text
stringlengths 2
132k
| source
dict |
|---|---|
(i.e. the χ2 is greater than the degrees of freedom (df)), then the standard error should be increased by ( χ2 /df) 1/2 . Many computer programs calculate SE values 49 that are based on the sum of squares, instead of the Poisson estimate of the variance, which may lead to a false underestimation of the Poisson error. For this reason, when SE are calculated using this method, and the df is greater than the χ2 , it is a good practice to increase the SE by (df /χ2 )1/2 .Table 5 indicates the fitted coefficients when the data from Table 4 are used and the SEs calculated using Poisson assumptions. TABLE 5. THE RESULTS OF FITTING THE DICENTRIC DATA FROM TABLE 4 γ-rays (Cobalt-60) C ± SE α (Gy -1 ) ±SE β (Gy -2 ) ±SE χ2 df 0.00128 ± 0.00047 0.02103 ± 0.00516 0.06307 ± 0.00401 6.61 8 F = 4.08, p<0.03 F = 15.73, p<0.01 p = 0.58 20 MeV 4 He particles C ± SE α (Gy -1 ) ±SE β (Gy -2 ) ±SE χ2 df 0.00143 ± 0.00093 0.32790 ± 0.02875 0.02932 ± 0.01636 7.40 6 F = 11.41, p<0.01 F = 1.79, p = 0.25 p = 0.39 0,00193 ± 0,00097 0.37290 ± 0.01787 10,91 7 F = 20.87, p 1 <0.01 p = 0.14 1 F.05 [7, 7] = 3.77 The p values of the χ2 -test shown in Table 5 indicate that the fitted data points were not statistically different from the observed ones confirming a good fit. Moreover the significance of the linear and quadratic coefficients was also confirmed by the F-test, the ratio between each coefficient and its SE; for each coefficient the F value was higher than 3.44 (the cut off value for F.05 [8, 8] )
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
and the z value was higher than 1.96 (the cut off value for the normal distribution; both values can be found in the standard tables). The F test, which is described in Annex VI, is a ratio of two Chi-squared distributions and F.05 [8, 8] means the cut off value for alpha = 0.05, for 8 degrees of freedom for the numerator and 8 degrees of freedom for the denominator. For 4 He particles, weights were decreased by the average value of σ2 /y, 1.19. The β coefficient of the linear-quadratic was not significant (cut off value for F.05 [6, 6] = 4.28); z-test p = 0.12) and for this reason a linear fit is also presented. Opinions vary on how to treat the background level of aberrations in fitting dose response data. In general there are three approaches: a dose point at zero Gy is included in the curve fitting procedure, the zero dose point is ignored, or else the zero dose point is represented in every fitting procedure by a standard background value. If the measured yield at zero dose is used as one of the data points for the curve fitting (as used in the curve fitting presented above), the background becomes a variable parameter. However, since the yield in unirradiated cells is usually low, often none are observed so the measured yield at zero dose is zero. As discussed, at low doses, the statistical resolution of the data points is generally low. Thus, including the zero dose point in the curve fitting procedure can sometimes lead to negative estimates of the background value (C) and negative linear coefficients ( α), which obviously have no biological basis. Some investigators resolve this problem by ignoring zero dose data points and constraining the curve to pass through the origin.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
There are, however, sufficient data published from surveys of subjects exposed only to background radiation to show that there is a small positive background level of aberrations. An alternative method adopted by some workers is 50 therefore to use a small positive background value as a data point and to ascribe a large percentage of uncertainty to it. Ideally a laboratory should generate its own background data, although this requires the analysis of many thousands of cells. A consensus has emerged that the background level of dicentrics is ~0.5–1.0 per 1000 cells whilst for translocations and micronuclei the control values are higher. There are several programs that can be used for curve fitting such as a Poisson Iteratively Reweigthed Least Squares (PIRLS) computer program for additive, multiplicative, power, and non-linear models developed by Peterson , or the generalized linear interactive modelling (GLIM, www.nag.co.uk./stats/GDGE-soft.asp), or using R-based tools 2 . This should be combined with a routine specifically written for curve fitting that is available in this publication in Annex VI-3. Additionally, a number of specialized curve fitting computer programs have been recently developed from within the radiation cytogenetics community . CABAS uses maximum likelihood methods to fit calibration data to the linear quadratic equation. Dose Estimate is a similar tool and this allows both linear quadratic and linear fitting. Apart from curve fitting both CABAS and Dose Estimate have additional tools that assist with processing data from radiation accident cases in order to derive dose estimates when the circumstances depart from recent acute and whole body exposure. These cover the range of calculations described later in Section 9.7. Whether these or other software are employed, the program should give sufficient information regarding the methods used and provide details of the variances and covariances on the
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
fitted coefficients, as these are required for calculation of the uncertainties on dose estimates (Section 9.7.3). 2 R is a free software environment for statistical computing and is available for download from 51 9. DICENTRIC ANALYSIS The following text concerning cell culture, fixation and slide staining is written in the context of the dicentric assay but much of the material is also relevant to the other assays covered in this publication. By fully discussing the procedures here, the sections covering the other assays need simply to detail how the processes depart from that for dicentrics. In brief the important differences are: (a) for the dicentric assay two day cultures are used whilst they are extended to 3 days for the micronucleus and/or nucleoplasmic bridge assays (cytokinesis-block micronucleus cytome assay (CBMN Cyt)). (b) premature chromosome condensation by mitotic fusion requires no cell culturing whilst chemically induced methods generally do. (c) the dicentric and other assays on metaphases require mitotic arrest with Colcemid while the CBMN Cyt assay does not. (d) instead the CBMN Cyt assay requires cytokinesis blocking with cytochalasin B. 9.1. CULTURING On receipt of a blood specimen several replicate cultures should be set up. 9.1.1. Choice of culture medium There are several defined culture media which may be employed. All are commercially available and have been shown to be suitable for lymphocyte culture. Media formulated without folic acid, in order to detect inherited fragile sites on chromosomes, should not be used. Certain media (F-10 and RPMI-1640) appear to encourage faster growth than, for example, MEM and TC-199 . Although the numbers of second in vitro metaphase (M2) cells can be determined by fluorescence plus Giemsa (FPG) staining, it is a good policy to use routinely a culture procedure which generally gives a minimal number of M2 cells at
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
48 hours. Medium should be supplemented with L-glutamine, heparin and antibiotics. Penicillin and streptomycin are commonly used (details found in Annex I). Depending on the manufacturer, many media already contain these antibiotics. However, antibiotics may need to be added when diluting the medium to working strength, if concentrated or powdered media are purchased. Some laboratories prefer to use media without antibiotics, in which case aseptic working procedures, including the use of sterile blood specimen tubes, are essential. 9.1.2. Choice of serum Foetal calf or human AB serum should be used. As there may be considerable variations between batches of sera, new consignments should be quality tested for their ability to support cell growth. The serum should be heat inactivated at 56 ± 1°C for 0.5–1 hours in a water bath as this helps to reduce batch variability. It is possible also to grow lymphocytes in serum free medium, and such media are commercially available. 9.1.3. Bromodeoxyuridine Bromodeoxyuridine (BrdU) should be included in the cultures in order to permit fluorescence plus Giemsa (FPG) staining . This thymidine analogue is taken up preferentially into replicating DNA. When one chromatid is bifiliarly and the other one 53 unifiliarly substituted, FPG staining produces a ‘harlequin’ effect in the metaphase chromosome of cells which are in their second or later post-substitution division. (Fig. 22). FIG. 22. A second division metaphase stained by the FGP method, exhibiting differential staining of the sister chromatids; the ‘harlequin’ effect. There is no universally established concentration of BrdU that can be used. The optimum will vary depending on such factors as the thymidine concentration in the particular culture medium employed. A laboratory should experiment for itself to determine a satisfactory level. It is customary to add the BrdU to the culture medium at a concentration such that the concentration in
