wish to thank Nathalie Dunkel and her team for their help in accessing the data of SPECTRUM and providing information and documents upon our request.
The authors would like to thank all of the investigators involved in the SPECTRUM study.
Funding
This study was funded by Correvio International Sàrl, Geneva, Switzerland. The authors received no funding for their participation in this manuscript.
Data Availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Compliance with Ethical Standards
Conflict of Interest
Professor Levy reports no conflicts of interest. Professor Hartikainen has been an investigator in studies sponsored by AstraZeneca, Biosense Webster, Boehringer Ingelheim, Correvio International Sàrl, Medtronic, and St. Jude Medical. Dr. Ritz is an employee of Correvio International Sàrl, Geneva, Switzerland. Dr. Juhlin has received speaker honoraria from Correvio International Sàrl. Professor Domanovits reports no conflicts of interest. Dr. Carbajosa-Dalmau has received speaker honoraria from Correvio International Sàrl.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. | FLECAINIDE, VERNAKALANT | DrugsGivenReaction | CC BY | 33206300 | 19,673,350 | 2021-04 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Drug interaction'. | Vernakalant for Rapid Cardioversion of Recent-Onset Atrial Fibrillation: Results from the SPECTRUM Study.
Rapid restoration of sinus rhythm using pharmacological cardioversion is commonly indicated in patients with symptomatic recent-onset atrial fibrillation (AF). The objectives of this large, international, multicenter observational study were to determine the safety and effectiveness of intravenous (IV) vernakalant for conversion of AF to sinus rhythm in daily practice.
Consenting patients with symptomatic recent-onset AF (< 7 days) treated with IV vernakalant were enrolled and followed up to 24 h after the last infusion or until discharge, in order to determine the incidence of predefined serious adverse events (SAEs) and other observed SAEs and evaluate the conversion rate within the first 90 min. Overall, 2009 treatment episodes in 1778 patients were analyzed. The age of patients was 62.3 ± 13.0 years (mean ± standard deviation). Median AF duration before treatment was 11.1 h (IQR 5.4-27.0 h). A total of 28 SAEs occurred in 26 patients including 19 predefined SAEs, i.e., sinus arrest (n = 4, 0.2%), significant bradycardia (n = 11, 0.5%), significant hypotension (n = 2, 0.1%), and atrial flutter with 1:1 conduction (n = 2, 0.1%). There were no cases of sustained ventricular arrhythmias or deaths. All patients who experienced SAEs recovered fully (n = 25) or with sequelae (n = 1). Conversion rate to sinus rhythm was 70.2%, within a median of 12 min (IQR 8.0-28.0 min).
This large multicenter, international observational study confirms the good safety profile and the high effectiveness of vernakalant for the rapid cardioversion of recent-onset AF in daily hospital practice.
Introduction and Purpose of the Study
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, with an estimated 33.5 million people affected worldwide [1]. One in four adults over 55 years of age in Europe and the USA develop AF, with greater prevalence in older populations [1, 2]. Patients with AF are at increased risk of stroke and heart failure [3, 4]. A significant number of patients with recent-onset AF seen in the emergency departments (EDs) undergo commonly in Europe pharmacological cardioversion.
Vernakalant is a partial atrial-selective antiarrhythmic agent by its action through IKur and IKACh channel inhibition [5]. However, it has a modest effect on the ventricle via Ina and IKr channels resulting in a limited effect on ventricular repolarization (QT interval) [5]. Vernakalant is contra-indicated in patients with prolonged QT interval.
Intravenous vernakalant has been approved by the European Medicine Agency [2010] for the rapid conversion of recent-onset AF [6]. To date, a number of studies have shown vernakalant to be well tolerated and effective for cardioversion of AF [7–18].
The FDA (Food and Drug Administration agency) decided in 2008 and in December 2019 not to approve to market vernakalant in the USA for safety concerns. In 2010, the EMA requested a post-authorization safety study to better define the risk benefit ratio in routine clinical practice. The objectives of SPECTRUM (Surveillance of Pharmacologic thErapy for Cardioversion in aTrial fibrillation Registry Using IV treatMent) (NCT01370629 and EUPAS2078) study were to assess the rates of adverse events and to estimate the effectiveness of the drug in a large cohort of patients with recent-onset AF.
Methods
Definitions
Recent-onset AF was defined as symptomatic episode within 7 days that will be undergoing cardioversion taking into account that about 70% of patients with symptomatic AF < 72 h were reported to convert spontaneously [19]. Beyond 7 days, AF is likely to persist and the chances of pharmacological cardioversion to be successful become low. Hypertension was reported when documented on the medical record or the patient report. Coronary artery disease (CAD) was diagnosed when the patient had a documented history of CAD and/or a history of coronary revascularization.
Patients and Procedures
Adult patients (≥ 18 years) with recent-onset AF occurring between September 1, 2011 and April 11, 2018 who received vernakalant for cardioversion were eligible for inclusion in this international, multicenter, observational, post-authorization study. Fifty-five hospitals in Austria, Denmark, Germany, Spain, Sweden, and Finland participated in the study, 53 of which enrolled patients. While administration of vernakalant was at the discretion of the treating physician, consecutively treated patients were enrolled and reasons for non-participation were documented. A preinfusion checklist and healthcare provider educational card were implemented during the study period to assist in identifying patients for treatment consistent with the approved indications and contraindications.
Patients were required to give informed consent for participation in the study and could be enrolled more than once if they presented on multiple occasions for AF episodes. Patients who had participated in an investigational drug/device clinical trial within 30 days prior to enrollment were not eligible. In order to enhance enrollment and reach the EMA required target of 2000 episodes, a protocol amendment was made in September 2016, which permitted retrospective inclusion of patients who had received vernakalant between April 2013 and the end of the study, provided that they fulfilled the established eligibility criteria. For prospectively enrolled patients, data were collected from both medical records and supplemental standardized data collection forms. For retrospectively enrolled patients, only medical records were available. The study period comprised a baseline assessment and up to 24-h follow-up after completion of the last infusion or until discharge. This study was mandated and approved by the European Committee for Medicinal Products for Human Use. The study protocol was approved by the appropriate local research ethics committees for all participating centers, and the study was conducted in accordance with applicable national and local regulations/guidelines, accepted standards for Good Clinical Practice, Guidelines for Good Pharmacoepidemiology Practices, and the Declaration of Helsinki [20].
Study Objectives and Endpoints
The primary objectives of the study was to estimate the incidence of clinically predefined serious adverse events (SAEs), i.e., significant hypotension (systolic blood pressure < 90 mmHg or requiring vasopressors); sustained (> 30 s) ventricular arrhythmias, Torsade de Pointes (>10 s) or ventricular fibrillation, atrial flutter with 1:1 conduction, bradycardia requiring temporary electrical pacing, or sinus arrest (> 3 s). Definition of these predefined SAEs was based on events from previous controlled studies on IV vernakalant [7, 8, 11, 12] and from the reported adverse events (AEs) on other antiarrhythmic agents. Secondary objectives included the rates of all other SAEs. Each SAE was reviewed and adjudicated by an independent expert Safety Review Committee (SRC). This study had also the objective to determine the conversion rate to sinus rhythm in a large population of patients outside the setting of controlled clinical trials.
The duration of the index AF episode was calculated as the time between the patient-reported time of symptom onset and the start of the first vernakalant infusion. Successful cardioversion was defined as conversion to sinus rhythm within 90 min of the start of vernakalant infusion. Conversion rate was calculated in all patients, as well as in an effectiveness population excluding all treatment episodes in which patients received another therapy for cardioversion within 90 min of the start of vernakalant administration (e.g., electrical or pharmacological cardioversion). Vernakalant is recommended to be administered in a step-dose fashion. Each treatment episode can comprise up to two infusions, separated by a 15-min observation period. The recommended doses for the first and second infusions are 3.0 mg/kg and 2.0 mg/kg, respectively, each administered over 10 min. For patients above 113 kg, vernakalant has a fixed initial dose of 339 mg. If conversion to sinus rhythm does not occur within 15 min after the end of the initial infusion, a second 10-min infusion of 226 mg may be administered.
Statistics and Analyses
A target sample size of 2000 vernakalant IV treatment episodes was chosen to allow adequate statistical precision, as expressed by a two-sided 95% confidence limit. Enrollment per site was capped at 10% of the total study population and 40% per country to minimize any potential bias in practice patterns. Categorical variable frequency, along with 95% confidence intervals (CIs), was determined for the summed treatment episodes. Continuous variables were summarized using descriptive statistics. Data were analyzed based on enrollment method (prospective vs retrospective) and reported as stratified and unstratified CIs. All analyses were performed using Statistical Analysis System v9.2, or later, software.
Results
Study Population
A total of 1778 patients who presented with 2009 treatment episodes were included: 1580 episodes were in prospectively enrolled patients and 429 in retrospectively enrolled patients (Table 1). The majority of patients were treated in the ED for 1289 (64.1%) AF episodes and 563 (28.0%) AF episodes in the coronary or intensive care units, with the remainder 157 (7.8%) episodes being treated in other hospital settings. As seen in Fig. 1, the main reason for non-inclusion in the study was lack of informed consent. In 1905 (94.7%) AF episodes, vernakalant was administered to non-surgery patients, and in 104 (5.2%) to post-cardiac surgery patients. The later are among the prospectively included patients. The mean age of the overall patient population at time of treatment was 62.3 ± 13.0 years (mean ± standard deviation [SD]), ranging from 18.0 to 94.0 years, and 1222 (60.8%) episodes occurred in men (Table 1). At baseline, systolic blood pressure (BP) was 132.5 ± 19.5 mmHg and heart rate (HR) was 112.9 ± 25.5/min (mean ± SD). The median duration of AF episode prior to treatment was 11.1 (5.4–27.0) hours (median [interquartile range, IQR]). In 88.9% of episodes, the patients were treated within 48 h of the onset of symptoms, and in 72.5% within 24 h. Duration of AF before treatment in 104 post-cardiac surgery patients was shorter than in the overall population, with 3.6 h (range 0.8–15.4) (median [IQR]). Baseline demographics and characteristics were similar between patients enrolled prospectively and retrospectively. Total length of ED stay was 7.5 (5.0–13.5) hours (median [IQR]). Only 167 (13.0%) of patients initially managed in the ED were in hospital for 24 h or longer. The number of vernakalant infusions was available in 1990 patients. Of these, 1201 (60.4%) received one vernakalant infusion and 789 (39.6) received a total of 2 infusions.Table 1 Clinical characteristics of patients
Total Prospective Retrospective
No. of patients 2009 1580 429
Age (years) mean ± SD 62.3 ± 13.0 61.9 ± 13.5 63.6 ± 11.2
Range (years) 18.0–94 18–93 30–94
Male, n (%) 1222 (60.8) 998 (63.2) 224 (52.2)
Body weight (kg) mean ± SD 84.1 ± 16.5 84.3 ± 16.5 83.1 (16.9)
Range (kg) 45.0–189.0 45.0–189.0 45.0–165.0
Body mass index (kg/m2) 27.8 ± 4.9 27.7 ± 4.8 28.2 ± 5.1
Associated conditions, n (%)
Hypertension 1103 (54.9) 884 (55.9) 219 (51.0)
Coronary artery disease 118 (5.9) 82 (5.2) 36 (8.4)
Cardiomyopathy 33 (1.6) 31 (2.0) 2 (0.5%)
Heart failure (history) 63 (3.1) 59 (3.7) 4 (0.9)
Diabetes 199 (9.9) 165 (10.4) 34 (7.9)
Stroke (history) 91 (4.5) 68 (4.3) 23 (5.4)
Pacemaker/ICD 36 (1.8) 24 (1.5) 12 (2.8)
Type of AF episode
First detected 477 (23.7) 393 (24.9) 84 (19.6)
Previous history of AF 1458 (72.6) 1115 (70.6) 343 (80.0)
Onset unknown/not assessed 5 (0.2) 3 (0.2) 2 (0.5)
Post-surgery 69 (3.4) 69 (4.4) 0 (0.0)
Symptoms on admission, n (%)
Palpitations, irregular heart beat 1749 (87.1) 1337 (84.6) 412 (96.0)
Dyspnea or shortness of breath 352 (17.5) 306 (19.4) 46 (10.7)
Dizziness, light-headedness 320 (15.9) 251 (15.9) 69 (16.1)
Chest pain 271 (13.5) 220 (13.9) 51 (11.9)
Syncope, near syncope 61 (3.0) 53 (3.4) 8 (1.9)
Duration of the index episode
Less than 24 h, n (%) 1438 (72.5) 1107 (70.2) 331 (81.5)
24–48 h, n (%) 347 (17.5) 288 (18.3) 59 (14.5)
More than 48 h 199 (10.0) 183 (11.6) 16 (3.9)
Mean duration ± SD (h) 23.2 ± 44.9 24.9 ± 45.8 16.8 ± 40.6
Median (IQR 25–75) (h) 11.1 (5.44–27.03) 11.9 (5.8–29.7) 8.2 (4.8–18.3)
Antiarrhythmic agents, n (%)
Betablockers 1055 (52.5) 800 (50.6) 255 (59.4)
Calcium channels blockers 22 (1.1) 20 (1.3) 2 (0.5)
Class I agents* 85 (4.2) 71 (4.5) 14 (3.3)
