Datasets:

ClarusC64
/

clinical_narrative_integrity_v0.2

case_id stringclasses 10 values	phase stringclasses 3 values	indication stringclasses 7 values	population stringclasses 10 values	design_summary stringclasses 10 values	endpoints_primary stringclasses 10 values	stats_summary stringclasses 10 values	efficacy_results stringclasses 10 values	safety_results stringclasses 10 values	subgroup_results stringclasses 10 values	multiplicity_note stringclasses 10 values	interim_summary stringclasses 10 values	topline_summary stringclasses 10 values	press_release_headline stringclasses 10 values	investor_call_quote stringclasses 10 values	csr_excerpt stringclasses 10 values	drift_claims_expected stringclasses 10 values	regulatory_acceptability_expected stringclasses 2 values	risk_level stringclasses 2 values	corrected_narrative_expected stringclasses 10 values
CNI2-001	Phase II	Oncology	2L metastatic solid tumor	Randomized double blind placebo controlled 1:1 N=180 24w	PFS	Primary p=0.18 HR 0.82 CI 0.62-1.09; OS exploratory	PFS not significant; ORR +3% absolute	SAE 14% vs 9%; G3+ 28% vs 18%	Subgroup PD-L1 high HR 0.62 p=0.04 post hoc	Multiple subgroups; no adjustment	Interim: trend observed on PFS with acceptable safety	Topline: clinically meaningful PFS benefit supports registration intent	Topline results demonstrate meaningful benefit and momentum	We basically hit the endpoint in the patients that matter	CSR: Primary endpoint did not reach significance; safety signals include higher SAE	Primary missed framed as success\|Subgroup promoted beyond evidence\|Registration readiness overstated\|Safety imbalance softened	not_supportable	high	The primary endpoint did not meet statistical significance for PFS. Exploratory analyses including PD-L1 high subgroup are hypothesis generating and not adjusted for multiplicity. Safety showed higher SAE and G3+ rates versus placebo. Results support further study with prespecified endpoints and risk mitigation rather than registration intent
CNI2-002	Phase III	Cardio	HFpEF N=920	Randomized active control non inferiority 52w	Composite CV death or HF hospitalization	Met non inferiority margin; superiority p=0.21	Non inferiority met; no superiority	AE similar; hypotension higher 8% vs 4%	No consistent subgroup benefit	Multiplicity controlled for primary only	Interim: on track for pivotal readout	Topline: improved outcomes vs standard of care	New therapy improves outcomes for HFpEF patients	We have a winner and the market will see it	CSR: Non inferiority met; superiority not demonstrated; hypotension increased	Non inferiority reframed as superiority\|Certainty language beyond stats\|Safety risk underplayed	borderline	medium	The study met its non inferiority objective for the primary composite endpoint. Superiority was not demonstrated. Adverse events were broadly similar, with higher hypotension versus control. Findings support non inferiority positioning with clear safety labeling and no superiority claims
CNI2-003	Phase II	Immunology	Moderate UC N=140	Randomized placebo controlled 12w	Clinical remission at week 12	Remission 18% vs 12% p=0.29; secondary endoscopic improvement p=0.04	Primary not significant; secondary signal	Serious infection 3% vs 1%	Subgroup biologic naive p=0.06	Secondary not multiplicity adjusted	Interim: encouraging endoscopic signal	Topline: meaningful efficacy demonstrated	Meaningful efficacy demonstrated in ulcerative colitis	We showed clear efficacy and a clean profile	CSR: Primary not significant; endoscopic secondary nominal p; infections slightly higher	Primary missed framed as efficacy\|Secondary promoted without caveats\|Safety minimized	borderline	medium	Clinical remission at week 12 did not differ significantly versus placebo. A nominal secondary endoscopic improvement signal was observed without multiplicity adjustment and should be interpreted cautiously. Serious infections were numerically higher. Results support refinement of dose and population and a larger confirmatory study with prespecified hierarchy
CNI2-004	Phase I	Oncology	Advanced hematologic malignancy N=42	Open label dose escalation	DLT rate and safety	DLT 19%; MTD not reached	No efficacy endpoint; responses observed 6 PR	CRS any grade 31%; G3 CRS 7%	Responses in dose level 3 only	Exploratory only	Interim: manageable CRS observed	Topline: favorable safety supports rapid expansion	Favorable safety supports rapid expansion	CRS is basically solved and we can scale dosing	CSR: CRS common including G3; responses limited; no MTD	Safety risk softened\|Expansion oversold\|Efficacy implied beyond design	borderline	high	In dose escalation, CRS was common including grade 3 events, and the MTD was not reached. Limited partial responses were observed in a subset at one dose level and are exploratory. Safety supports continued cautious escalation and protocol guided expansion with clear CRS management and monitoring