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the final culture mixture does not exceed about 50 μM (15.4 μg/mL). Above this level there is the possibility of BrdU causing excessive mitotic delay . With fresh (<24 hours) blood specimens, a final culture concentration of about 15 μM is often satisfactory. If blood specimens are delayed in transit so that they are more than 24 hours old, the BrdU concentration may have to be increased to, say, 40 μM in order to achieve reliable FPG staining . It should be noted that BrdU is light sensitive, and therefore the cultures should be prepared in subdued lighting (e.g. a yellow safe light) and then incubated in the dark. It can be helpful to wrap the culture vessels in aluminium foil. When using the BrdU method, an optimized fixation time should be chosen for which a high proportion of analysable cells are at the first division stage. Unfortunately, it is not always possible to predict the optimal fixation time. Differences may be encountered not only because of individual variations but also because of radiation effects on cell cycle times. Highly damaged cells may have a significantly stalled response to mitogenic stimulation. In practice, laboratories culture for a single time, usually 48 hours. The alternative would be to set up a large number of cultures and fix replicates at a range of different incubation times and then select that which contains the highest frequency of cells in first mitosis. This is however time consuming, costly and impractical especially in situations where many patients 54 may need to be evaluated rapidly. Therefore a slight modification to the culture has been proposed. Cytochalasin B (Cyt-B), which is normally used for micronucleus preparations (Section 12) may be added to the metaphase cultures and this enables cells after different cell divisions to be distinguished and,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
for analysis purposes, select only those cells in first mitosis. The Cyt-B is added at 24 hours into the culture period at a final concentration of 2 μg/mL. This technique was first used by Hayata et. al. to identify cells in the first cell cycle, where differentiation was based on the number of chromosomes rather than the harlequin staining of sister chromatids with BrdU. The Cyt-B technique is not widely used for the dicentric assay but has been employed successfully in some dose estimation interlaboratory comparison exercises [118,119]. 9.1.4. Mitogens Several mitogens, mostly plant lectins, are commercially available. In most cases the particular populations of lymphocytes which they stimulate have not been precisely defined. It is recommended that phytohaemagglutinin (PHA), which is the most widely used mitogen, should be employed. Several manufacturers market two versions of PHA, sometimes called types M and P. The more expensive and highly purified material (P) is not necessary for routine whole-blood cultures; some laboratories, however, consider it advisable to use it for culturing isolated lymphocytes. There are other mitogens available, e.g. concanavalin A or pokeweed mitogen, which stimulate particular subsets of lymphocytes. These have applications in certain experimental systems and with non-human cells. None are as broadly acting as PHA, and for biological dosimetry they should not be used. 9.1.5. The cultures Autoclavable glass or sterile, disposable plastic containers may be used. It is common practice to culture in 10 mL round-bottomed disposable tubes. These should be held at about a 45º angle with loosened caps in an incubator at 37°C with 5% CO 2 . It is also possible to culture the cells without a CO 2 incubator but then the caps should be closed. Cells should be incubated at 37.0 ± 0.5°C. The thermostability of the incubator is important, and it
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
is advisable to monitor its performance with, for example, a thermocouple and a chart recorder. Too low a temperature will result in a poor yield, if any, of metaphases after 48 hours. If the temperature is high (38°C, or above), cells will progress more quickly through the cycle so that unacceptably high numbers of second-division metaphases may be present by 48 hours . In a busy laboratory where a communally used incubator may be opened and closed frequently there is a danger that even with fan assistance the temperature of the cultures may fall below the optimum for an appreciable time. An alternative is to incubate in a thermostatically controlled water bath. This provides a more rapid heat transfer to the culture than via air and greater thermal stability throughout the 48 hours. If this method is used, the head space above a 5 mL culture should be at least 10 mL, and gassed with filtered 5% CO 2 in air. The vessels’ lids should then be sealed tight. The water bath should have a lid so that the cultures containing BrdU are in darkness. The culturing methods are based, with modifications, on the techniques originally published by Moorhead et al. and Hungerford . In brief, one may set up cultures with whole blood or with separated lymphocytes. The advantages and disadvantages of the techniques concern the volumes of blood sample supplied, the time taken in setting up a culture and the number of scorable metaphases (higher mitotic index) which result. The criteria for determining mitotic index is found in Annex V. 55 9.1.5.1. Whole blood This method can be used with smaller blood samples (1–2 mL) and, if necessary, can be performed with blood collected from a finger prick. A further advantage is the speed and ease with
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
which cultures can be set up. However, the number of resultant metaphases per microscope slide is generally smaller than with the other methods. The procedure is to add 0.3 mL of whole blood and 0.1 mL of PHA working solution to a vessel containing 4 mL of medium and 1 mL of serum and then to incubate. 9.1.5.2. Separated lymphocytes In this method, an enriched inoculum of lymphocytes is added to the medium. It is suitable for cases where a blood sample greater than 3 mL is available. There are two techniques for producing enriched inocula: (a) Firstly, 0.15 mL of PHA is added to 2 mL of blood, and the mixture is then gently agitated. Blood will agglutinate on the walls of the vessel. Then, 2 mL of serum is added, gently mixed and centrifuged for one minute at 50 g. The supernatant of about 3 mL, comprising serum, plasma and buffy coat, is removed with a syringe, leaving behind most of the agglutinated red cells. It helps to disturb the buffy coat with the tip of the needle while drawing up the supernatant. Use a wide bore needle to minimize sheering stress on the cells. The 3 mL of fluid is sufficient to make two cultures and is divided equally into two vessels, each containing 4 mL of medium. (b) Secondly, lymphocytes may be separated from whole blood by layering onto a sterile Ficoll Hypaque column. Ready-to-use tubes for such lymphocyte separation are commercially available. The tubes are centrifuged and the lymphocyte rich layer is removed. This is washed in phosphate buffered saline and placed in culture. The concentration of viable cells can be established by dye exclusion of a small aliquot counted in a haemocytometer chamber so that the cell concentration in the cultures can be adjusted
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
to an optimum value. This value is likely to vary between laboratories and so should be independently established but it is likely to be in the range 0.5–2.0 x 10 6/mL. A detailed protocol for this has been given by Hayata et al . and McFee et al. [123, 124], who point out that the method is particularly suitable for producing clean preparations with a lot of metaphases. Some laboratories find it better to use separated lymphocyte cultures for FISH analysis and also when preparing slides for scanning with an automated metaphase finder (Sections, 10 and 13.3.1). It is probably unnecessarily complicated for scoring conventional Giemsa staining with a normal light microscope, where method (a) above, or whole-blood cultures, are sufficient. 9.1.6. Mitotic arrest Colchicine or its synthetic analogue, demecolcine (Colcemid) can be used, with the latter being the arresting agent preferred by most researchers. A suitable stock solution will contain 10 μg/mL of Colcemid in physiological saline and, if prepared aseptically and stored at 4°C, will keep for six months. Adding 25–50 μL of this solution to each culture of 5.0 mL (final concentration: 0.05–0.1 μg/mL) should provide a sufficient number of metaphases while avoiding problems of cell toxicity which occur with higher concentrations. Colcemid is usually added 2 or 3 hours before terminating the cultures. A few researchers prefer to add the Colcemid midway through the culture period, i.e. after about 24 hours or in some cases at the start of cultures . This should prevent cells from progressing beyond the first metaphase and is thus an alternative means of avoiding the analysis of M2 cells. It should be noted that early addition of Colcemid could produce excessive contraction of the chromosomes unless the final concentration in the culture is substantially lowered to about 56 0.05 μg/mL