Class III agents* 98 (4.9) 89 (5.6) 9 (2.1)
Digitalis glycosides 22 (1.1) 18 (1.1) 4 (0.9)
*Using the Vaughan-Williams classification
Fig. 1 Study flow chart. Flow chart showing patient enrollment in the SPECTRUM study. The term patient here refers to individual treatment episodes (asterisk). Owing to lack of informed consent (n = 500) (dagger). Other reasons included patient enrollment in an investigational drug trial in the past 30 days, spontaneous conversion to sinus rhythm, ejection fraction 30–35%, electrical cardioversion preferred, missing information regarding start of atrial fibrillation, inclusion criteria not met, other, or no reason provided or known. Source data could not be verified to confirm that vernakalant IV was administered (double dagger). Spontaneous conversion to sinus rhythm before vernakalant IV administration (section sign). Patient decision and lack of follow-up after cardioversion in one case each (double vertical line). IV intravenous
Predefined Serious Adverse Events and Other Adverse Events
No deaths were recorded in our study. Nineteen predefined SAEs were reported during or after 17 treatment episodes (cumulative incidence 0.8%; CI 0.5–1.4%) (Table 2). Eighteen of the 19 events occurred within 2 h from the start of infusion. The remaining event was an episode of atrial flutter with 1:1 conduction which occurred 3.1 h after drug infusion and was terminated by electrical shock. Symptomatic bradycardia was the most common event occurring in 11 (0.5%; CI 0.4–1.2%) episodes (Table 2). Conversion to sinus rhythm occurred in 10 of these cases. A pause described as sinus arrest preceding the restoration of sinus rhythm occurred in 4 patients. In 2 patients, sinus arrest was associated with sinus bradycardia. In all bradycardia and sinus arrest cases, the vernakalant infusion was immediately discontinued. One of these 4 sinus arrests occurred in a 66-year-old man, sportive cyclist with no history of heart disease, admitted for a first episode of AF with a mean ventricular response of 95 beats/min. He received 300 mg orally of flecainide which failed to restore sinus rhythm. The treating physician decided 4 h later, to administer IV vernakalant. At the end of the infusion, a pause of 6 s, with a brief dizziness, occurred and resolved spontaneously, followed by a normal sinus rhythm with a HR of 47 beats/min which was patient usual HR and a BP of 120/85 mmHg. This event was considered a SAE although there was probably an interaction between oral flecainide still active and vernakalant in this event. One of the bradycardia events occurred in a retrospectively enrolled 69-year-old woman on bisoprolol with a history of hypertension and CAD, who developed 8 min after the second infusion of vernakalant a sinus bradycardia which rapidly resolved with IV atropine. Two bradycardia episodes occurred in post-cardiac surgery patients requiring temporary electrical pacing through the electrodes left in place by the surgeon. Both patients converted to sinus rhythm. None of the non-surgery patients required temporary electrical pacing. Significant hypotension occurred on two (0.1%; CI < 0.1–0.4%) occasions, associated with sinus bradycardia in both instances. Both events resolved with intravenous atropine and fluid. There were two cases of atrial flutter with 1:1 ventricular conduction terminated with electrical shock whereas no cases of sustained ventricular tachycardia (VT), ventricular fibrillation, or Torsade de Pointes were observed. In addition to the predefined SAEs, there were 9 other SAEs, one of which occurred in a retrospectively enrolled patient (Table 2). They included two instances of hypotension not requiring vasopressor agents, 2 non-sustained VT which deserve special attention. The first non-sustained VT occurred in a 48-year-old man with asthma admitted with fever, palpitations, dyspnea, and first episode of AF with a ventricular rate of 144 bpm. During vernakalant infusion, 5 beats of non-sustained VT was observed. Among the tests done, coronary angiography was reported as normal. The same run of 5 beats of non-sustained VT was observed 20 h after infusion (next day) making the causal effect of vernakalant unlikely. The other event occurred in a 57-year-old patient with a 6-year history of recurrent symptomatic AF and arterial hypertension with left ventricular hypertrophy. He was admitted with palpitations, irregular heartbeats, and dizziness. He was on dronedarone, and ECG showed AF with a ventricular rate of 135 bpm. During infusion of vernakalant, he had 6 s of non-sustained VT observed on the monitor and was given 5 mg of bisoprolol which reduced the heart rate to 120 beats/min and relieved patient symptoms. The Safety Review Committee considered that in the first case, the wide QRS complexes were due to aberrant conduction during rapid AF (Ashman phenomenon). Among the non-predefined SAEs, one supraventricular tachycardia (120 beats/min) and a single report each of angina pectoris, pericardial effusion, transient visual disturbance, and vernakalant overdose (Table 2). A total of 188 non-serious AEs were reported, the most common of which were dysgeusia (n = 35) and sneezing (n = 27). All patients with vernakalant-related AEs recovered without sequelae. All but 6 of the 28 SAEs were considered by the investigators and the SRC to be related to vernakalant administration.Table 2 Adverse events in 2009 episodes during treatment and observation periods
Event type Number of events Incidence (95% CI) Considered drug-related, n (%)
All SAEs 28 1.3% (0.8–1.9) 22 (78.6)
Predefined SAEs 19 0.8% (0.5–1.4) 18 (94.7)
Significant hypotension 2 0.1% (< 0.1–0.4) 2 (100.0)
Bradycardiaα 11 0.5% (0.3–10) 10 (93.3)
Sinus arrest (> 3 s)β 4 0.2% (< 0.1–0.4) 4 (100.0)
Atrial flutter with 1: 1 AV conduction 2 0.1% (0.1–0.4) 2 (100.0)
Ventricular tachycardia γ 0 0 0 (0.0)
Other than predefined SAEs 9 0.45% 5 (55.6)
Hypotension 2 0.1% 1 (50.0)
Supraventricular tachycardiaδ 1 < 0.1% 1 (100.0)
Non-sustained ventricular tachycardiaε 2 < 0.1% 1 (50.0)
Angina pectoris 1 (< 0.1) < 0.1% 0 (0.0)
Pericardial effusion 1 (< 0.1) < 0.1% 0 (0.0)
Visual disturbance 1 (< 0.1) < 0.1% 0 (0.0)
Vernakalant overdoseζ 1 (< 0.1) < 0.1% 1 (100.0)
αNine cases of sinus bradycardia and 2 reported as significant bradycardia
βOne patient had both sinus arrest followed by sinus bradycardia
γOne event reclassified as atrial flutter with 1:1 conduction
δAtrial arrhythmia other than atrial flutter
εSee text, exceeding 5% of the weight-based dosing recommendation. In this case, the administered dose was 51% in excess of the recommended dose
Rates of Conversion to Sinus Rhythm
Overall, conversion to sinus rhythm at any time following vernakalant infusion occurred in 1448 out of 2009 (72.1%) treatment episodes. Successful cardioversion was recorded in 70.2% (CI 68.1–72.2%) of the 1936 episodes of the effectiveness population excluding those in which either electrical cardioversion (n = 68) or an additional intravenous Class I/III antiarrhythmic drug (n = 6) was given within 90 min of infusion initiation. The rate of cardioversion was similar between the 1107 of 1580 (70.1%) episodes included prospectively and the 297 of 421 (70.5%) episodes of retrospectively enrolled patients. Successful cardioversion of AF was reported in 68 of 104 (65.4%) of treatment episodes in the post-cardiac surgery patients. Time to cardioversion was recorded in 1413 of 1448 episodes with successful conversion to sinus rhythm. The median time to conversion was 12.0 (8.0–28.0) minutes (median [IQR]) Fig. 2). One thousand one hundred eight of 1413 (78.4%) successful cardioversions were treated with only one drug infusion. The percentage of successful cardioversion was 70.1% in the prospective patients and 70.5% in the retrospective patients. The median hospital stay time in those treated in the ED was 7.5 h allowing patient discharges when their condition was clinically stable.Fig. 2 Time to conversion to sinus rhythm. Time to conversion to sinus rhythm with vernakalant IV in the effectiveness analysis population (N = 1936). Time to conversion was not recorded in 29 treatment episodes in which patients converted to sinus rhythm; these episodes are not displayed on the graph but are taken into account for the proportion calculation. IV intravenous
Anticoagulation
About a quarter of patients presenting with recent-onset AF at baseline were on vitamin K antagonists or direct oral anticoagulants. Investigators respected current guidelines [3] on anticoagulation both peri-procedurally and after hospital discharge.
Discussion
The SPECTRUM study included a large real-world patient population of 1778 patients with 2009 recent-onset AF episodes in whom pharmacological cardioversion was performed with vernakalant. About 70% of patients were cardioverted within 12 min from onset of infusion and 11 h from the AF onset. Our findings confirm the safety and efficacy of vernakalant reported in previous studies [7–18, 21–25] and extend their consistency to routine hospital use in large populations. To our knowledge, the present study provides the largest series of patients with recent-onset AF undergoing pharmacological cardioversion with a specific antiarrhythmic agent. The safety was the main objective of this study. We found the incidence of both predefined and other SAEs to be lower than expected. There were no death and no sustained ventricular arrhythmia. Overall, 28 SAEs (1.3%) were recorded. The majority of patients were AF treated in ED and intensive care units.
Pharmacological cardioversion is frequently indicated as part of a rhythm control strategy or as a tool to control patient symptoms and avoid hospitalization in clinically stable condition [25, 26]. It is often preferred to electrical cardioversion in patients with hemodynamically stable condition as it does not require general anesthesia or sedation. Among agents currently available for rapid termination of recent-onset AF, vernakalant represents an option [3]. However, there has been to our knowledge, no large study exploring the safety of vernakalant in daily practice.
There is no universal definition for recent-onset AF. In current literature, the duration limits of AF episodes range from < 24 [27] to < 48 h and even < 7 days [28, 29]. The prevalence of recent-onset AF among all AF subsets varies from 11% when restricted to the first detected episode (new onset) [30] to 26% [31]. The characteristics of patients were similar to those of other AF cohorts [31, 32].
As with electrical cardioversion, pharmacological cardioversion can be associated with post-cardioversion bradyarrhythmias, often unmasking pre-existing sinus node dysfunction or atrioventricular conduction abnormalities and can result in ventricular escape rhythms or prolonged ventricular pauses. Of interest, these pauses were first reported by Lown [33], following electrical cardioversion as the possible reflect of sinus dysfunction. Another possible mechanism for these sinus arrests is right atrial stunning [34]. The cumulative incidences of bradycardia (0.5%), sinus arrest (0.2%), and hypotension (0.1%) observed in this study were also low. The incidence of atrial flutter with 1:1 conduction was lower than that reported with oral Class Ic antiarrhythmics, such as flecainide or propafenone. The “pill in the pocket” approach requires initiation of therapy in hospital to verify its safety [35]. No cases of Torsade de Pointes or sustained VT was observed, which is in line with the low risk of ventricular proarrhythmia associated with vernakalant owing to its electrophysiological properties [5]. This contrasts with the reported incidence of Torsade de Pointes [24, 36, 37] in patients with AF/atrial flutter of 4.3% with intravenous ibutilide in the report of Kowey et al. including 1.7% of which required cardioversion [36]. Of note, all but one of the predefined SAEs in this study occurred within 2 h of the start of infusion. As aforementioned, the remaining patient had atrial flutter with 1:1 ventricular conduction which occurred 3.1 h following infusion initiation, indicating that close cardiac monitoring should be available during and after treatment in some patients.
Conversion to sinus rhythm with vernakalant was rapid (median time of 12.0 min) similar to what was previously reported [9–18]. The conversion median time of ibutilide was significantly longer than that of vernakalant (26 min versus 10 min, P = 0.01) in a randomized comparison [18]. Furthermore, in this particularly large real-world study, the median duration of AF episode was short (11.1 h) as there is important evidence, and relevant guidelines [3] suggesting that prompt cardioversion could be associated with benefits in terms of lower risk of thromboembolic events [4, 38, 39].
Although the baseline characteristics of the study population were consistent with AF population-based studies [31, 32] and clinical studies with vernakalant, the conversion rate was higher than that observed in recent review and meta-analysis (~ 50%) [22–25]. This seems likely to be due to patients being treated soon after symptom onset in European clinical practice. Other recent but smaller observational studies [13–15, 17, 18], which collectively included almost 1300 patients, have found similarly high conversion rates (65–86%) when vernakalant was administered soon after the onset of AF, particularly within the first 48 h [15, 17, 21]. Vernakalant has also been shown to induce a higher rate of cardioversion compared with flecainide (67% vs 46%) in a non-randomized cohort study [21]. Similarly, in randomized studies, vernakalant was more effective than amiodarone [12] (52% vs 5%; after 90 min) and ibutilide [18, 24] (69% vs 43%; within 90 min). The SPECTRUM results are consistent with previous reports that vernakalant is safe and effective for the rapid cardioversion of recent-onset AF and extends them to daily practice.