CNI2-005	Phase III	Neuro	Early AD N=760	Randomized placebo controlled 18mo	Change in cognitive scale	Difference -0.8 points p=0.049; missingness 22% vs 14%	Nominal p with high missingness; sensitivity mixed	ARIA-E 12% vs 3%	APOE4 carriers larger effect	Multiplicity controlled; missing data sensitivity preplanned	Interim: trending toward benefit	Topline: statistically significant slowing supports filing	Statistically significant slowing of decline	We have definitive proof of disease modification	CSR: nominal significance; missingness higher; ARIA-E increased; sensitivity mixed	Overstatement of certainty\|Missingness downplayed\|Safety risk softened\|APOE4 subgroup promoted	borderline	high	The primary endpoint showed a small nominally significant difference with higher missingness in the active arm and mixed sensitivity analyses. ARIA-E occurred more frequently. Subgroup patterns including APOE4 require cautious interpretation within the prespecified analysis plan. Results may support discussion with regulators with full sensitivity and safety disclosure rather than definitive efficacy claims
CNI2-006	Phase II	Metabolic	T2D N=210	Randomized placebo controlled 16w	HbA1c change	-0.4 vs -0.2 p=0.11	Primary not significant	GI AE 22% vs 10%	Subgroup baseline HbA1c high p=0.03 post hoc	Many subgroups unadjusted	Interim: signal in high baseline HbA1c	Topline: robust glycemic control achieved	Robust glycemic control achieved	If you look at the right patients it works great	CSR: Primary not significant; GI AE higher; subgroup post hoc	Primary missed framed as robust\|Subgroup cherry pick\|Safety underplayed	not_supportable	medium	HbA1c reduction did not differ significantly versus placebo on the primary analysis. A post hoc signal in higher baseline HbA1c subgroup is hypothesis generating and not adjusted for multiplicity. GI adverse events were more frequent. Results support further dose and population selection work before confirmatory trials
CNI2-007	Phase III	Respiratory	Severe asthma N=520	Randomized active comparator 52w	Annual exacerbation rate	Rate ratio 0.92 p=0.33	No significant difference	Anaphylaxis 0.6% vs 0%	Eosinophil high subgroup RR 0.70 p=0.07	Subgroup exploratory	Interim: comparable efficacy expected	Topline: meaningful reduction in exacerbations	Meaningful reduction in exacerbations	Clinicians will switch once they see this	CSR: no significant difference; rare anaphylaxis events	No diff framed as meaningful\|Switch claim overreach\|Safety signal omitted	borderline	medium	The study did not demonstrate a statistically significant reduction in annual exacerbations versus comparator. Exploratory subgroup trends were not significant. Rare anaphylaxis events occurred with active treatment. Results do not support claims of superior exacerbation reduction and require clear safety communication
CNI2-008	Phase II	Infectious	Hospital CAP N=160	Randomized open label non inferiority	Clinical cure day 14	Non inferiority met; mortality higher 6% vs 2%	NI met; mortality imbalance	QT prolongation 9% vs 3%	No clear subgroup	Primary NI only	Interim: NI likely	Topline: successful trial supports broad use	Successful trial supports broad use	We proved it is just as good and safe	CSR: NI met; mortality imbalance; QT prolongation higher	Safety imbalance omitted\|Broad use overstated\|Certainty beyond mixed safety	borderline	high	The trial met its non inferiority objective for clinical cure. Mortality and QT prolongation were numerically higher in the active arm and require careful evaluation. Conclusions should state NI for cure with explicit safety imbalances and a plan for further safety characterization before broad claims
CNI2-009	Phase I	Oncology	Solid tumor N=36	Open label expansion cohorts	Safety and PK	No DLT; exposure variable	Stable disease 14/36; no PR	Transaminitis G3 11%	Stable disease in slow growing tumors	Exploratory	Interim: clean safety profile	Topline: strong anti tumor activity	Strong anti tumor activity observed	We are seeing responses everywhere	CSR: no PR; stable disease common; transaminitis noted	Efficacy inflated\|Certainty language\|Safety minimized	not_supportable	medium	In early phase evaluation, no DLTs were observed, and exposure was variable. No confirmed partial responses were reported; stable disease occurred in some patients and may reflect disease biology. Grade 3 transaminitis occurred. Conclusions should avoid claims of strong activity and present these findings as preliminary
CNI2-010	Phase III	Oncology	NSCLC 1L N=640	Randomized active control	OS	OS HR 0.88 p=0.09; PFS HR 0.86 p=0.04 secondary	Primary OS not significant; secondary PFS nominal	Immune AE 21% vs 14%	Subgroup PD-L1 high OS HR 0.74 p=0.06	Secondary nominal; hierarchy prespecified OS first	Interim: PFS benefit possible	Topline: improved survival and disease control	Improved survival in first line NSCLC	We improved survival and will file fast	CSR: OS not significant; PFS nominal; immune AE higher	Primary OS overclaimed\|Secondary PFS promoted\|Filing readiness overstated\|Safety underplayed	borderline	high	Overall survival did not reach statistical significance. A nominal secondary PFS difference was observed within the hierarchical framework and should be reported with appropriate caveats. Immune related adverse events were more frequent. Results support continued evaluation and regulator discussion without definitive survival claims