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
. Early addition of Colcemid could allow cultures to be prolonged beyond 48 hours to allow for longer cell cycle times in some individuals, e.g. the elderly. 9.2. FIXATION PROCEDURE Lymphocyte cultures are conventionally incubated for 48 hours, although the exact time may vary between laboratories from 46 to 52 hours. Laboratories should establish the optimum time that normally produces good yields of M1 metaphases with their routine procedure. It is also advisable to fix only some of the replicate cultures at the routine time, leaving the remainder in the incubator. This allows for the possibility of cells from some donors taking longer to reach metaphase, and also offers the opportunity for scoring later cells if a high dose may have caused mitotic delay. On terminating the cultures it is no longer necessary to observe aseptic procedures, and, except where specified, further processing may be carried out at room temperature. However, it is important to maintain safe-handling practices as the blood samples may contain human pathogens. The cultures should be centrifuged and the supernatant removed and replaced by a hypotonic solution (5 mL) of 0.075 M potassium chloride. If the supernatant is to be removed by suction, the centrifuge speed should be 200 g for 10 min. If, however, the supernatant is to be tipped off, a firmer pellet is required (600 g for 3 min), though this can lead to more broken cells. The tubes should be left to stand for approximately 15 min at 37°C but when isolated lymphocytes are used, 3–5 min are enough. It is also possible to add about 1 mL of fixative to the hypotonic solution for 5 to 10 min to minimize cell lyses upon centrifugation. The tubes should then be spun again, the hypotonic solution removed and the cell pellet resuspended in
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
5–10 mL of freshly prepared fixative (3:1 methanol/acetic acid). The fixative should be added slowly, but at a constant rate, while the tube is agitated with, for example, a vortex mixer. This is important since it ensures that the cells are dispersed into a uniform suspension. The cells should then be spun down again and resuspended in three changes of fixative. The cells may, if required, be stored long term in fixative, ideally in a -20°C freezer. Alternatively, slides can be prepared either immediately or the next day, and for short term storage the cell suspension can be kept at 4°C. The final wash of fixative should be removed, leaving a sufficient quantity of it (0.25 mL) to give a suitable volume of suspension for dispensing onto slides. However, the final volume depends on the cell density and can be diluted with more fixative solution if found necessary. Clean and grease free slides should be used. While some manufacturers claim that the slides that they supply are sufficiently clean, many laboratories prefer to make doubly certain and store the slides in a degreasing fluid. This can be a 1:1 mixture of acetone and methanol or a 1:1 mixture of ether and ethanol, or 1% concentrated hydrochloric acid in methanol. When needed, the slides can be dried and polished with clean tissue paper. One should note that better quality paper handkerchiefs are not suitable because they have lanoline added to make them soft. Separation of the chromosomes is improved if the slides are cold and wet. This can be achieved by storing the slides in a freezer, taking them out just prior to use and melting the frost with one’s breath a few moments before dispensing the cells. Alternatively, the slides can be dipped for a few seconds into a beaker
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
of distilled water and ice cubes. Improved wetting of the slides is obtained if some methanol is poured on top of the iced water, but not stirred in. Surface liquid should be shaken from the slide a moment before the cells are dispensed. Experience has shown that spreading of the chromosomes can be strongly influenced by the ambient temperature and relative humidity in the laboratory. Variable quality due to these factors can be overcome by dispensing the cells in a controlled environment cabinet. Cabinets designed specifically for cytogenetics laboratories are commercially available. 57 The cells should be thoroughly suspended in the remaining fixative by bubbling with a pipette and dispensing two or three drops onto the slide. The cells from one culture should be dispensed onto at least two slides and many workers prefer to produce up to ten slides from a culture. Before dispensing all the cells from a culture onto slides, it is a good policy to place one drop of the suspension on a test slide. This enables the concentration of metaphases to be judged, and, if necessary, the remaining suspension can be further concentrated or diluted with fixative. If the appearance of the metaphases on the test slide is poor, i.e. badly spread clusters of chromosomes and an excessive amount of debris, it often helps to add one more wash of fixative, stopper and store the tubes overnight in a refrigerator and then spin down and dispense the cells on the following day. The slides should be allowed to air dry, and this can be speeded up by gentle heating over a hot plate, by placing them in a gentle draught of warmed or ambient air from a fan, or by waving them through a spirit lamp (avoid igniting the fixative). 9.3. STAINING Fluorescence plus Giemsa
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
(FPG) staining is recommended as this permits the analysis to be confined to the first in vitro division metaphases (M1) . However, this method has certain drawbacks which can be overcome by using conventional Giemsa staining, as well as FPG. Many workers have noted considerable variation in the quality of FPG staining between replicate slides and also between different patches on the same slide. The FPG technique is most successful if delayed until a few days (up to five) after the slides are made. The rest of the slides can be put in a box and kept at -20°C before use. The quality is poorer if fresher slides are used and also if the slides are more than two or three weeks old. Storage of FPG stained slides for more than a few weeks before scoring can result in their deterioration. Thus, there is the risk that the images of FPG stained metaphases may not be clear enough for accurate discernment of all aberrations. However, the quality is usually sufficient to determine the relative proportions of M1 cells, which are not differentially stained, and M2 cells which display the harlequin effect. As a positive control that the staining has worked, the batch of slides should also include a few slides prepared from longer (72 hours) cultures known to contain M2 cells. Thus, the recommended protocol is to FPG stain one or more replicate slides from each culture. If the staining is good these may be used for scoring aberrations in the M1 cells. If not, the slides should be used to check the M1/M2 ratio, and aberration analysis should be done with replicate slides from the same culture which have been stained with Giemsa provided that the level of M2 cells is less than 5%, as assessed by FPG. If
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the level is higher, this may require an adjustment of the aberration yield which could introduce some extra error. This would require certain assumptions regarding, for example, the proportion of dicentrics in M2 cells which are still accompanied by an acentric fragment. As stressed earlier, it is a better policy to adopt a culture method which usually results in few M2 cells although of course this cannot be predicted for any individual because people behave differently in their lymphocytes’ stimulation and proliferation capacity . Adaptation to alternative culture techniques such as culturing with Cyt-B with Colcemid or with early Colcemid may provide an easier and faster alternative to FPG staining [117–119, 125]. These techniques may be of particular use in triage scenarios where rapid dose estimates are required. 9.3.1. Pretreatment A pre-treatment of slides with RNase A, prior to staining, can remove residual stainable cytoplasmic material . This is an optional procedure that can provide much clearer images of the chromosomes for scoring block stained, harlequin stained or banded preparations. Additionally, it has proved useful for slides assessed with automatic image analysis systems. 58 The protocol is as follows: A stock solution of 10 mg/mL RNase A in Tris EDTA buffer is heated for 10 min at 70°C and then allowed to cool slowly. Aliquots may be stored for several years at -20°C. Slides are rinsed in distilled water and placed in 0.5 mg/mL RNase A solution (stock solution: distilled water 1: 20) for 10 min at 37°C. This may be done either in a prewarmed staining jar or, to be more economical, a smaller volume of the diluted stock solution can be placed on the slide beneath a coverslip. The slides are then washed in distilled water, placed in 3:1 methanol: acetic acid fixative for 2 min, dried and