Owing to the rapid time to conversion with vernakalant, the median hospital stay time for those treated in the ED was 7.5 h. This is encouraging given that a study in France reported that hospitalization constitutes 60% of the cost of care for patients with AF [40].
Study Limitations
This multicenter international study was observational as the main objective was to determine the safety of vernakalant as used in daily hospital practice without interfering on the management of recent-onset AF by the treating physician. For these reasons, the adverse events were expected to be higher in an “uncontrolled” setting with no guidance on patient selection than those reported in controlled studies with strict protocols. In fact, SAEs were low in this study. Data collection for prospectively enrolled patients was comprehensive owing to use of both study-specific tools and medical records. However, for retrospectively enrolled patients, it was not possible to routinely collect all data of interest in a standardized manner. Nevertheless, baseline characteristics, medical histories, and SAEs in the retrospective cohort were similar to those in the prospective cohort, supporting the use of a retrospective analysis.
Conclusions
The results of this large multicenter study showed that vernakalant has a good safety profile and is effective in enabling rapid cardioversion in clinical practice. Moreover, the rates of serious complications were lower than those observed in early trials reflecting appropriate patient selection in clinical practice. In conclusion, vernakalant provides a rapid and effective means of pharmacological conversion in patients with recent-onset AF undergoing cardioversion undergoing cardioversion in daily hospital practice.
Appendix. List of Investigators
Austria: Michael Joannidis, Klemens Zotter Frank Hartig, Anton Sandhofer, Alois Süssenbacher, Bernd Eber, Elisabeth Lassnig, Ulrike Pfeifenberger, Michaela Steiner, Hans Domanovits, Alexander Simon, Alexander Spiel, Jan Niederdockl, Nikola Schuetz, Daniel Wehinger, Franz Xaver Roithinger, Isabella M. von Katzler, Katharina Bichler, Robert Schoenbauer, Lukas Fiedler, Michael Pfeffer, Markus Peck, Florian Benische, Michael Hackl, Susane Demschar, Astrid Ebner, Melanie Eder, Rainer Huditz, Arnulf Isak, Michael Moser, Georg Pinter, Thomas Singer, Claudia Waldhauser, Helmut Pürerfellner, Martin Martinek, Sandra Muellner, Andrea Ploechl, Tanja Koppler, Elisabeth Sigmund, Michael Derndorfer, Sabine Metz, Karin Streicher, Clemens Steinwender, Karim Saleh, Andreas Lueger, Petra Fladerer, Eiko Meister, Heinz Drexel, Alexandra Schuler, Susanne Waeger, Karl-Martin Ebner, Christine Heinzle, Arthur Mader, Peter Schwerzler, Berta Patsch, Abdurahman Said, Claudia Stoeckloecker, Daniela Zanolin, Jutta Bergler-Klein, Ljubica Mandic, Mariann Gyöngyösi, Neraida Cene, Zsuzsanna Szankai, Abelina Zimba.
Denmark: Henrik Nielsen, Bjerre Flemming, Michaelsen Michaelsen, Elisa Stokholm, Katja Holm, Charlotte Schmidt Skov, Pauline Gøgsig Johansen, Soren shjortshoj, Thomas Melchior, Ole Dyg Pedersen, Sanne Heinsvig, Inge Larsen, Vibeke Perret-Gentil, Thomas Wagner Nielsen, Axel Brandes, Marianne Jensen, Ida Rosenlund, Liv Gøtzsche, Heidi Munk Andersen.
Germany: Andreas Götte, Matthias Hammwohner, Britta Möehring, Jutta Schaertl, Daniel Steven, Iris Berg, Alexandra Kuehn, Hannes Reuter, Elena Terentieva, Christian Loges, Christine Lindner, Hendrik Bonnemeier, Christan Wulff, Thomas Demming, Svenja Gediehn, Johanna Parlitz, Wilhelm Haverkamp, Buehner Kathrin, Hubert Katja, Iacovella Ines, Bernhard Korbmacher, Marc Thone, Hannan Dalyanoglu, Da Un Chung, Naujoks Angela, Dirk Weismann, Björn Lengenfelder, Jan Becher, Klaus Meyer, Irina Turkin, Sebastian Maier, Marcus Koller, Alban Glaser, Lisa Gebele, Jale Goezuebueyuek, Ralph Hampe, Barbara Ruemmler, Hagen Schrötter, Manja Hubald, Cornelia Fritz, Martin Domhardt, Kathrin Haacke, Nicole Schmiedehausen, Ruth Strasser, Kristof Graf, Lidia Fischer, Roland Thieme, Karlheinz Seidl, Martin Kulzer, Monika Zackel, Gerian Grönefeld, Christina Paitazoglou, Simone Müller, ThoraBotschafter Britta Goldmann, Andrea Moeller, Sindy Bartel, Joern Schmitt, Damir Erkapic, Gabriele Hellwig-Bahavar, Ritvan Chasan, Christopher Gemein, Victoria Johnson, Christiane Kelm, Kay Weipert, Johannes Brachmann, Michael Held, Andrea Höhn, Ute Goebel, Andrea Linss, Swetlana Rube, Ahmed Saleh, Steffen Schnupp, Yeong-Hoon Choi, Vera Wolf, Andrea Plate, Anton Sabashnikov, Antje-Christin Deppe, Petra Krause.
Spain: Ignacio Fernandez Lozano, Manuel Sánchez, Francisco Hernández, Alfonso Martin Martinez, Pedro Vazquez, Esther Alvarez, Pascual Lopez, Raquel Torres, José Carbajosa Dalmau, Laura Parades, Nestor Hernandez, Inmaculada Jimenez Ruiz, Ana Maria Lopez, Luis López-Andujar, Alexandre Noguera, Francisco Roman Cerdan, Carmen del Arco Galan, Daniel Afonso, Raquel Caminero, Manual Lunquera, Monica Negro, Cristina Santiago, Nestor Villalba, José Manuel Garrido Castilla, Roberto Martinez Asenjo, Elena Mejia Martinez, José Luís Merino Llorens, Maria Jesus Diaz-Pintado, Jorge Alejandro Figueroa, Alberto Borobia Perez, Sergio Castrejon Castrejon, David Filgueiras Rama, Manuel Quintana Diaz, Maria Angelica Ribera Nunez, Miguel Angel Ramirez Marrero, Antonio Martin, José Miguel Ormaetxe Merodio, Mercedes Varona Peinador, Maria Fe Arcocha, Larraitz Gaztañaga, F. Xavier Palom Rico, Javier Jacob Rodriguez, Pascual Piñera Salmeron, Juan Cosin Sales, Isabel Navarro, Francisco Buendia Funetes.
Sweden: Henrik Wallentin, Kerstin Roos, Arash Mokhtari, Hans-Jörgen Nilsson, Peter Vasko, Terese Nyström, Martina Gustensson, Susanne Johansson, Inga Uggeldahl, Göran Andersson, Olle Bergström, Thomas Aronsson, Mehmet Hamid, Kerstin Giocondi, Deborah Svanerö, Qassim Awad, Tord Juhlin, Hjördis Jernhed, Stefan Berglund, Magnus Forsgren, Michael Guggi, Pär-Lennart Agren, Kristina Eriksson, Kristina Karlsson, Per Blomström, Alejandro Utreras, Caroline Lundgren, Maria Soderlund, Frederik Buijs, Solveig Östberg, Johan-Emil Bager, Ingrid Hendequist, Maria Just, Siv Heden, Liselott Lisjo, Chrichan Mansson, Helen Svanstrom, Mikael Dellborg, Helena Dellborg, Gorel Hultsberg Olsson, Linus Hansson.
Finland: Hannu Sulonen, Hanna Suurmunne, Anna Petrovskaja, Johanna Markkanen, Juha Hartikainen, Lari Kujanen, Antti Heikkola, Hanna Pohjantahti MaarooS.
Safety Review Committee
Lars Kober, MD, Samuel Lévy, MD (Chair), Cristina Varas-Lorenzo, MD, MSc, PhD, Manel Pladevall-Vila MD, MS.
The authors wish to thank Nathalie Dunkel and her team for their help in accessing the data of SPECTRUM and providing information and documents upon our request.
The authors would like to thank all of the investigators involved in the SPECTRUM study.
Funding
This study was funded by Correvio International Sàrl, Geneva, Switzerland. The authors received no funding for their participation in this manuscript.
Data Availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Compliance with Ethical Standards
Conflict of Interest
Professor Levy reports no conflicts of interest. Professor Hartikainen has been an investigator in studies sponsored by AstraZeneca, Biosense Webster, Boehringer Ingelheim, Correvio International Sàrl, Medtronic, and St. Jude Medical. Dr. Ritz is an employee of Correvio International Sàrl, Geneva, Switzerland. Dr. Juhlin has received speaker honoraria from Correvio International Sàrl. Professor Domanovits reports no conflicts of interest. Dr. Carbajosa-Dalmau has received speaker honoraria from Correvio International Sàrl.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. | FLECAINIDE, VERNAKALANT | DrugsGivenReaction | CC BY | 33206300 | 19,695,564 | 2021-04 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Sinus arrest'. | Vernakalant for Rapid Cardioversion of Recent-Onset Atrial Fibrillation: Results from the SPECTRUM Study.
Rapid restoration of sinus rhythm using pharmacological cardioversion is commonly indicated in patients with symptomatic recent-onset atrial fibrillation (AF). The objectives of this large, international, multicenter observational study were to determine the safety and effectiveness of intravenous (IV) vernakalant for conversion of AF to sinus rhythm in daily practice.
Consenting patients with symptomatic recent-onset AF (< 7 days) treated with IV vernakalant were enrolled and followed up to 24 h after the last infusion or until discharge, in order to determine the incidence of predefined serious adverse events (SAEs) and other observed SAEs and evaluate the conversion rate within the first 90 min. Overall, 2009 treatment episodes in 1778 patients were analyzed. The age of patients was 62.3 ± 13.0 years (mean ± standard deviation). Median AF duration before treatment was 11.1 h (IQR 5.4-27.0 h). A total of 28 SAEs occurred in 26 patients including 19 predefined SAEs, i.e., sinus arrest (n = 4, 0.2%), significant bradycardia (n = 11, 0.5%), significant hypotension (n = 2, 0.1%), and atrial flutter with 1:1 conduction (n = 2, 0.1%). There were no cases of sustained ventricular arrhythmias or deaths. All patients who experienced SAEs recovered fully (n = 25) or with sequelae (n = 1). Conversion rate to sinus rhythm was 70.2%, within a median of 12 min (IQR 8.0-28.0 min).
This large multicenter, international observational study confirms the good safety profile and the high effectiveness of vernakalant for the rapid cardioversion of recent-onset AF in daily hospital practice.
Introduction and Purpose of the Study
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, with an estimated 33.5 million people affected worldwide [1]. One in four adults over 55 years of age in Europe and the USA develop AF, with greater prevalence in older populations [1, 2]. Patients with AF are at increased risk of stroke and heart failure [3, 4]. A significant number of patients with recent-onset AF seen in the emergency departments (EDs) undergo commonly in Europe pharmacological cardioversion.
Vernakalant is a partial atrial-selective antiarrhythmic agent by its action through IKur and IKACh channel inhibition [5]. However, it has a modest effect on the ventricle via Ina and IKr channels resulting in a limited effect on ventricular repolarization (QT interval) [5]. Vernakalant is contra-indicated in patients with prolonged QT interval.
Intravenous vernakalant has been approved by the European Medicine Agency [2010] for the rapid conversion of recent-onset AF [6]. To date, a number of studies have shown vernakalant to be well tolerated and effective for cardioversion of AF [7–18].
The FDA (Food and Drug Administration agency) decided in 2008 and in December 2019 not to approve to market vernakalant in the USA for safety concerns. In 2010, the EMA requested a post-authorization safety study to better define the risk benefit ratio in routine clinical practice. The objectives of SPECTRUM (Surveillance of Pharmacologic thErapy for Cardioversion in aTrial fibrillation Registry Using IV treatMent) (NCT01370629 and EUPAS2078) study were to assess the rates of adverse events and to estimate the effectiveness of the drug in a large cohort of patients with recent-onset AF.
Methods
Definitions
Recent-onset AF was defined as symptomatic episode within 7 days that will be undergoing cardioversion taking into account that about 70% of patients with symptomatic AF < 72 h were reported to convert spontaneously [19]. Beyond 7 days, AF is likely to persist and the chances of pharmacological cardioversion to be successful become low. Hypertension was reported when documented on the medical record or the patient report. Coronary artery disease (CAD) was diagnosed when the patient had a documented history of CAD and/or a history of coronary revascularization.
Patients and Procedures
Adult patients (≥ 18 years) with recent-onset AF occurring between September 1, 2011 and April 11, 2018 who received vernakalant for cardioversion were eligible for inclusion in this international, multicenter, observational, post-authorization study. Fifty-five hospitals in Austria, Denmark, Germany, Spain, Sweden, and Finland participated in the study, 53 of which enrolled patients. While administration of vernakalant was at the discretion of the treating physician, consecutively treated patients were enrolled and reasons for non-participation were documented. A preinfusion checklist and healthcare provider educational card were implemented during the study period to assist in identifying patients for treatment consistent with the approved indications and contraindications.
Patients were required to give informed consent for participation in the study and could be enrolled more than once if they presented on multiple occasions for AF episodes. Patients who had participated in an investigational drug/device clinical trial within 30 days prior to enrollment were not eligible. In order to enhance enrollment and reach the EMA required target of 2000 episodes, a protocol amendment was made in September 2016, which permitted retrospective inclusion of patients who had received vernakalant between April 2013 and the end of the study, provided that they fulfilled the established eligibility criteria. For prospectively enrolled patients, data were collected from both medical records and supplemental standardized data collection forms. For retrospectively enrolled patients, only medical records were available. The study period comprised a baseline assessment and up to 24-h follow-up after completion of the last infusion or until discharge. This study was mandated and approved by the European Committee for Medicinal Products for Human Use. The study protocol was approved by the appropriate local research ethics committees for all participating centers, and the study was conducted in accordance with applicable national and local regulations/guidelines, accepted standards for Good Clinical Practice, Guidelines for Good Pharmacoepidemiology Practices, and the Declaration of Helsinki [20].
Study Objectives and Endpoints
The primary objectives of the study was to estimate the incidence of clinically predefined serious adverse events (SAEs), i.e., significant hypotension (systolic blood pressure < 90 mmHg or requiring vasopressors); sustained (> 30 s) ventricular arrhythmias, Torsade de Pointes (>10 s) or ventricular fibrillation, atrial flutter with 1:1 conduction, bradycardia requiring temporary electrical pacing, or sinus arrest (> 3 s). Definition of these predefined SAEs was based on events from previous controlled studies on IV vernakalant [7, 8, 11, 12] and from the reported adverse events (AEs) on other antiarrhythmic agents. Secondary objectives included the rates of all other SAEs. Each SAE was reviewed and adjudicated by an independent expert Safety Review Committee (SRC). This study had also the objective to determine the conversion rate to sinus rhythm in a large population of patients outside the setting of controlled clinical trials.
The duration of the index AF episode was calculated as the time between the patient-reported time of symptom onset and the start of the first vernakalant infusion. Successful cardioversion was defined as conversion to sinus rhythm within 90 min of the start of vernakalant infusion. Conversion rate was calculated in all patients, as well as in an effectiveness population excluding all treatment episodes in which patients received another therapy for cardioversion within 90 min of the start of vernakalant administration (e.g., electrical or pharmacological cardioversion). Vernakalant is recommended to be administered in a step-dose fashion. Each treatment episode can comprise up to two infusions, separated by a 15-min observation period. The recommended doses for the first and second infusions are 3.0 mg/kg and 2.0 mg/kg, respectively, each administered over 10 min. For patients above 113 kg, vernakalant has a fixed initial dose of 339 mg. If conversion to sinus rhythm does not occur within 15 min after the end of the initial infusion, a second 10-min infusion of 226 mg may be administered.
Statistics and Analyses
A target sample size of 2000 vernakalant IV treatment episodes was chosen to allow adequate statistical precision, as expressed by a two-sided 95% confidence limit. Enrollment per site was capped at 10% of the total study population and 40% per country to minimize any potential bias in practice patterns. Categorical variable frequency, along with 95% confidence intervals (CIs), was determined for the summed treatment episodes. Continuous variables were summarized using descriptive statistics. Data were analyzed based on enrollment method (prospective vs retrospective) and reported as stratified and unstratified CIs. All analyses were performed using Statistical Analysis System v9.2, or later, software.
Results
Study Population
A total of 1778 patients who presented with 2009 treatment episodes were included: 1580 episodes were in prospectively enrolled patients and 429 in retrospectively enrolled patients (Table 1). The majority of patients were treated in the ED for 1289 (64.1%) AF episodes and 563 (28.0%) AF episodes in the coronary or intensive care units, with the remainder 157 (7.8%) episodes being treated in other hospital settings. As seen in Fig. 1, the main reason for non-inclusion in the study was lack of informed consent. In 1905 (94.7%) AF episodes, vernakalant was administered to non-surgery patients, and in 104 (5.2%) to post-cardiac surgery patients. The later are among the prospectively included patients. The mean age of the overall patient population at time of treatment was 62.3 ± 13.0 years (mean ± standard deviation [SD]), ranging from 18.0 to 94.0 years, and 1222 (60.8%) episodes occurred in men (Table 1). At baseline, systolic blood pressure (BP) was 132.5 ± 19.5 mmHg and heart rate (HR) was 112.9 ± 25.5/min (mean ± SD). The median duration of AF episode prior to treatment was 11.1 (5.4–27.0) hours (median [interquartile range, IQR]). In 88.9% of episodes, the patients were treated within 48 h of the onset of symptoms, and in 72.5% within 24 h. Duration of AF before treatment in 104 post-cardiac surgery patients was shorter than in the overall population, with 3.6 h (range 0.8–15.4) (median [IQR]). Baseline demographics and characteristics were similar between patients enrolled prospectively and retrospectively. Total length of ED stay was 7.5 (5.0–13.5) hours (median [IQR]). Only 167 (13.0%) of patients initially managed in the ED were in hospital for 24 h or longer. The number of vernakalant infusions was available in 1990 patients. Of these, 1201 (60.4%) received one vernakalant infusion and 789 (39.6) received a total of 2 infusions.Table 1 Clinical characteristics of patients
Total Prospective Retrospective
No. of patients 2009 1580 429
Age (years) mean ± SD 62.3 ± 13.0 61.9 ± 13.5 63.6 ± 11.2
Range (years) 18.0–94 18–93 30–94
Male, n (%) 1222 (60.8) 998 (63.2) 224 (52.2)
Body weight (kg) mean ± SD 84.1 ± 16.5 84.3 ± 16.5 83.1 (16.9)
Range (kg) 45.0–189.0 45.0–189.0 45.0–165.0
Body mass index (kg/m2) 27.8 ± 4.9 27.7 ± 4.8 28.2 ± 5.1
Associated conditions, n (%)
Hypertension 1103 (54.9) 884 (55.9) 219 (51.0)
Coronary artery disease 118 (5.9) 82 (5.2) 36 (8.4)
Cardiomyopathy 33 (1.6) 31 (2.0) 2 (0.5%)
Heart failure (history) 63 (3.1) 59 (3.7) 4 (0.9)
Diabetes 199 (9.9) 165 (10.4) 34 (7.9)
Stroke (history) 91 (4.5) 68 (4.3) 23 (5.4)
Pacemaker/ICD 36 (1.8) 24 (1.5) 12 (2.8)
Type of AF episode
First detected 477 (23.7) 393 (24.9) 84 (19.6)
Previous history of AF 1458 (72.6) 1115 (70.6) 343 (80.0)
Onset unknown/not assessed 5 (0.2) 3 (0.2) 2 (0.5)
Post-surgery 69 (3.4) 69 (4.4) 0 (0.0)
Symptoms on admission, n (%)
Palpitations, irregular heart beat 1749 (87.1) 1337 (84.6) 412 (96.0)
Dyspnea or shortness of breath 352 (17.5) 306 (19.4) 46 (10.7)
Dizziness, light-headedness 320 (15.9) 251 (15.9) 69 (16.1)
Chest pain 271 (13.5) 220 (13.9) 51 (11.9)
Syncope, near syncope 61 (3.0) 53 (3.4) 8 (1.9)
Duration of the index episode
Less than 24 h, n (%) 1438 (72.5) 1107 (70.2) 331 (81.5)
24–48 h, n (%) 347 (17.5) 288 (18.3) 59 (14.5)
More than 48 h 199 (10.0) 183 (11.6) 16 (3.9)
Mean duration ± SD (h) 23.2 ± 44.9 24.9 ± 45.8 16.8 ± 40.6
Median (IQR 25–75) (h) 11.1 (5.44–27.03) 11.9 (5.8–29.7) 8.2 (4.8–18.3)
Antiarrhythmic agents, n (%)
Betablockers 1055 (52.5) 800 (50.6) 255 (59.4)
Calcium channels blockers 22 (1.1) 20 (1.3) 2 (0.5)
Class I agents* 85 (4.2) 71 (4.5) 14 (3.3)
Class III agents* 98 (4.9) 89 (5.6) 9 (2.1)
Digitalis glycosides 22 (1.1) 18 (1.1) 4 (0.9)
*Using the Vaughan-Williams classification
Fig. 1 Study flow chart. Flow chart showing patient enrollment in the SPECTRUM study. The term patient here refers to individual treatment episodes (asterisk). Owing to lack of informed consent (n = 500) (dagger). Other reasons included patient enrollment in an investigational drug trial in the past 30 days, spontaneous conversion to sinus rhythm, ejection fraction 30–35%, electrical cardioversion preferred, missing information regarding start of atrial fibrillation, inclusion criteria not met, other, or no reason provided or known. Source data could not be verified to confirm that vernakalant IV was administered (double dagger). Spontaneous conversion to sinus rhythm before vernakalant IV administration (section sign). Patient decision and lack of follow-up after cardioversion in one case each (double vertical line). IV intravenous
Predefined Serious Adverse Events and Other Adverse Events
No deaths were recorded in our study. Nineteen predefined SAEs were reported during or after 17 treatment episodes (cumulative incidence 0.8%; CI 0.5–1.4%) (Table 2). Eighteen of the 19 events occurred within 2 h from the start of infusion. The remaining event was an episode of atrial flutter with 1:1 conduction which occurred 3.1 h after drug infusion and was terminated by electrical shock. Symptomatic bradycardia was the most common event occurring in 11 (0.5%; CI 0.4–1.2%) episodes (Table 2). Conversion to sinus rhythm occurred in 10 of these cases. A pause described as sinus arrest preceding the restoration of sinus rhythm occurred in 4 patients. In 2 patients, sinus arrest was associated with sinus bradycardia. In all bradycardia and sinus arrest cases, the vernakalant infusion was immediately discontinued. One of these 4 sinus arrests occurred in a 66-year-old man, sportive cyclist with no history of heart disease, admitted for a first episode of AF with a mean ventricular response of 95 beats/min. He received 300 mg orally of flecainide which failed to restore sinus rhythm. The treating physician decided 4 h later, to administer IV vernakalant. At the end of the infusion, a pause of 6 s, with a brief dizziness, occurred and resolved spontaneously, followed by a normal sinus rhythm with a HR of 47 beats/min which was patient usual HR and a BP of 120/85 mmHg. This event was considered a SAE although there was probably an interaction between oral flecainide still active and vernakalant in this event. One of the bradycardia events occurred in a retrospectively enrolled 69-year-old woman on bisoprolol with a history of hypertension and CAD, who developed 8 min after the second infusion of vernakalant a sinus bradycardia which rapidly resolved with IV atropine. Two bradycardia episodes occurred in post-cardiac surgery patients requiring temporary electrical pacing through the electrodes left in place by the surgeon. Both patients converted to sinus rhythm. None of the non-surgery patients required temporary electrical pacing. Significant hypotension occurred on two (0.1%; CI < 0.1–0.4%) occasions, associated with sinus bradycardia in both instances. Both events resolved with intravenous atropine and fluid. There were two cases of atrial flutter with 1:1 ventricular conduction terminated with electrical shock whereas no cases of sustained ventricular tachycardia (VT), ventricular fibrillation, or Torsade de Pointes were observed. In addition to the predefined SAEs, there were 9 other SAEs, one of which occurred in a retrospectively enrolled patient (Table 2). They included two instances of hypotension not requiring vasopressor agents, 2 non-sustained VT which deserve special attention. The first non-sustained VT occurred in a 48-year-old man with asthma admitted with fever, palpitations, dyspnea, and first episode of AF with a ventricular rate of 144 bpm. During vernakalant infusion, 5 beats of non-sustained VT was observed. Among the tests done, coronary angiography was reported as normal. The same run of 5 beats of non-sustained VT was observed 20 h after infusion (next day) making the causal effect of vernakalant unlikely. The other event occurred in a 57-year-old patient with a 6-year history of recurrent symptomatic AF and arterial hypertension with left ventricular hypertrophy. He was admitted with palpitations, irregular heartbeats, and dizziness. He was on dronedarone, and ECG showed AF with a ventricular rate of 135 bpm. During infusion of vernakalant, he had 6 s of non-sustained VT observed on the monitor and was given 5 mg of bisoprolol which reduced the heart rate to 120 beats/min and relieved patient symptoms. The Safety Review Committee considered that in the first case, the wide QRS complexes were due to aberrant conduction during rapid AF (Ashman phenomenon). Among the non-predefined SAEs, one supraventricular tachycardia (120 beats/min) and a single report each of angina pectoris, pericardial effusion, transient visual disturbance, and vernakalant overdose (Table 2). A total of 188 non-serious AEs were reported, the most common of which were dysgeusia (n = 35) and sneezing (n = 27). All patients with vernakalant-related AEs recovered without sequelae. All but 6 of the 28 SAEs were considered by the investigators and the SRC to be related to vernakalant administration.Table 2 Adverse events in 2009 episodes during treatment and observation periods