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
stained as described below. The RNase A cleaning procedure can also be used after destaining old slides or on micronucleus preparations. For these applications, concentrations and times may vary . 9.3.2. Fluorescence plus Giemsa (FPG) staining This method is derived from that published by Perry and Wolff with some modifications. About ten drops of Hoechst 33258 stain (0.5 μg/mL in pH 6.8 phosphate buffer) should be placed on the slide beneath a coverslip, ensuring that no air bubbles are trapped. At this point workers with a fluorescence microscope can, if they wish, make a quick check of the M1/M2 ratio using Latt’s method , which produces a harlequin effect, but which fades very rapidly. Otherwise the slides can be illuminated under a 20 W UV lamp (>310 nm) for 0.5 hour or, alternatively, a 30 W fluorescent strip lamp for about 1.5 hours. After careful removal of the coverslips, the slides should be washed well with pH 6.8 phosphate buffer. At this point some workers put the slides into 2 x SSC (0.3M sodium chloride and 0.03M trisodium citrate) at 60°C for about 20–30 min. Experience has shown that this SSC stage can be omitted if it results in an undesirable swelling of the chromatids which makes microscope analysis more difficult. The use of 2 x SSC, however, removes some cellular debris and so leads to cleaner preparations. The slides are then washed in distilled water, followed by immersion in Giemsa stain (5–10% in pH 6.8 buffer, Gurr R66) for 3 to 5 min. They are then rinsed in the buffer, then with distilled water and allowed to dry. The slides can be examined at this stage under the microscope or cleared and mounted beneath a coverslip. 9.3.3. Conventional Giemsa staining The slides should be immersed in 2% (Gurr
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
R66 improved) Giemsa stain in pH 6.8 phosphate buffer for 5 min, washed in buffer, briefly rinsed in distilled water and allowed to dry, finally mounted with a cover glass using a mounting medium. Figs 10 and 11 show Giemsa-stained metaphases. It is possible to modify the staining specifically to highlight centromeres although for experienced scorers this is not normally necessary . Such highlighting can be achieved by FISH, using a pancentromeric probe (Fig. 27) or with Giemsa stain using the C-banding method (Fig. 23). 59 FIG. 23. A metaphase stained with Giemsa by the C-banding method which highlights centromeres. The C-banding protocol is: (1) Place the slides in 0.2 N hydrochloric acid at room temperature for 30 min. (2) Wash three times in distilled water. (3) Place slides in 5% barium hydroxide at 60 oC for 1 min. (4) Wash in 0.2 N HCl for 2 min. (5) Wash in distilled water for 2 min. (6) Place in 2x SSC at 60 oC for 45 min. (7) Wash in distilled water. (8) Allow to air dry and stain in 2% Giemsa in pH 6.8 phosphate buffer for 10 min. (9) If the staining intensity is insufficient, the slides can be reimmersed in the Giemsa stain for a further 5–10 min. 9.4. ANALYSIS OF SLIDES 9.4.1. Conventional microscopy The slides should be coded to prevent bias in the scoring and should be scanned methodically so that the entire area is covered. The scanning should be done at low magnification (about x 100 to x 200). At this level, it is not possible to count whether all the chromosomes are present, nor is it possible to detect aberrations. However, with practice the scorer can identify those spreads which have about 40 or more pieces and an appearance which is likely at
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
higher magnification to be of analysable quality. It is important that this initial scanning be done at a magnification low enough to prevent a bias towards selecting cells 60 which contain aberrations. Having found a likely metaphase, the scorer should switch to high magnification (about x 1000 to x 2000), ignore, if possible, the presence of any aberrations and make a snap judgement on whether the chromosomes are of a quality suitable for scoring. This will be based on the sharpness of the images and the amount of twisting and overlapping of chromosomes. With FPG stained material the cell should be rejected if it displays the harlequin effect, indicating that it is not an M1 spread. If the decision is taken to analyse the spread, then the number of individual chromosome pieces should be counted and the presence of aberrations noted. It is recommended that only complete metaphases be recorded, i.e. those with 46 centromeres. If the cell contains unstable aberrations, then it should balance. For example, a spread containing a dicentric should also have an acentric fragment, yet still count to 46 pieces. By contrast, a centric ring will also have an accompanying fragment, but the total number of objects in the cell will count to 47. Each excess acentric, i.e. one not associated with a dicentric or centric ring, will increase the count of pieces beyond 46. When recording the aberrations, the fragments associated with a dicentric or ring must not be included with the count of excess acentrics. When high radiation doses are involved there may be more than one aberration in the spread, but the pieces should still balance. Tricentric aberrations are equivalent to two dicentrics and should have two accompanying fragments, while quadricentrics will have three fragments, and so on. All abnormalities in the cell
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
should be recorded, although for dosimetry purposes only the data on dicentrics, or dicentrics plus rings will normally be used. The x and y stage co-ordinates of all complete cells analysed, including those free from aberrations, should be recorded for possible future reference. 9.4.2. Computer assisted microscopy Metaphase finding by automated pattern recognition systems has been introduced into many labs and several commercial systems are available. These instruments also include semi-automated analysis of digitized images that assist with locating aberrant chromosomes. However, no system is fully automatic; all incorporate steps where the operator’s judgement and decision are required. Use of these instruments should be such that the same recommended criteria as outlined above are maintained, namely, selection of candidate metaphases for scoring should not introduce bias likely to distort aberration yields and only complete spreads of chromosomes should be scored. Automated cell finding and scoring systems are discussed in detail in Section 13. 9.5. RECORDING OF DATA Good laboratory practice requires that a unique identifier code or labelling system be devised for specimens, slides and associated paperwork. The receipt and processing of specimens, whether for experiments or for overdose investigations, should be recorded in a laboratory diary. Electronic systems for data storage and handling are available (see Section 13.4). However, many researchers still work by recording their microscope observations onto a score sheet and most laboratories have evolved their own preferred way of recording the data. Electronic systems can have the data stored and displayed in a variety of ways to suit the laboratory. It is important that the primary data comprising the observations made on every cell can be retrieved so that later on all possible compilations and aggregations of data can be made. Table 6 illustrates a simple layout of a data sheet for recording aberrations. 61 TABLE
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
6. LAYOUT OF A DATA SHEET FOR RECORDING ABERRATIONS Slide No: Scorer: Microscope No: NDate: Cell No. Stage coordinates No. of chromosomal pieces Dicentrics Centric rings Excess acentrics Remarks x y1 100.1 1.2 46 2 103.4 1.5 47 1 13 105.4 1.2 49 2 1 24 112.4 1.6 46 5 112.7 1.8 48 26 120.1 1.2 46 17 122.7 1.5 47 18 124.1 1.4 45 Chromatid exchange 9 126.8 1.7 46 2* *= 1 tricentric etc. From the information on this sheet any cell can be identified for re-examination on a future occasion. Using the conventional Giemsa staining technique, data on unstable aberrations are most important for biological dosimetry, although no attempt has been made to list separately the minutes, fragments and acentric rings. This is because accurate discrimination between them is not always possible. However, if it is preferred, they could be recorded as M, F and AR, respectively, instead of numerals in the column headed ‘Excess acentrics’. The Remarks column can be used to record other abnormalities, e.g. numerical aberrations, stable chromosome damage or chromatid aberrations. Any other numerical information which may be required, such as the percentage of cells with damage, or distributions of aberrations among the cells, can be easily extracted from the sheet. 9.6. STORAGE OF INFORMATION AND SLIDES Clearly, research data have to be filed and stored for future reference. It is worth emphasizing that files relating to overdose cases may need to be re-examined long afterwards. In the event of a person developing a malignancy, perhaps decades later, the case may be reopened to resolve a claim for compensation. Most laboratories would wish or are obliged to store the microscope slides as well, and this can create some problems. Conventionally Giemsa stained preparations have a tendency to fade and FPG stained material creates