Event type Number of events Incidence (95% CI) Considered drug-related, n (%)
All SAEs 28 1.3% (0.8–1.9) 22 (78.6)
Predefined SAEs 19 0.8% (0.5–1.4) 18 (94.7)
Significant hypotension 2 0.1% (< 0.1–0.4) 2 (100.0)
Bradycardiaα 11 0.5% (0.3–10) 10 (93.3)
Sinus arrest (> 3 s)β 4 0.2% (< 0.1–0.4) 4 (100.0)
Atrial flutter with 1: 1 AV conduction 2 0.1% (0.1–0.4) 2 (100.0)
Ventricular tachycardia γ 0 0 0 (0.0)
Other than predefined SAEs 9 0.45% 5 (55.6)
Hypotension 2 0.1% 1 (50.0)
Supraventricular tachycardiaδ 1 < 0.1% 1 (100.0)
Non-sustained ventricular tachycardiaε 2 < 0.1% 1 (50.0)
Angina pectoris 1 (< 0.1) < 0.1% 0 (0.0)
Pericardial effusion 1 (< 0.1) < 0.1% 0 (0.0)
Visual disturbance 1 (< 0.1) < 0.1% 0 (0.0)
Vernakalant overdoseζ 1 (< 0.1) < 0.1% 1 (100.0)
αNine cases of sinus bradycardia and 2 reported as significant bradycardia
βOne patient had both sinus arrest followed by sinus bradycardia
γOne event reclassified as atrial flutter with 1:1 conduction
δAtrial arrhythmia other than atrial flutter
εSee text, exceeding 5% of the weight-based dosing recommendation. In this case, the administered dose was 51% in excess of the recommended dose
Rates of Conversion to Sinus Rhythm
Overall, conversion to sinus rhythm at any time following vernakalant infusion occurred in 1448 out of 2009 (72.1%) treatment episodes. Successful cardioversion was recorded in 70.2% (CI 68.1–72.2%) of the 1936 episodes of the effectiveness population excluding those in which either electrical cardioversion (n = 68) or an additional intravenous Class I/III antiarrhythmic drug (n = 6) was given within 90 min of infusion initiation. The rate of cardioversion was similar between the 1107 of 1580 (70.1%) episodes included prospectively and the 297 of 421 (70.5%) episodes of retrospectively enrolled patients. Successful cardioversion of AF was reported in 68 of 104 (65.4%) of treatment episodes in the post-cardiac surgery patients. Time to cardioversion was recorded in 1413 of 1448 episodes with successful conversion to sinus rhythm. The median time to conversion was 12.0 (8.0–28.0) minutes (median [IQR]) Fig. 2). One thousand one hundred eight of 1413 (78.4%) successful cardioversions were treated with only one drug infusion. The percentage of successful cardioversion was 70.1% in the prospective patients and 70.5% in the retrospective patients. The median hospital stay time in those treated in the ED was 7.5 h allowing patient discharges when their condition was clinically stable.Fig. 2 Time to conversion to sinus rhythm. Time to conversion to sinus rhythm with vernakalant IV in the effectiveness analysis population (N = 1936). Time to conversion was not recorded in 29 treatment episodes in which patients converted to sinus rhythm; these episodes are not displayed on the graph but are taken into account for the proportion calculation. IV intravenous
Anticoagulation
About a quarter of patients presenting with recent-onset AF at baseline were on vitamin K antagonists or direct oral anticoagulants. Investigators respected current guidelines [3] on anticoagulation both peri-procedurally and after hospital discharge.
Discussion
The SPECTRUM study included a large real-world patient population of 1778 patients with 2009 recent-onset AF episodes in whom pharmacological cardioversion was performed with vernakalant. About 70% of patients were cardioverted within 12 min from onset of infusion and 11 h from the AF onset. Our findings confirm the safety and efficacy of vernakalant reported in previous studies [7–18, 21–25] and extend their consistency to routine hospital use in large populations. To our knowledge, the present study provides the largest series of patients with recent-onset AF undergoing pharmacological cardioversion with a specific antiarrhythmic agent. The safety was the main objective of this study. We found the incidence of both predefined and other SAEs to be lower than expected. There were no death and no sustained ventricular arrhythmia. Overall, 28 SAEs (1.3%) were recorded. The majority of patients were AF treated in ED and intensive care units.
Pharmacological cardioversion is frequently indicated as part of a rhythm control strategy or as a tool to control patient symptoms and avoid hospitalization in clinically stable condition [25, 26]. It is often preferred to electrical cardioversion in patients with hemodynamically stable condition as it does not require general anesthesia or sedation. Among agents currently available for rapid termination of recent-onset AF, vernakalant represents an option [3]. However, there has been to our knowledge, no large study exploring the safety of vernakalant in daily practice.
There is no universal definition for recent-onset AF. In current literature, the duration limits of AF episodes range from < 24 [27] to < 48 h and even < 7 days [28, 29]. The prevalence of recent-onset AF among all AF subsets varies from 11% when restricted to the first detected episode (new onset) [30] to 26% [31]. The characteristics of patients were similar to those of other AF cohorts [31, 32].
As with electrical cardioversion, pharmacological cardioversion can be associated with post-cardioversion bradyarrhythmias, often unmasking pre-existing sinus node dysfunction or atrioventricular conduction abnormalities and can result in ventricular escape rhythms or prolonged ventricular pauses. Of interest, these pauses were first reported by Lown [33], following electrical cardioversion as the possible reflect of sinus dysfunction. Another possible mechanism for these sinus arrests is right atrial stunning [34]. The cumulative incidences of bradycardia (0.5%), sinus arrest (0.2%), and hypotension (0.1%) observed in this study were also low. The incidence of atrial flutter with 1:1 conduction was lower than that reported with oral Class Ic antiarrhythmics, such as flecainide or propafenone. The “pill in the pocket” approach requires initiation of therapy in hospital to verify its safety [35]. No cases of Torsade de Pointes or sustained VT was observed, which is in line with the low risk of ventricular proarrhythmia associated with vernakalant owing to its electrophysiological properties [5]. This contrasts with the reported incidence of Torsade de Pointes [24, 36, 37] in patients with AF/atrial flutter of 4.3% with intravenous ibutilide in the report of Kowey et al. including 1.7% of which required cardioversion [36]. Of note, all but one of the predefined SAEs in this study occurred within 2 h of the start of infusion. As aforementioned, the remaining patient had atrial flutter with 1:1 ventricular conduction which occurred 3.1 h following infusion initiation, indicating that close cardiac monitoring should be available during and after treatment in some patients.
Conversion to sinus rhythm with vernakalant was rapid (median time of 12.0 min) similar to what was previously reported [9–18]. The conversion median time of ibutilide was significantly longer than that of vernakalant (26 min versus 10 min, P = 0.01) in a randomized comparison [18]. Furthermore, in this particularly large real-world study, the median duration of AF episode was short (11.1 h) as there is important evidence, and relevant guidelines [3] suggesting that prompt cardioversion could be associated with benefits in terms of lower risk of thromboembolic events [4, 38, 39].
Although the baseline characteristics of the study population were consistent with AF population-based studies [31, 32] and clinical studies with vernakalant, the conversion rate was higher than that observed in recent review and meta-analysis (~ 50%) [22–25]. This seems likely to be due to patients being treated soon after symptom onset in European clinical practice. Other recent but smaller observational studies [13–15, 17, 18], which collectively included almost 1300 patients, have found similarly high conversion rates (65–86%) when vernakalant was administered soon after the onset of AF, particularly within the first 48 h [15, 17, 21]. Vernakalant has also been shown to induce a higher rate of cardioversion compared with flecainide (67% vs 46%) in a non-randomized cohort study [21]. Similarly, in randomized studies, vernakalant was more effective than amiodarone [12] (52% vs 5%; after 90 min) and ibutilide [18, 24] (69% vs 43%; within 90 min). The SPECTRUM results are consistent with previous reports that vernakalant is safe and effective for the rapid cardioversion of recent-onset AF and extends them to daily practice.
Owing to the rapid time to conversion with vernakalant, the median hospital stay time for those treated in the ED was 7.5 h. This is encouraging given that a study in France reported that hospitalization constitutes 60% of the cost of care for patients with AF [40].
Study Limitations
This multicenter international study was observational as the main objective was to determine the safety of vernakalant as used in daily hospital practice without interfering on the management of recent-onset AF by the treating physician. For these reasons, the adverse events were expected to be higher in an “uncontrolled” setting with no guidance on patient selection than those reported in controlled studies with strict protocols. In fact, SAEs were low in this study. Data collection for prospectively enrolled patients was comprehensive owing to use of both study-specific tools and medical records. However, for retrospectively enrolled patients, it was not possible to routinely collect all data of interest in a standardized manner. Nevertheless, baseline characteristics, medical histories, and SAEs in the retrospective cohort were similar to those in the prospective cohort, supporting the use of a retrospective analysis.
Conclusions
The results of this large multicenter study showed that vernakalant has a good safety profile and is effective in enabling rapid cardioversion in clinical practice. Moreover, the rates of serious complications were lower than those observed in early trials reflecting appropriate patient selection in clinical practice. In conclusion, vernakalant provides a rapid and effective means of pharmacological conversion in patients with recent-onset AF undergoing cardioversion undergoing cardioversion in daily hospital practice.
Appendix. List of Investigators
Austria: Michael Joannidis, Klemens Zotter Frank Hartig, Anton Sandhofer, Alois Süssenbacher, Bernd Eber, Elisabeth Lassnig, Ulrike Pfeifenberger, Michaela Steiner, Hans Domanovits, Alexander Simon, Alexander Spiel, Jan Niederdockl, Nikola Schuetz, Daniel Wehinger, Franz Xaver Roithinger, Isabella M. von Katzler, Katharina Bichler, Robert Schoenbauer, Lukas Fiedler, Michael Pfeffer, Markus Peck, Florian Benische, Michael Hackl, Susane Demschar, Astrid Ebner, Melanie Eder, Rainer Huditz, Arnulf Isak, Michael Moser, Georg Pinter, Thomas Singer, Claudia Waldhauser, Helmut Pürerfellner, Martin Martinek, Sandra Muellner, Andrea Ploechl, Tanja Koppler, Elisabeth Sigmund, Michael Derndorfer, Sabine Metz, Karin Streicher, Clemens Steinwender, Karim Saleh, Andreas Lueger, Petra Fladerer, Eiko Meister, Heinz Drexel, Alexandra Schuler, Susanne Waeger, Karl-Martin Ebner, Christine Heinzle, Arthur Mader, Peter Schwerzler, Berta Patsch, Abdurahman Said, Claudia Stoeckloecker, Daniela Zanolin, Jutta Bergler-Klein, Ljubica Mandic, Mariann Gyöngyösi, Neraida Cene, Zsuzsanna Szankai, Abelina Zimba.
Denmark: Henrik Nielsen, Bjerre Flemming, Michaelsen Michaelsen, Elisa Stokholm, Katja Holm, Charlotte Schmidt Skov, Pauline Gøgsig Johansen, Soren shjortshoj, Thomas Melchior, Ole Dyg Pedersen, Sanne Heinsvig, Inge Larsen, Vibeke Perret-Gentil, Thomas Wagner Nielsen, Axel Brandes, Marianne Jensen, Ida Rosenlund, Liv Gøtzsche, Heidi Munk Andersen.
Germany: Andreas Götte, Matthias Hammwohner, Britta Möehring, Jutta Schaertl, Daniel Steven, Iris Berg, Alexandra Kuehn, Hannes Reuter, Elena Terentieva, Christian Loges, Christine Lindner, Hendrik Bonnemeier, Christan Wulff, Thomas Demming, Svenja Gediehn, Johanna Parlitz, Wilhelm Haverkamp, Buehner Kathrin, Hubert Katja, Iacovella Ines, Bernhard Korbmacher, Marc Thone, Hannan Dalyanoglu, Da Un Chung, Naujoks Angela, Dirk Weismann, Björn Lengenfelder, Jan Becher, Klaus Meyer, Irina Turkin, Sebastian Maier, Marcus Koller, Alban Glaser, Lisa Gebele, Jale Goezuebueyuek, Ralph Hampe, Barbara Ruemmler, Hagen Schrötter, Manja Hubald, Cornelia Fritz, Martin Domhardt, Kathrin Haacke, Nicole Schmiedehausen, Ruth Strasser, Kristof Graf, Lidia Fischer, Roland Thieme, Karlheinz Seidl, Martin Kulzer, Monika Zackel, Gerian Grönefeld, Christina Paitazoglou, Simone Müller, ThoraBotschafter Britta Goldmann, Andrea Moeller, Sindy Bartel, Joern Schmitt, Damir Erkapic, Gabriele Hellwig-Bahavar, Ritvan Chasan, Christopher Gemein, Victoria Johnson, Christiane Kelm, Kay Weipert, Johannes Brachmann, Michael Held, Andrea Höhn, Ute Goebel, Andrea Linss, Swetlana Rube, Ahmed Saleh, Steffen Schnupp, Yeong-Hoon Choi, Vera Wolf, Andrea Plate, Anton Sabashnikov, Antje-Christin Deppe, Petra Krause.
Spain: Ignacio Fernandez Lozano, Manuel Sánchez, Francisco Hernández, Alfonso Martin Martinez, Pedro Vazquez, Esther Alvarez, Pascual Lopez, Raquel Torres, José Carbajosa Dalmau, Laura Parades, Nestor Hernandez, Inmaculada Jimenez Ruiz, Ana Maria Lopez, Luis López-Andujar, Alexandre Noguera, Francisco Roman Cerdan, Carmen del Arco Galan, Daniel Afonso, Raquel Caminero, Manual Lunquera, Monica Negro, Cristina Santiago, Nestor Villalba, José Manuel Garrido Castilla, Roberto Martinez Asenjo, Elena Mejia Martinez, José Luís Merino Llorens, Maria Jesus Diaz-Pintado, Jorge Alejandro Figueroa, Alberto Borobia Perez, Sergio Castrejon Castrejon, David Filgueiras Rama, Manuel Quintana Diaz, Maria Angelica Ribera Nunez, Miguel Angel Ramirez Marrero, Antonio Martin, José Miguel Ormaetxe Merodio, Mercedes Varona Peinador, Maria Fe Arcocha, Larraitz Gaztañaga, F. Xavier Palom Rico, Javier Jacob Rodriguez, Pascual Piñera Salmeron, Juan Cosin Sales, Isabel Navarro, Francisco Buendia Funetes.
Sweden: Henrik Wallentin, Kerstin Roos, Arash Mokhtari, Hans-Jörgen Nilsson, Peter Vasko, Terese Nyström, Martina Gustensson, Susanne Johansson, Inga Uggeldahl, Göran Andersson, Olle Bergström, Thomas Aronsson, Mehmet Hamid, Kerstin Giocondi, Deborah Svanerö, Qassim Awad, Tord Juhlin, Hjördis Jernhed, Stefan Berglund, Magnus Forsgren, Michael Guggi, Pär-Lennart Agren, Kristina Eriksson, Kristina Karlsson, Per Blomström, Alejandro Utreras, Caroline Lundgren, Maria Soderlund, Frederik Buijs, Solveig Östberg, Johan-Emil Bager, Ingrid Hendequist, Maria Just, Siv Heden, Liselott Lisjo, Chrichan Mansson, Helen Svanstrom, Mikael Dellborg, Helena Dellborg, Gorel Hultsberg Olsson, Linus Hansson.
Finland: Hannu Sulonen, Hanna Suurmunne, Anna Petrovskaja, Johanna Markkanen, Juha Hartikainen, Lari Kujanen, Antti Heikkola, Hanna Pohjantahti MaarooS.
Safety Review Committee
Lars Kober, MD, Samuel Lévy, MD (Chair), Cristina Varas-Lorenzo, MD, MSc, PhD, Manel Pladevall-Vila MD, MS.
The authors wish to thank Nathalie Dunkel and her team for their help in accessing the data of SPECTRUM and providing information and documents upon our request.
The authors would like to thank all of the investigators involved in the SPECTRUM study.
Funding
This study was funded by Correvio International Sàrl, Geneva, Switzerland. The authors received no funding for their participation in this manuscript.
Data Availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Compliance with Ethical Standards
Conflict of Interest
Professor Levy reports no conflicts of interest. Professor Hartikainen has been an investigator in studies sponsored by AstraZeneca, Biosense Webster, Boehringer Ingelheim, Correvio International Sàrl, Medtronic, and St. Jude Medical. Dr. Ritz is an employee of Correvio International Sàrl, Geneva, Switzerland. Dr. Juhlin has received speaker honoraria from Correvio International Sàrl. Professor Domanovits reports no conflicts of interest. Dr. Carbajosa-Dalmau has received speaker honoraria from Correvio International Sàrl.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. | FLECAINIDE, VERNAKALANT | DrugsGivenReaction | CC BY | 33206300 | 19,695,564 | 2021-04 |
Give an alphabetized list of all active substances of drugs taken by the patient who experienced 'Treatment failure'. | Vernakalant for Rapid Cardioversion of Recent-Onset Atrial Fibrillation: Results from the SPECTRUM Study.
Rapid restoration of sinus rhythm using pharmacological cardioversion is commonly indicated in patients with symptomatic recent-onset atrial fibrillation (AF). The objectives of this large, international, multicenter observational study were to determine the safety and effectiveness of intravenous (IV) vernakalant for conversion of AF to sinus rhythm in daily practice.
Consenting patients with symptomatic recent-onset AF (< 7 days) treated with IV vernakalant were enrolled and followed up to 24 h after the last infusion or until discharge, in order to determine the incidence of predefined serious adverse events (SAEs) and other observed SAEs and evaluate the conversion rate within the first 90 min. Overall, 2009 treatment episodes in 1778 patients were analyzed. The age of patients was 62.3 ± 13.0 years (mean ± standard deviation). Median AF duration before treatment was 11.1 h (IQR 5.4-27.0 h). A total of 28 SAEs occurred in 26 patients including 19 predefined SAEs, i.e., sinus arrest (n = 4, 0.2%), significant bradycardia (n = 11, 0.5%), significant hypotension (n = 2, 0.1%), and atrial flutter with 1:1 conduction (n = 2, 0.1%). There were no cases of sustained ventricular arrhythmias or deaths. All patients who experienced SAEs recovered fully (n = 25) or with sequelae (n = 1). Conversion rate to sinus rhythm was 70.2%, within a median of 12 min (IQR 8.0-28.0 min).
This large multicenter, international observational study confirms the good safety profile and the high effectiveness of vernakalant for the rapid cardioversion of recent-onset AF in daily hospital practice.
Introduction and Purpose of the Study
Atrial fibrillation (AF) is the most common sustained cardiac arrhythmia, with an estimated 33.5 million people affected worldwide [1]. One in four adults over 55 years of age in Europe and the USA develop AF, with greater prevalence in older populations [1, 2]. Patients with AF are at increased risk of stroke and heart failure [3, 4]. A significant number of patients with recent-onset AF seen in the emergency departments (EDs) undergo commonly in Europe pharmacological cardioversion.
Vernakalant is a partial atrial-selective antiarrhythmic agent by its action through IKur and IKACh channel inhibition [5]. However, it has a modest effect on the ventricle via Ina and IKr channels resulting in a limited effect on ventricular repolarization (QT interval) [5]. Vernakalant is contra-indicated in patients with prolonged QT interval.
Intravenous vernakalant has been approved by the European Medicine Agency [2010] for the rapid conversion of recent-onset AF [6]. To date, a number of studies have shown vernakalant to be well tolerated and effective for cardioversion of AF [7–18].
The FDA (Food and Drug Administration agency) decided in 2008 and in December 2019 not to approve to market vernakalant in the USA for safety concerns. In 2010, the EMA requested a post-authorization safety study to better define the risk benefit ratio in routine clinical practice. The objectives of SPECTRUM (Surveillance of Pharmacologic thErapy for Cardioversion in aTrial fibrillation Registry Using IV treatMent) (NCT01370629 and EUPAS2078) study were to assess the rates of adverse events and to estimate the effectiveness of the drug in a large cohort of patients with recent-onset AF.
Methods
Definitions
Recent-onset AF was defined as symptomatic episode within 7 days that will be undergoing cardioversion taking into account that about 70% of patients with symptomatic AF < 72 h were reported to convert spontaneously [19]. Beyond 7 days, AF is likely to persist and the chances of pharmacological cardioversion to be successful become low. Hypertension was reported when documented on the medical record or the patient report. Coronary artery disease (CAD) was diagnosed when the patient had a documented history of CAD and/or a history of coronary revascularization.
Patients and Procedures
Adult patients (≥ 18 years) with recent-onset AF occurring between September 1, 2011 and April 11, 2018 who received vernakalant for cardioversion were eligible for inclusion in this international, multicenter, observational, post-authorization study. Fifty-five hospitals in Austria, Denmark, Germany, Spain, Sweden, and Finland participated in the study, 53 of which enrolled patients. While administration of vernakalant was at the discretion of the treating physician, consecutively treated patients were enrolled and reasons for non-participation were documented. A preinfusion checklist and healthcare provider educational card were implemented during the study period to assist in identifying patients for treatment consistent with the approved indications and contraindications.
Patients were required to give informed consent for participation in the study and could be enrolled more than once if they presented on multiple occasions for AF episodes. Patients who had participated in an investigational drug/device clinical trial within 30 days prior to enrollment were not eligible. In order to enhance enrollment and reach the EMA required target of 2000 episodes, a protocol amendment was made in September 2016, which permitted retrospective inclusion of patients who had received vernakalant between April 2013 and the end of the study, provided that they fulfilled the established eligibility criteria. For prospectively enrolled patients, data were collected from both medical records and supplemental standardized data collection forms. For retrospectively enrolled patients, only medical records were available. The study period comprised a baseline assessment and up to 24-h follow-up after completion of the last infusion or until discharge. This study was mandated and approved by the European Committee for Medicinal Products for Human Use. The study protocol was approved by the appropriate local research ethics committees for all participating centers, and the study was conducted in accordance with applicable national and local regulations/guidelines, accepted standards for Good Clinical Practice, Guidelines for Good Pharmacoepidemiology Practices, and the Declaration of Helsinki [20].
Study Objectives and Endpoints
The primary objectives of the study was to estimate the incidence of clinically predefined serious adverse events (SAEs), i.e., significant hypotension (systolic blood pressure < 90 mmHg or requiring vasopressors); sustained (> 30 s) ventricular arrhythmias, Torsade de Pointes (>10 s) or ventricular fibrillation, atrial flutter with 1:1 conduction, bradycardia requiring temporary electrical pacing, or sinus arrest (> 3 s). Definition of these predefined SAEs was based on events from previous controlled studies on IV vernakalant [7, 8, 11, 12] and from the reported adverse events (AEs) on other antiarrhythmic agents. Secondary objectives included the rates of all other SAEs. Each SAE was reviewed and adjudicated by an independent expert Safety Review Committee (SRC). This study had also the objective to determine the conversion rate to sinus rhythm in a large population of patients outside the setting of controlled clinical trials.
The duration of the index AF episode was calculated as the time between the patient-reported time of symptom onset and the start of the first vernakalant infusion. Successful cardioversion was defined as conversion to sinus rhythm within 90 min of the start of vernakalant infusion. Conversion rate was calculated in all patients, as well as in an effectiveness population excluding all treatment episodes in which patients received another therapy for cardioversion within 90 min of the start of vernakalant administration (e.g., electrical or pharmacological cardioversion). Vernakalant is recommended to be administered in a step-dose fashion. Each treatment episode can comprise up to two infusions, separated by a 15-min observation period. The recommended doses for the first and second infusions are 3.0 mg/kg and 2.0 mg/kg, respectively, each administered over 10 min. For patients above 113 kg, vernakalant has a fixed initial dose of 339 mg. If conversion to sinus rhythm does not occur within 15 min after the end of the initial infusion, a second 10-min infusion of 226 mg may be administered.
Statistics and Analyses
A target sample size of 2000 vernakalant IV treatment episodes was chosen to allow adequate statistical precision, as expressed by a two-sided 95% confidence limit. Enrollment per site was capped at 10% of the total study population and 40% per country to minimize any potential bias in practice patterns. Categorical variable frequency, along with 95% confidence intervals (CIs), was determined for the summed treatment episodes. Continuous variables were summarized using descriptive statistics. Data were analyzed based on enrollment method (prospective vs retrospective) and reported as stratified and unstratified CIs. All analyses were performed using Statistical Analysis System v9.2, or later, software.