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
more difficulty as it frequently fades after several months. It is advisable to keep the stained slides in a box in a dry place at room temperature. However, faded slides can be retrieved by carefully soaking off the coverslip and restaining 62 with conventional Giemsa. Attempts to restain with FPG will not succeed. Stored replicate slides, kept at -20°C, that have never been stained can also be stained with conventional Giemsa many years later. It is also good practice to store surplus fixed cells from overdose investigations. For ease of storage they can be concentrated down into small (2 mL) ampoules and kept at -20°C. Slides made from this material can, years later if required, be stained conventionally, with FPG or FISH. 9.7. DOSE ASSESSMENT 9.7.1. Choice of curve The sources of radiation to which personnel are usually exposed are gamma-, X-rays and, occasionally, degraded neutrons. It is commonly found that there is a difference between the yield curves of X and gamma rays, particularly at low doses (<0.5 Gy). Therefore it is advisable to have a calibration curve for a suitable energy of X rays (e.g. 200–250 kVp) as well as for either 60 Co or 137 Cs. In general most research laboratories have more ready access to a 60 Co source rather than 137 Cs. For neutrons a degraded energy spectrum is similar to a fission spectrum. Available evidence indicates that the dose response curves for fission spectrum neutrons are linear and do not alter much with neutron energy. Thus one calibration curve produced with a fission spectrum would suffice. In industrial radiography 192 Ir is commonly used and its gamma energy is on average 400 keV. Few laboratories have access to this isotope to produce a calibration curve which should lie somewhere between the X and 60
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Co / 137 Cs gamma ray curves. However, it is generally considered to lie closer to the latter, and so it is recommended that the gamma ray curve be used. 9.7.2. Number of cells to be analysed In order to produce a dose estimate with a statistical uncertainty small enough to be of value, a large number of cells usually needs to be scored. The decision on how many to analyse is a compromise based on the importance of the case, the available labour and the quality of the preparations. For example, after exposure with a dose of several Gy and higher, the subject’s lymphocyte count may be severely depleted and this will be reflected in a low number of metaphases on the slides. However, as the number of aberrations per cell will be high, a reasonable estimate could be made from the analysis of just a few tens of cells. Consideration of the dose uncertainty versus number of cells scored is important when deploying the dicentric assay as a triage tool for rapid assessment after a mass casualties event. This topic is considered more fully later (Section 14). For lower doses, where the number of available cells is not the limiting factor, a dose estimate could be based on about 500 cells. This may require 2–3 person-days at aconventional microscope, although in an emergency several people can collaborate in scoring replicate slides. For a low or zero dicentric yield, the confidence limits resulting from 500 scored cells are usually sufficient. The decision to extend scoring beyond 500 to 1000 or more cells depends on whether there is evidence of a serious overexposure justifying an extended analysis, or if the continued employment of a radiation worker is in jeopardy. Clearly, there is no single number of cells that can be
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
recommended as being applicable in all cases. However, as a general rule it is suggested that 500 cells or 100 dicentrics should be scored in order to give a reasonably accurate estimate of dose. Table 7 shows the limits calculated using this method for several dose estimates up to 1.0 Gy. 63 TABLE 7. THE EFFECT OF INCREASING THE NUMBER OF CELLS EXAMINED ON THE LOWER AND UPPER 95% CONFIDENCE LIMITS FOR FOUR ESTIMATES OF ACUTE GAMMA DOSE (based on the curve shown in Fig. 24) > Dose estimate, mGy Confidence limits No. of cells examined 500 1000 100 Upper 320 245 Lower < 0 16 250 Upper 448 380 Lower 111 141 500 Upper 677 627 Lower 333 383 1000 Upper 1178 1127 Lower 830 881 A simple method for calculating confidence limits on dose estimates is discussed in the following section, 9.7.3. 9.7.3. Uncertainty on dose estimates While there is no difficulty in deriving a dose from a measured yield of dicentrics, there are a number of different ways in which the uncertainty on the yield can be derived. The aim is to express uncertainty in terms of a confidence interval and it is standard practice to calculate 95% limits. The 95% confidence limits define an interval that will encompass the true dose on at least 95% of occasions. The difficulty in the computation of confidence limits arises because there are two components to the uncertainty: one from the Poisson nature of the yield of aberrations, seen in the sample from the overexposed subject, and the other from uncertainties associated with the calibration curve which are approximately normally distributed. The problem has been discussed in the literature by Savage et al. , Merkle and more recently, Sz łui ńska et al. . The simplest solution was proposed
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
by Merkle, it allows both the Poisson error on the yield and the errors on the calibration curve to be taken into account. Merkle’s approach, illustrated in Fig. 24, involves the following steps: (1) Assuming the Poisson distribution, calculate the yields corresponding to the lower and upper 95% confidence limits on the observed yield (Y L and Y U). (2) Calculate the dose at which Y L crosses the upper curve. This is the lower confidence limit (D L). (3) Calculate the dose at which Y U crosses the lower curve. This is the upper confidence limit (D U). 64 0.1 0.09 0.08 0.07 0.06 0.05 0.04 0.03 0.02 0.01 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 100 FIG. 24. A dose–response calibration curve with its 95% confidence limits, used to estimate uncertainties. Example : Five hundred cells were analysed and 25 of them were observed each to contain one dicentric. This gives a yield (Y) of 0.05 dicentrics / cell and the dispersion index and the u test of 0.95 and -0.78 respectively. Dose was estimated using the dose-effect curve for 60 Co shown in Fig. 23 for which coefficients, variances and covariances are listed below. C = 1.28E-3 α = 2.10E-2 β = 6.31E-2 var C = 2.22E-07 var α = 2.66E-05 var β =1.61E-05 covar (C, α) = -9.95E-07 covar (C, β) = 4.38E-07 covar ( α,β) = -1.512E-05 (1) Since the dose-effect curve is linear-quadratic (Y= C+ αD+ βD2 ), the estimated dose which is 0.73 Gy is obtained by solving the equation: # ( )ββαα 242 CYD −++−= (6) (2) Y L and Y U are obtained from standard statistical tables of confidence limits for the expectation of a Poisson variable . Table 8 shows the 95% limits for values of 65 observed
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
dicentrics from 0 to 103. For the 25 dicentrics observed in this example the Y Lis 16.768/500 = 0.034 and Y U is 36.03/500 = 0.072. TABLE 8. THE POISSON UPPER AND LOWER 95% CONFIDENCE LIMITS ON OBSERVED NUMBERS (X) OF DICENTRICS (adapted from ) X Lower Upper X Lower Upper X Lower Upper X Lower Upper 0 0 3.285 26 16.77 37.67 52 38.165 66.76 78 61.9 96.06 1 0.051 5.323 27 17.63 38.165 53 39.76 68.1 79 62.81 97.545 2 0.355 6.686 28 19.05 39.76 54 4.094 69.62 80 62.81 99.17 3 0.818 8.102 29 19.05 40.94 55 40.94 71.09 81 63.49 99.17 4 1.366 9.598 30 20.335 41.75 56 41.75 71.28 82 64.95 100.32 5 1.97 11.177 31 21.36 43.45 57 43.45 72.66 83 66.76 101.71 6 2.613 12.817 32 21.36 44.26 58 44.26 74.22 84 66.76 103.315 7 3.285 13.765 33 22.945 45.28 59 44.26 75.49 85 66.76 104.4 8 3.285 14.921 34 23.76 47.025 60 45.28 75.785 86 68.1 104.58 9 4.46 16.768 35 23.76 47.69 61 47.025 77.16 87 69.62 105.905 10 5.323 17.633 36 25.4 48.74 62 47.69 78.73 88 71.02 107.32 11 5.323 19.05 37 26.31 50.42 63 47.69 79.98 89 71.09 109.11 12 6.686 20.335 38 26.31 51.29 64 48.74 80.25 90 71.28 109.61 13 6.686 21.364 39 27.735 52.15 65 50.42 81.61 91 72.66 110.11 14 8.102 22.945 40 28.97 53.72 66 51.29 83.14 92 74.22 111.44 15 8.102 23.762 41 28.97 54.99 67 51.29 84.57 93 75.49 112.87 16 9.598 25.4 42 30.02 55.51 68 52.15 84.67 94 75.49 114.84 17 9.598 26.306 43 31.675 56.99 69 53.72 86.01 95 75.785 114.84 18 11.177 27.735 44 31.675 58.72 70 54.99 87.48 96 77.16 115.605 19 11.177 28.966 45 32.28 58.84 71 54.99 89.23 97 78.73 116.93 20 12.817 30.017