Results
Study Population
A total of 1778 patients who presented with 2009 treatment episodes were included: 1580 episodes were in prospectively enrolled patients and 429 in retrospectively enrolled patients (Table 1). The majority of patients were treated in the ED for 1289 (64.1%) AF episodes and 563 (28.0%) AF episodes in the coronary or intensive care units, with the remainder 157 (7.8%) episodes being treated in other hospital settings. As seen in Fig. 1, the main reason for non-inclusion in the study was lack of informed consent. In 1905 (94.7%) AF episodes, vernakalant was administered to non-surgery patients, and in 104 (5.2%) to post-cardiac surgery patients. The later are among the prospectively included patients. The mean age of the overall patient population at time of treatment was 62.3 ± 13.0 years (mean ± standard deviation [SD]), ranging from 18.0 to 94.0 years, and 1222 (60.8%) episodes occurred in men (Table 1). At baseline, systolic blood pressure (BP) was 132.5 ± 19.5 mmHg and heart rate (HR) was 112.9 ± 25.5/min (mean ± SD). The median duration of AF episode prior to treatment was 11.1 (5.4–27.0) hours (median [interquartile range, IQR]). In 88.9% of episodes, the patients were treated within 48 h of the onset of symptoms, and in 72.5% within 24 h. Duration of AF before treatment in 104 post-cardiac surgery patients was shorter than in the overall population, with 3.6 h (range 0.8–15.4) (median [IQR]). Baseline demographics and characteristics were similar between patients enrolled prospectively and retrospectively. Total length of ED stay was 7.5 (5.0–13.5) hours (median [IQR]). Only 167 (13.0%) of patients initially managed in the ED were in hospital for 24 h or longer. The number of vernakalant infusions was available in 1990 patients. Of these, 1201 (60.4%) received one vernakalant infusion and 789 (39.6) received a total of 2 infusions.Table 1 Clinical characteristics of patients
Total Prospective Retrospective
No. of patients 2009 1580 429
Age (years) mean ± SD 62.3 ± 13.0 61.9 ± 13.5 63.6 ± 11.2
Range (years) 18.0–94 18–93 30–94
Male, n (%) 1222 (60.8) 998 (63.2) 224 (52.2)
Body weight (kg) mean ± SD 84.1 ± 16.5 84.3 ± 16.5 83.1 (16.9)
Range (kg) 45.0–189.0 45.0–189.0 45.0–165.0
Body mass index (kg/m2) 27.8 ± 4.9 27.7 ± 4.8 28.2 ± 5.1
Associated conditions, n (%)
Hypertension 1103 (54.9) 884 (55.9) 219 (51.0)
Coronary artery disease 118 (5.9) 82 (5.2) 36 (8.4)
Cardiomyopathy 33 (1.6) 31 (2.0) 2 (0.5%)
Heart failure (history) 63 (3.1) 59 (3.7) 4 (0.9)
Diabetes 199 (9.9) 165 (10.4) 34 (7.9)
Stroke (history) 91 (4.5) 68 (4.3) 23 (5.4)
Pacemaker/ICD 36 (1.8) 24 (1.5) 12 (2.8)
Type of AF episode
First detected 477 (23.7) 393 (24.9) 84 (19.6)
Previous history of AF 1458 (72.6) 1115 (70.6) 343 (80.0)
Onset unknown/not assessed 5 (0.2) 3 (0.2) 2 (0.5)
Post-surgery 69 (3.4) 69 (4.4) 0 (0.0)
Symptoms on admission, n (%)
Palpitations, irregular heart beat 1749 (87.1) 1337 (84.6) 412 (96.0)
Dyspnea or shortness of breath 352 (17.5) 306 (19.4) 46 (10.7)
Dizziness, light-headedness 320 (15.9) 251 (15.9) 69 (16.1)
Chest pain 271 (13.5) 220 (13.9) 51 (11.9)
Syncope, near syncope 61 (3.0) 53 (3.4) 8 (1.9)
Duration of the index episode
Less than 24 h, n (%) 1438 (72.5) 1107 (70.2) 331 (81.5)
24–48 h, n (%) 347 (17.5) 288 (18.3) 59 (14.5)
More than 48 h 199 (10.0) 183 (11.6) 16 (3.9)
Mean duration ± SD (h) 23.2 ± 44.9 24.9 ± 45.8 16.8 ± 40.6
Median (IQR 25–75) (h) 11.1 (5.44–27.03) 11.9 (5.8–29.7) 8.2 (4.8–18.3)
Antiarrhythmic agents, n (%)
Betablockers 1055 (52.5) 800 (50.6) 255 (59.4)
Calcium channels blockers 22 (1.1) 20 (1.3) 2 (0.5)
Class I agents* 85 (4.2) 71 (4.5) 14 (3.3)
Class III agents* 98 (4.9) 89 (5.6) 9 (2.1)
Digitalis glycosides 22 (1.1) 18 (1.1) 4 (0.9)
*Using the Vaughan-Williams classification
Fig. 1 Study flow chart. Flow chart showing patient enrollment in the SPECTRUM study. The term patient here refers to individual treatment episodes (asterisk). Owing to lack of informed consent (n = 500) (dagger). Other reasons included patient enrollment in an investigational drug trial in the past 30 days, spontaneous conversion to sinus rhythm, ejection fraction 30–35%, electrical cardioversion preferred, missing information regarding start of atrial fibrillation, inclusion criteria not met, other, or no reason provided or known. Source data could not be verified to confirm that vernakalant IV was administered (double dagger). Spontaneous conversion to sinus rhythm before vernakalant IV administration (section sign). Patient decision and lack of follow-up after cardioversion in one case each (double vertical line). IV intravenous
Predefined Serious Adverse Events and Other Adverse Events
No deaths were recorded in our study. Nineteen predefined SAEs were reported during or after 17 treatment episodes (cumulative incidence 0.8%; CI 0.5–1.4%) (Table 2). Eighteen of the 19 events occurred within 2 h from the start of infusion. The remaining event was an episode of atrial flutter with 1:1 conduction which occurred 3.1 h after drug infusion and was terminated by electrical shock. Symptomatic bradycardia was the most common event occurring in 11 (0.5%; CI 0.4–1.2%) episodes (Table 2). Conversion to sinus rhythm occurred in 10 of these cases. A pause described as sinus arrest preceding the restoration of sinus rhythm occurred in 4 patients. In 2 patients, sinus arrest was associated with sinus bradycardia. In all bradycardia and sinus arrest cases, the vernakalant infusion was immediately discontinued. One of these 4 sinus arrests occurred in a 66-year-old man, sportive cyclist with no history of heart disease, admitted for a first episode of AF with a mean ventricular response of 95 beats/min. He received 300 mg orally of flecainide which failed to restore sinus rhythm. The treating physician decided 4 h later, to administer IV vernakalant. At the end of the infusion, a pause of 6 s, with a brief dizziness, occurred and resolved spontaneously, followed by a normal sinus rhythm with a HR of 47 beats/min which was patient usual HR and a BP of 120/85 mmHg. This event was considered a SAE although there was probably an interaction between oral flecainide still active and vernakalant in this event. One of the bradycardia events occurred in a retrospectively enrolled 69-year-old woman on bisoprolol with a history of hypertension and CAD, who developed 8 min after the second infusion of vernakalant a sinus bradycardia which rapidly resolved with IV atropine. Two bradycardia episodes occurred in post-cardiac surgery patients requiring temporary electrical pacing through the electrodes left in place by the surgeon. Both patients converted to sinus rhythm. None of the non-surgery patients required temporary electrical pacing. Significant hypotension occurred on two (0.1%; CI < 0.1–0.4%) occasions, associated with sinus bradycardia in both instances. Both events resolved with intravenous atropine and fluid. There were two cases of atrial flutter with 1:1 ventricular conduction terminated with electrical shock whereas no cases of sustained ventricular tachycardia (VT), ventricular fibrillation, or Torsade de Pointes were observed. In addition to the predefined SAEs, there were 9 other SAEs, one of which occurred in a retrospectively enrolled patient (Table 2). They included two instances of hypotension not requiring vasopressor agents, 2 non-sustained VT which deserve special attention. The first non-sustained VT occurred in a 48-year-old man with asthma admitted with fever, palpitations, dyspnea, and first episode of AF with a ventricular rate of 144 bpm. During vernakalant infusion, 5 beats of non-sustained VT was observed. Among the tests done, coronary angiography was reported as normal. The same run of 5 beats of non-sustained VT was observed 20 h after infusion (next day) making the causal effect of vernakalant unlikely. The other event occurred in a 57-year-old patient with a 6-year history of recurrent symptomatic AF and arterial hypertension with left ventricular hypertrophy. He was admitted with palpitations, irregular heartbeats, and dizziness. He was on dronedarone, and ECG showed AF with a ventricular rate of 135 bpm. During infusion of vernakalant, he had 6 s of non-sustained VT observed on the monitor and was given 5 mg of bisoprolol which reduced the heart rate to 120 beats/min and relieved patient symptoms. The Safety Review Committee considered that in the first case, the wide QRS complexes were due to aberrant conduction during rapid AF (Ashman phenomenon). Among the non-predefined SAEs, one supraventricular tachycardia (120 beats/min) and a single report each of angina pectoris, pericardial effusion, transient visual disturbance, and vernakalant overdose (Table 2). A total of 188 non-serious AEs were reported, the most common of which were dysgeusia (n = 35) and sneezing (n = 27). All patients with vernakalant-related AEs recovered without sequelae. All but 6 of the 28 SAEs were considered by the investigators and the SRC to be related to vernakalant administration.Table 2 Adverse events in 2009 episodes during treatment and observation periods
Event type Number of events Incidence (95% CI) Considered drug-related, n (%)
All SAEs 28 1.3% (0.8–1.9) 22 (78.6)
Predefined SAEs 19 0.8% (0.5–1.4) 18 (94.7)
Significant hypotension 2 0.1% (< 0.1–0.4) 2 (100.0)
Bradycardiaα 11 0.5% (0.3–10) 10 (93.3)
Sinus arrest (> 3 s)β 4 0.2% (< 0.1–0.4) 4 (100.0)
Atrial flutter with 1: 1 AV conduction 2 0.1% (0.1–0.4) 2 (100.0)
Ventricular tachycardia γ 0 0 0 (0.0)
Other than predefined SAEs 9 0.45% 5 (55.6)
Hypotension 2 0.1% 1 (50.0)
Supraventricular tachycardiaδ 1 < 0.1% 1 (100.0)
Non-sustained ventricular tachycardiaε 2 < 0.1% 1 (50.0)
Angina pectoris 1 (< 0.1) < 0.1% 0 (0.0)
Pericardial effusion 1 (< 0.1) < 0.1% 0 (0.0)
Visual disturbance 1 (< 0.1) < 0.1% 0 (0.0)
Vernakalant overdoseζ 1 (< 0.1) < 0.1% 1 (100.0)
αNine cases of sinus bradycardia and 2 reported as significant bradycardia
βOne patient had both sinus arrest followed by sinus bradycardia
γOne event reclassified as atrial flutter with 1:1 conduction
δAtrial arrhythmia other than atrial flutter
εSee text, exceeding 5% of the weight-based dosing recommendation. In this case, the administered dose was 51% in excess of the recommended dose
Rates of Conversion to Sinus Rhythm
Overall, conversion to sinus rhythm at any time following vernakalant infusion occurred in 1448 out of 2009 (72.1%) treatment episodes. Successful cardioversion was recorded in 70.2% (CI 68.1–72.2%) of the 1936 episodes of the effectiveness population excluding those in which either electrical cardioversion (n = 68) or an additional intravenous Class I/III antiarrhythmic drug (n = 6) was given within 90 min of infusion initiation. The rate of cardioversion was similar between the 1107 of 1580 (70.1%) episodes included prospectively and the 297 of 421 (70.5%) episodes of retrospectively enrolled patients. Successful cardioversion of AF was reported in 68 of 104 (65.4%) of treatment episodes in the post-cardiac surgery patients. Time to cardioversion was recorded in 1413 of 1448 episodes with successful conversion to sinus rhythm. The median time to conversion was 12.0 (8.0–28.0) minutes (median [IQR]) Fig. 2). One thousand one hundred eight of 1413 (78.4%) successful cardioversions were treated with only one drug infusion. The percentage of successful cardioversion was 70.1% in the prospective patients and 70.5% in the retrospective patients. The median hospital stay time in those treated in the ED was 7.5 h allowing patient discharges when their condition was clinically stable.Fig. 2 Time to conversion to sinus rhythm. Time to conversion to sinus rhythm with vernakalant IV in the effectiveness analysis population (N = 1936). Time to conversion was not recorded in 29 treatment episodes in which patients converted to sinus rhythm; these episodes are not displayed on the graph but are taken into account for the proportion calculation. IV intravenous
Anticoagulation
About a quarter of patients presenting with recent-onset AF at baseline were on vitamin K antagonists or direct oral anticoagulants. Investigators respected current guidelines [3] on anticoagulation both peri-procedurally and after hospital discharge.
Discussion
The SPECTRUM study included a large real-world patient population of 1778 patients with 2009 recent-onset AF episodes in whom pharmacological cardioversion was performed with vernakalant. About 70% of patients were cardioverted within 12 min from onset of infusion and 11 h from the AF onset. Our findings confirm the safety and efficacy of vernakalant reported in previous studies [7–18, 21–25] and extend their consistency to routine hospital use in large populations. To our knowledge, the present study provides the largest series of patients with recent-onset AF undergoing pharmacological cardioversion with a specific antiarrhythmic agent. The safety was the main objective of this study. We found the incidence of both predefined and other SAEs to be lower than expected. There were no death and no sustained ventricular arrhythmia. Overall, 28 SAEs (1.3%) were recorded. The majority of patients were AF treated in ED and intensive care units.
Pharmacological cardioversion is frequently indicated as part of a rhythm control strategy or as a tool to control patient symptoms and avoid hospitalization in clinically stable condition [25, 26]. It is often preferred to electrical cardioversion in patients with hemodynamically stable condition as it does not require general anesthesia or sedation. Among agents currently available for rapid termination of recent-onset AF, vernakalant represents an option [3]. However, there has been to our knowledge, no large study exploring the safety of vernakalant in daily practice.
There is no universal definition for recent-onset AF. In current literature, the duration limits of AF episodes range from < 24 [27] to < 48 h and even < 7 days [28, 29]. The prevalence of recent-onset AF among all AF subsets varies from 11% when restricted to the first detected episode (new onset) [30] to 26% [31]. The characteristics of patients were similar to those of other AF cohorts [31, 32].
As with electrical cardioversion, pharmacological cardioversion can be associated with post-cardioversion bradyarrhythmias, often unmasking pre-existing sinus node dysfunction or atrioventricular conduction abnormalities and can result in ventricular escape rhythms or prolonged ventricular pauses. Of interest, these pauses were first reported by Lown [33], following electrical cardioversion as the possible reflect of sinus dysfunction. Another possible mechanism for these sinus arrests is right atrial stunning [34]. The cumulative incidences of bradycardia (0.5%), sinus arrest (0.2%), and hypotension (0.1%) observed in this study were also low. The incidence of atrial flutter with 1:1 conduction was lower than that reported with oral Class Ic antiarrhythmics, such as flecainide or propafenone. The “pill in the pocket” approach requires initiation of therapy in hospital to verify its safety [35]. No cases of Torsade de Pointes or sustained VT was observed, which is in line with the low risk of ventricular proarrhythmia associated with vernakalant owing to its electrophysiological properties [5]. This contrasts with the reported incidence of Torsade de Pointes [24, 36, 37] in patients with AF/atrial flutter of 4.3% with intravenous ibutilide in the report of Kowey et al. including 1.7% of which required cardioversion [36]. Of note, all but one of the predefined SAEs in this study occurred within 2 h of the start of infusion. As aforementioned, the remaining patient had atrial flutter with 1:1 ventricular conduction which occurred 3.1 h following infusion initiation, indicating that close cardiac monitoring should be available during and after treatment in some patients.
Conversion to sinus rhythm with vernakalant was rapid (median time of 12.0 min) similar to what was previously reported [9–18]. The conversion median time of ibutilide was significantly longer than that of vernakalant (26 min versus 10 min, P = 0.01) in a randomized comparison [18]. Furthermore, in this particularly large real-world study, the median duration of AF episode was short (11.1 h) as there is important evidence, and relevant guidelines [3] suggesting that prompt cardioversion could be associated with benefits in terms of lower risk of thromboembolic events [4, 38, 39].
Although the baseline characteristics of the study population were consistent with AF population-based studies [31, 32] and clinical studies with vernakalant, the conversion rate was higher than that observed in recent review and meta-analysis (~ 50%) [22–25]. This seems likely to be due to patients being treated soon after symptom onset in European clinical practice. Other recent but smaller observational studies [13–15, 17, 18], which collectively included almost 1300 patients, have found similarly high conversion rates (65–86%) when vernakalant was administered soon after the onset of AF, particularly within the first 48 h [15, 17, 21]. Vernakalant has also been shown to induce a higher rate of cardioversion compared with flecainide (67% vs 46%) in a non-randomized cohort study [21]. Similarly, in randomized studies, vernakalant was more effective than amiodarone [12] (52% vs 5%; after 90 min) and ibutilide [18, 24] (69% vs 43%; within 90 min). The SPECTRUM results are consistent with previous reports that vernakalant is safe and effective for the rapid cardioversion of recent-onset AF and extends them to daily practice.
Owing to the rapid time to conversion with vernakalant, the median hospital stay time for those treated in the ED was 7.5 h. This is encouraging given that a study in France reported that hospitalization constitutes 60% of the cost of care for patients with AF [40].
Study Limitations
This multicenter international study was observational as the main objective was to determine the safety of vernakalant as used in daily hospital practice without interfering on the management of recent-onset AF by the treating physician. For these reasons, the adverse events were expected to be higher in an “uncontrolled” setting with no guidance on patient selection than those reported in controlled studies with strict protocols. In fact, SAEs were low in this study. Data collection for prospectively enrolled patients was comprehensive owing to use of both study-specific tools and medical records. However, for retrospectively enrolled patients, it was not possible to routinely collect all data of interest in a standardized manner. Nevertheless, baseline characteristics, medical histories, and SAEs in the retrospective cohort were similar to those in the prospective cohort, supporting the use of a retrospective analysis.
Conclusions
The results of this large multicenter study showed that vernakalant has a good safety profile and is effective in enabling rapid cardioversion in clinical practice. Moreover, the rates of serious complications were lower than those observed in early trials reflecting appropriate patient selection in clinical practice. In conclusion, vernakalant provides a rapid and effective means of pharmacological conversion in patients with recent-onset AF undergoing cardioversion undergoing cardioversion in daily hospital practice.
Appendix. List of Investigators
Austria: Michael Joannidis, Klemens Zotter Frank Hartig, Anton Sandhofer, Alois Süssenbacher, Bernd Eber, Elisabeth Lassnig, Ulrike Pfeifenberger, Michaela Steiner, Hans Domanovits, Alexander Simon, Alexander Spiel, Jan Niederdockl, Nikola Schuetz, Daniel Wehinger, Franz Xaver Roithinger, Isabella M. von Katzler, Katharina Bichler, Robert Schoenbauer, Lukas Fiedler, Michael Pfeffer, Markus Peck, Florian Benische, Michael Hackl, Susane Demschar, Astrid Ebner, Melanie Eder, Rainer Huditz, Arnulf Isak, Michael Moser, Georg Pinter, Thomas Singer, Claudia Waldhauser, Helmut Pürerfellner, Martin Martinek, Sandra Muellner, Andrea Ploechl, Tanja Koppler, Elisabeth Sigmund, Michael Derndorfer, Sabine Metz, Karin Streicher, Clemens Steinwender, Karim Saleh, Andreas Lueger, Petra Fladerer, Eiko Meister, Heinz Drexel, Alexandra Schuler, Susanne Waeger, Karl-Martin Ebner, Christine Heinzle, Arthur Mader, Peter Schwerzler, Berta Patsch, Abdurahman Said, Claudia Stoeckloecker, Daniela Zanolin, Jutta Bergler-Klein, Ljubica Mandic, Mariann Gyöngyösi, Neraida Cene, Zsuzsanna Szankai, Abelina Zimba.
Denmark: Henrik Nielsen, Bjerre Flemming, Michaelsen Michaelsen, Elisa Stokholm, Katja Holm, Charlotte Schmidt Skov, Pauline Gøgsig Johansen, Soren shjortshoj, Thomas Melchior, Ole Dyg Pedersen, Sanne Heinsvig, Inge Larsen, Vibeke Perret-Gentil, Thomas Wagner Nielsen, Axel Brandes, Marianne Jensen, Ida Rosenlund, Liv Gøtzsche, Heidi Munk Andersen.
Germany: Andreas Götte, Matthias Hammwohner, Britta Möehring, Jutta Schaertl, Daniel Steven, Iris Berg, Alexandra Kuehn, Hannes Reuter, Elena Terentieva, Christian Loges, Christine Lindner, Hendrik Bonnemeier, Christan Wulff, Thomas Demming, Svenja Gediehn, Johanna Parlitz, Wilhelm Haverkamp, Buehner Kathrin, Hubert Katja, Iacovella Ines, Bernhard Korbmacher, Marc Thone, Hannan Dalyanoglu, Da Un Chung, Naujoks Angela, Dirk Weismann, Björn Lengenfelder, Jan Becher, Klaus Meyer, Irina Turkin, Sebastian Maier, Marcus Koller, Alban Glaser, Lisa Gebele, Jale Goezuebueyuek, Ralph Hampe, Barbara Ruemmler, Hagen Schrötter, Manja Hubald, Cornelia Fritz, Martin Domhardt, Kathrin Haacke, Nicole Schmiedehausen, Ruth Strasser, Kristof Graf, Lidia Fischer, Roland Thieme, Karlheinz Seidl, Martin Kulzer, Monika Zackel, Gerian Grönefeld, Christina Paitazoglou, Simone Müller, ThoraBotschafter Britta Goldmann, Andrea Moeller, Sindy Bartel, Joern Schmitt, Damir Erkapic, Gabriele Hellwig-Bahavar, Ritvan Chasan, Christopher Gemein, Victoria Johnson, Christiane Kelm, Kay Weipert, Johannes Brachmann, Michael Held, Andrea Höhn, Ute Goebel, Andrea Linss, Swetlana Rube, Ahmed Saleh, Steffen Schnupp, Yeong-Hoon Choi, Vera Wolf, Andrea Plate, Anton Sabashnikov, Antje-Christin Deppe, Petra Krause.
Spain: Ignacio Fernandez Lozano, Manuel Sánchez, Francisco Hernández, Alfonso Martin Martinez, Pedro Vazquez, Esther Alvarez, Pascual Lopez, Raquel Torres, José Carbajosa Dalmau, Laura Parades, Nestor Hernandez, Inmaculada Jimenez Ruiz, Ana Maria Lopez, Luis López-Andujar, Alexandre Noguera, Francisco Roman Cerdan, Carmen del Arco Galan, Daniel Afonso, Raquel Caminero, Manual Lunquera, Monica Negro, Cristina Santiago, Nestor Villalba, José Manuel Garrido Castilla, Roberto Martinez Asenjo, Elena Mejia Martinez, José Luís Merino Llorens, Maria Jesus Diaz-Pintado, Jorge Alejandro Figueroa, Alberto Borobia Perez, Sergio Castrejon Castrejon, David Filgueiras Rama, Manuel Quintana Diaz, Maria Angelica Ribera Nunez, Miguel Angel Ramirez Marrero, Antonio Martin, José Miguel Ormaetxe Merodio, Mercedes Varona Peinador, Maria Fe Arcocha, Larraitz Gaztañaga, F. Xavier Palom Rico, Javier Jacob Rodriguez, Pascual Piñera Salmeron, Juan Cosin Sales, Isabel Navarro, Francisco Buendia Funetes.
Sweden: Henrik Wallentin, Kerstin Roos, Arash Mokhtari, Hans-Jörgen Nilsson, Peter Vasko, Terese Nyström, Martina Gustensson, Susanne Johansson, Inga Uggeldahl, Göran Andersson, Olle Bergström, Thomas Aronsson, Mehmet Hamid, Kerstin Giocondi, Deborah Svanerö, Qassim Awad, Tord Juhlin, Hjördis Jernhed, Stefan Berglund, Magnus Forsgren, Michael Guggi, Pär-Lennart Agren, Kristina Eriksson, Kristina Karlsson, Per Blomström, Alejandro Utreras, Caroline Lundgren, Maria Soderlund, Frederik Buijs, Solveig Östberg, Johan-Emil Bager, Ingrid Hendequist, Maria Just, Siv Heden, Liselott Lisjo, Chrichan Mansson, Helen Svanstrom, Mikael Dellborg, Helena Dellborg, Gorel Hultsberg Olsson, Linus Hansson.
Finland: Hannu Sulonen, Hanna Suurmunne, Anna Petrovskaja, Johanna Markkanen, Juha Hartikainen, Lari Kujanen, Antti Heikkola, Hanna Pohjantahti MaarooS.
Safety Review Committee
Lars Kober, MD, Samuel Lévy, MD (Chair), Cristina Varas-Lorenzo, MD, MSc, PhD, Manel Pladevall-Vila MD, MS.
The authors wish to thank Nathalie Dunkel and her team for their help in accessing the data of SPECTRUM and providing information and documents upon our request.
The authors would like to thank all of the investigators involved in the SPECTRUM study.
Funding
This study was funded by Correvio International Sàrl, Geneva, Switzerland. The authors received no funding for their participation in this manuscript.
Data Availability
The data underlying this article will be shared on reasonable request to the corresponding author.
Compliance with Ethical Standards
Conflict of Interest
Professor Levy reports no conflicts of interest. Professor Hartikainen has been an investigator in studies sponsored by AstraZeneca, Biosense Webster, Boehringer Ingelheim, Correvio International Sàrl, Medtronic, and St. Jude Medical. Dr. Ritz is an employee of Correvio International Sàrl, Geneva, Switzerland. Dr. Juhlin has received speaker honoraria from Correvio International Sàrl. Professor Domanovits reports no conflicts of interest. Dr. Carbajosa-Dalmau has received speaker honoraria from Correvio International Sàrl.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. | FLECAINIDE, VERNAKALANT | DrugsGivenReaction | CC BY | 33206300 | 19,695,564 | 2021-04 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.