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
46 34.05 60.24 72 55.51 89.23 98 79.98 118.35 21 12.817 31.675 47 34.665 61.9 73 56.99 90.37 99 79.98 120.36 22 13.765 32.277 48 34.665 62.81 74 58.72 91.78 100 80.25 120.36 23 14.921 34.048 49 36.03 63.49 75 58.72 93.48 101 81.61 121.06 24 14.921 34.665 50 37.67 64.95 76 58.84 94.23 102 83.14 122.57 25 16.768 36.03 51 37.67 66.76 77 60.24 94.705 103 84.57 123.77 (3) The lower and upper 95% confidence limits of the curve can be calculated by the equation: 32422 ),(cov 2),(cov 2),(cov 2var var var Dar DCar DCar DDCRDDCY βαβαβαβα +++++±++= (7) where: R2 is the regression confidence factor, and is the 95% confidence limit of a chi-square distribution, χ2 (df, 95%), with 2 or 3 degrees of freedom (df). For a linear-quadratic curve (df = 3) R 2 is 7.81, and for a linear curve is 5.99. In Eq. (7) a value of 2.79 should be used, in the case of a linear curve the value used should be 2.45. Because in both, dicentric yield observation and calibration curve the 95% confidence limits 66 are considered, some authors have proposed to use an 83% confidence limit of the regression curve instead of 95%, in order to reduce a possible overestimation of the uncertainty [134, 135]. In this case R 2 will be 5.02 for the linear-quadratic curve and 3.54 for the linear curve. (4) The calculation of the point where Y L intercepts the upper confidence curve, which is the lower 95% confidence limit of the dose estimated (D L), can be done by iteration. The Excel program contains a tool ‘Solver’ that can be used. In the same way the point where Y U intercepts the lower confidence curve (D U) can be obtained. Using the present example D L
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
and D U are 0.51 and 0.97 Gy respectively. If covariances are not available, the confidence limits can be approximated by Eq. (8). This equation is valid as the contribution to uncertainty from the covariances is comparatively small. 422 var var var DDCRDDCY βαβα ++±++= (8) With well-established calibration curves based on a large amount of scoring, the variance due to the curve is small compared with the variance on the observed yield from the subject and can be ignored. A simpler approximate estimate of D L and D U may be obtained directly from the calibration curve, by considering where Y L and Y U cross the solid line in Fig. 25. FIG. 25. A dose–response calibration curve used to estimate uncertainties ignoring the error due to the curve. 67 In the present example at 0.73 Gy the error associated with the curve is 0.002; this value is obtained by inserting 0.73 Gy for D in the last term of Eq. (7). The value obtained is smaller than the SE associated with the observed yield of dicentrics (25) 1/2 /500 which is equal to 0.01. With this approach D L and D U are 0.57 and 0.91. If the u test statistic is higher than 1.96, Y U and Y L should be corrected to consider the overdispersion by multiplying by the factor indicated below, where CL is the Poisson confidence limit indicated in standard tables, X the number of dicentrics observed and σ2 /y the observed dispersion index: > y XCL Factor > 2 > σ ⎟⎠⎞⎜⎝⎛= (9) Using the above example, if instead of 25 cells with one dicentric, 19 cells with one dicentric and three cells with two were observed then the σ2 /y will be 1.19, and the u value 3.19. In this case Y
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
U and Y L are: 1074 025 03 36 500 03 36 19 1 ...Y.U =⎟⎠⎞⎜⎝⎛×= (10) 0217 025 77 16 500 77 16 19 1 ...Y.L =⎟⎠⎞⎜⎝⎛×= (11) Using these values D L and D U are 0.39 and 1.19 Gy respectively. 9.7.4. Extension of dose calculations for more complex exposure scenarios The previous section applies to cases where a large acute accidental overexposure to relatively low LET radiation is uniformly distributed over the whole body and a blood sample is available promptly. The dicentric frequency per cell assessed against an appropriate acute in vitro dose response curve provides a reliable estimate of the average whole body absorbed dose. In practice, however, such ideal circumstances rarely occur and protracted or fractionated irradiations are common. It is more usual for accidental exposure to be non-uniform, perhaps involving only part of the body. A substantial time delay may also occur before a blood sample is taken for chromosome study. These factors will result in an inhomogeneous population of lymphocytes being sampled, and the resultant dicentric yield, when compared with a standard in vitro dose–response curve, will produce an unrealistic estimate of dose. Inhomogeneity produces a yield of dicentrics which does not conform to a Poisson distribution, but is generally overdispersed. For a partial body exposure this obviously arises because those lymphocytes in tissues outside the radiation field will not be damaged. In cases of highly localized exposure, the smaller than expected number of cells that are damaged may each contain several aberrations. Even when the radiation dose is uniform at the skin, its monotonic reduction with depth in tissue will result in a variety of doses being received by lymphocytes. This effect will be especially marked with weakly penetrating radiation, but for more strongly penetrating radiation, such as 250 kVp X
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
rays or gamma rays from 60 Co, 192 Ir and 137 Cs sources, the effect is sufficiently small for the dicentrics to have an approximately Poisson distribution. Accidental exposure to high LET radiation such as neutrons will also produce an overdispersed distribution because of the manner in which the dose is deposited at the cellular level (see Section 3). 68 Delays in blood sampling will influence the aberration yield, as cells containing unstable aberrations are lost from the circulation and replaced by newly produced, cells that contain no dicentrics. In this Section it is intended to discuss how the yield of chromosome aberrations is influenced by inhomogeneity of exposure, by delayed sampling and by protracted exposure and how the data might nevertheless be used to provide a meaningful estimate of dose. Reporting on emergencies involving very low doses that, because of statistical limitations, may be difficult to distinguish from zero dose is also considered. The following Section then contains worked examples for each of the emergency exposure situations. 9.7.4.1. Criticality Accident In a criticality accident the body is irradiated by both neutrons and gamma rays. If the ratio of neutron to gamma ray doses is known, and this information is usually available from physical measurements, it is possible, to estimate the separate neutron and gamma ray doses by iteration. The iteration process is implemented as follows: (1) Assume that all the aberrations are attributable to neutrons, and from the measured yield of dicentrics estimate a dose from the neutron curve; (2) Use the estimated neutron dose and the supplied neutron to gamma ray ratio to estimate the gamma ray dose; (3) Use the gamma ray dose to estimate the yield of dicentrics due to gamma rays; (4) Subtract this calculated gamma ray yield of dicentrics from the measured yield to
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
give a new value for the neutron yield; (5) Repeat steps 1 to 4 until self-consistent estimates are obtained. In the case where a physical estimate of the ratio of neutron: gamma dose is not available, the above method is not possible. One approach would be to express the dose in Gy-Eq as was done for the Tokai-mura accident victims. However, Brame and Groer have described a Bayesian approach to dose estimation in a criticality accident which allows estimation in the absence of the ratio estimate. The Bayesian method was found to give very similar results to the classical iterative approach in a simulated accident situation . 9.7.4.2. Low dose overexposure cases It is often stated that the lower limit of dose detection by dicentrics for low LET radiation is around 0.1–0.2 Gy. Sensitivity to low doses is a function of the background level of dicentrics (which for the general population is on the order of ~ 0.5–1/1000 cells) and the limit on the number of metaphases that can realistically be scored. Dose estimates at low doses therefore carry large statistical uncertainties. As discussed, these come mainly from the Poisson error in the yield but with a small contribution from the SE on the coefficients of the dose–response curve, of which α is the most important at low doses. For practical purposes the latter can be ignored unless the calibration data at low doses are sparse. Whilst 100-200 mGy is of minor concern when considering health consequences of exposure, in legal terms it is a high dose when compared with the ICRP recommended annual occupational dose limit of 20 mSv. There is often pressure on cytogenetics to try to resolve suspected low overdoses, perhaps pushing the method beyond its capabilities. 69 When reporting results, experience has shown that lay
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
persons rarely understand the concept of uncertainty. There are a number of approaches that can be used to aid interpretation of results. Firstly, although it is not strictly statistically accurate, it can be explained that there is only a 2.5% chance of the dose being greater than the upper 95% confidence limit. Additionally, the lower confidence limit can be used to define the ‘detection limit’ of the assay: using the figures in Table 9, a dose statistically greater than zero Gy would only be indicated by 4 or more dicentrics in 1000 cells, i.e. when the lower confidence limit is greater than zero Gy. TABLE 9. 95% DOSE CONFIDENCE LIMITS ON VARIOUS LOW YIELDS OF DICENTRICS IN 1000 CELLS AND THE ODDS RATIOS SHOWING THE LIKELIHOODS OF ZERO DOSE OR 0.25 Gy (doses calculated using Y = 0.0010 + 0.0164D + 0.0492D 2 ) > Observation (dicentrics) Dose (Gy) Odds ratio p(0 Gy):p(0.25 Gy) Lower confidence limit Mean Upper confidence limit 0—00.12 1306 : 1 1000.18 160 : 1 200.05 0.23 20 : 1 300.09 0.26 2 : 1 40.01 0.13 0.30 1 : 3 50.03 0.16 0.33 1 : 28 60.06 0.19 0.36 1 : 229 70.09 0.22 0.38 1 : 1868 Alternatively, if one considers just two possible scenarios: zero dose or the suspected (e.g. badge) dose, the relative probabilities of each can be used to calculate the odds ratio for the two doses. The chances are derived from the Poisson distribution, as follows. If the dose was zero, then from the dose-response curve, the background frequency of 1 dicentric in 1000 cells is expected. For a dose of 0.25 Gy a yield of 8.2/1000 is expected. From the Poisson distribution, the relative chance of seeing no dicentrics when 1 and 8.2 are expected is e – 1
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
/e –8.2 , which is 0.36788/0.00027 which is approximately 1300. The values of mean, lower and upper confidence intervals on dose, plus the odds ratio for zero dose: badge dose, are shown in the top line of Table 9, with other values below that would have been quoted if different numbers of dicentrics had been seen in 1000 cells. The reporting laboratory may use either or both approaches when presenting the results of the analysis, the decision depending on the particular circumstances of the case. 9.7.4.3. Partial body exposure The cytogenetic indication of a partial body exposure is a non-Poisson distribution of dicentrics among the patient‘s scored metaphases. Therefore the first step is to calculate the ratio of variance to mean ( σ2/y) and then use the u-test to determine if the ratio deviates significantly from unity (see Section 8.3). If the data are consistent with a Poisson distribution 70 the recommendation is to report an averaged whole body dose estimate. If the data are non-Poisson two methods, see below, have been proposed whereby an estimate of partial body dose may be derived rather than simply quoting the averaged whole body value. The incentive for determining an estimate of partial body dose may also be based on information about the circumstances of the overdose event. The u-test is the recommended method to assess Poisson distribution of data but many consider that this test is not very robust, especially in cases when low number of metaphase spreads are scored. Hence it is routine practice in many laboratories to also compare dose estimates for whole body as well as partial body using either of the two methods described below to data. If the whole body dose estimate is significantly different than the partial-body dose estimate, then the laboratory should consider the case
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
as potentially a partial-body exposure scenario. In cases of uncertainties in whether there are significant differences between the whole body and partial body dose estimates (Section 9.7.3), then the recommendation is to use the two methods described here only when the data are significantly non-Poisson. Method 1 This method was first proposed by Dolphin and is termed the contaminated Poisson method. It considers the overdispersed distribution of dicentrics among all the scored cells. The observed distribution is considered to be the sum of (a) a Poisson distribution which represents the irradiated fraction of the body and (b) the remaining unexposed fraction. Cells containing aberrations will obviously have been in the irradiated part of the body. Undamaged cells will comprise two subpopulations: those from the unexposed fraction and irradiated cells which received no damage (representing the first term (e –Y ) of the Poisson series). Eq. (12) describes the distribution of the damage in the cells: 01 nNXeYYF −=− (12) where: Y F is the mean yield of dicentrics in the irradiated fraction, e–Y represents the number of undamaged cells in the irradiated fraction, X is the the number of dicentrics observed, N is the total number of cells, and n0 is the number of cells free of dicentrics. Eq. (12) can be solved by iteration to find the maximum likelihood estimate of the yield, and Y F can then be used to calculate the fraction, f, of cells scored which were irradiated, using Eq. (13): N XfY F = (13) The dose to the irradiated fraction can then be calculated using Y F and the appropriate calibration curve. The size of the fraction of body irradiated may be derived from f after correction for the effects of interphase death and mitotic delay. These factors will cause irradiated cells, even
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
if free from aberrations, to be less likely than unexposed cells to reach metaphase by 48 hours in culture. If the fraction of irradiated cells which reach metaphase was p, the fraction of the body irradiated, F, is given by pffpfF +−= 1 (14) 71 The p value is estimated by the equation P = exp(-D/D o ) (15) where: D is the estimated dose, and there is experimental evidence of Do values between 2.7 and 3.5 [139, 140]. There are, however, a number of limitations to this approach: (1) The method assumes that the exposure to the irradiated fraction is homogeneous. (2) It derives the fraction of lymphocytes irradiated which can only be related to the fraction of body irradiated by making the simplifying assumption that lymphocytes are uniformly distributed throughout the body. (3) It requires a sufficiently high local dose so that there are a number of cells observed with two or more dicentrics. This is necessary for the best-fit calculation of the irradiated, but undamaged, cells. (4) The method assumes a minimal delay between irradiation and blood sampling, so that the dicentric yield is not significantly diluted by newly formed undamaged cells entering the circulation. Should dilution occur, then the fraction irradiated derived by this method is likely to be underestimated . Method 2 This approach has been proposed by Sasaki and Miyata and is termed the Qdr method. It considers the yield of dicentrics and rings only from those cells that contain unstable aberrations and assumes that these cells were present at the time of the accident. The method therefore circumvents problems of dilution by undamaged cells from an unexposed fraction of the body or post-irradiation replenishment from the stem cell pool. It also does not require the presence of heavily damaged cells containing two
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
or more aberrations. Qdr is the expected yield of dicentrics and rings among the damaged cells, N U, and is given by 2111 > YYU eYNXQdr −−== (16) where: X is the number of dicentrics and rings, and Y 1 and Y 2 are yields of dicentrics plus rings and of excess acentrics, respectively. As Y 1 and Y 2 are known functions of the dose and are derivable from in vitro dose– response curves, Qdr is a function of dose alone and hence permits a dose estimate to be made for the irradiated part of the body. There also are several limitations with this method: (1) It assumes, as does method 1, that the exposure to the irradiated fraction is uniform, but according to Sasaki and Miyata it provides no information on the size of this fraction. However, this can be derived, using essentially the same procedure as in method 1, by converting dose to yield and then using Eqs (13) and (14). (2) It assumes that the excess acentric aberrations also have Poisson distributions, but this is not borne out by data from in vitro experiments. If this limitation is thought to be 72 important, it could be avoided by considering the yield of dicentrics and rings in those damaged cells that contain just dicentrics and rings. Eq. (16) would now reduce to 111 > YU eYNXQdr −== (17) which is identical with Eq. (12). This simplified form will produce a dose estimate identical with that obtained by method 1 above. (3) The method assumes that all cells containing unstable aberrations were present at the time of irradiation and that there has been no recruitment of cells containing derived chromosome aberrations arising from chromatid damage in stem cells. 9.7.4.4. Delayed blood sampling It has been well documented
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
that some lymphocytes containing aberrations continue to exist in the peripheral circulation for many years after an irradiation. However, a delay of more than a few weeks between irradiation and sampling has been shown to reduce the aberration yield. This is particularly apparent following large doses that are sufficiently high to cause early deterministic reactions such as the depression of white blood cell counts. For lower doses, below the threshold for deterministic effect, the potential for late recognition of an overdose is greater. Therefore, some adjustment needs to be made in order to produce a more realistic estimate of dose. Unfortunately, there are few data which enable a reliable correction factor to be deduced. Indeed, since there is marked individual variation, depending on factors such as infections, the depression of aberration yield probably cannot be expressed simply as a function of time alone. Nevertheless, an exponential disappearance rate with a half-time of about three years has been suggested . As a general approximation this seems suitable when the sampling delay is long, say five or more years. However, when brief accidental exposures are being investigated there are rarely delays of this length. Typically, they range from a few days to a few weeks. Delay of a few weeks is likely if the exposure is only appreciated when a routine personal dosimeter is processed, with irradiation having occurred early in its period of issue. At most one might encounter a sampling delay of up to one year, and over this time span an exponential disappearance half-time of about three years is inappropriate. What is probably the most comprehensive body of data is that published by Buckton et al. [63, 143, 144] who, for over 30 years, repeatedly sampled a group of patients treated with fractionated X rays for ankylosing spondylitis. In
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
these studies there was a long initial plateau in aberration yield, lasting about 20 weeks, which was followed by a steep fall which persisted over four years. Over the first four years they calculated that the dicentric yield dropped at a rate of about 43% per year and thereafter the decline was about 14% per year. In view of the considerable variability in the limited data, no firm guidance can be given, especially for delays in excess of a few weeks. Uncorrected dicentric yields will, therefore, probably underestimate the dose, but the extent of the underestimate depends on generally unquantifiable factors particular to each individual. It was noted in the discussion of partial body irradiation (Section 9.7.4.3 above) that the Qdr method considers the yield of dicentrics and rings only in damaged cells. Therefore, applying this approach to delayed blood sampling could also avoid the problem of dilution with time by undamaged cells entering the circulation, provided that sufficient numbers of cells containing unstable aberrations are still observed. This is obviously not feasible for very long delays. In such cases it may be possible, however, to consider the persistence of cells with stable aberrations. For many years this was only possible by karyotyping many block-stained 73 and later banded preparations. By these methods the study of ankylosing spondylitis revealed that the level of these cells remained more or less constant over the 30 years of follow-up. Awa has also reported a good correlation between the frequency of stable aberrations and the DS86 estimates of dose in the atom bomb survivors. Dividing cells containing unstable aberrations are selectively eliminated by mitotic non-disjunction. The excess of stable aberrations with time is explained by assuming that cells with stable and unstable damage disappear at the same rate, but the loss of stable
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
damage is offset by unimpeded divisions from the stem cell pool. Laborious banded karyotyping has now been replaced by FISH as the optimum method for screening large numbers of cells for the presence of rare, random, non-constitutional stable translocations for retrospective biological dosimetry. This is described in Section 10. 9.7.4.5. Protracted and fractionated exposure Protraction or fractionation of the exposure may also produce a lower chromosome aberration yield than if the same dose is received acutely. For high LET radiation, where the dose–response relationship is close to linear, no dose rate or fractionation effect would be expected. For low LET radiation, however, the effect of dose protraction is to reduce the dose squared coefficient, β, in the yield Eq. (2). This term represents those aberrations, possibly of two track origin, which can be modified by repair mechanisms that have time to operate during the course of a protracted exposure or in the periods between intermittent acute exposures. A number of studies have shown that the decrease in the frequencies of aberrations appears to follow a single exponential function with a mean time of about 2 hours. The majority of lesions that are converted into chromosome aberrations will have been repaired or would become otherwise unavailable for interactions within about five to six hours after exposure. A time dependent factor known as the G function was proposed by Lea and Catcheside to enable modification of the dose squared coefficient and thus allow for the effects of dose protraction. The linear quadratic Eq. (2) may be modified, as shown in Eq. (18): 2)( DxGDCY βα ++= (18) where ]1[2)( 2 > x exxxG −+−= (19) and 0ttx = (20) where: t is the time over which the irradiation occurred, and t 0 is the mean lifetime of the breaks, which has
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
been shown to be on the order of ~ 2 hours [96, 147]. Therefore, in the case of continuous irradiation, it is necessary to know the length of time for which the exposure has lasted and to make the simplifying assumption that the dose rate during the exposure remained more or less constant. It is only worth attempting this procedure if the total dose involved is sufficiently large and the duration of the exposure is a matter of hours, up to a few days. Obviously, for small exposures (1.0 Gy) is involved, the yield becomes, in effect, Y = αD. For brief, intermittent exposures, where interfraction intervals of more than six hours are involved, the exposures may be considered as a number of isolated acute irradiations for each of which the induced aberration yields are additive. For shorter interfraction times, G(x) in Eq. (18) can be replaced by exp (–t 1 /t 0 ), where t 1 is the time between fractions. Experimental evidence which supports the G function hypothesis has been presented by Lloyd et al. and Bauchinger et al. . 9.7.4.6. Internal incorporation of radionuclides This constitutes a particular type of protracted irradiation with the added complication that exposure of the body is usually very uneven. This is because the sites of deposition of a radionuclide and its retention time depend on a large number of factors. These include the route of entry into the body, the physico-chemical form, the quality of the radiation
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
emitted, the metabolic pathways into which the nuclide may be incorporated and the subject’s physiological status. Chromosome aberrations in excess of background levels may be seen in lymphocytes taken from people who are internally contaminated. However, because of the many confounding factors, it is not possible to use the yield of aberrations to derive a meaningful estimate of radiation dose to the whole body or to specific organs. The aberration yield may be referred to a dose–response curve in which lymphocytes have been irradiated in vitro with the particular radionuclide, and this may enable an estimate to be made of the in vivo dose to the patient’s circulating lymphocytes. An example of this has been presented by DuFrain et al. for an accident in which a man received a massive contamination with 241 Am. The dose to lymphocytes, however, particularly in the case of alpha emitters, may grossly misrepresent the dose to other cells and tissues of the body. Thus, in general, cytogenetic studies are of limited value in cases of internally incorporated radionuclides. Exceptions exist when radionuclides disperse fairly uniformly around the body. Isotopes of caesium and tritiated water are two such examples. Caesium tends to concentrate in muscle which is rather ubiquitously distributed and has a biphasic clearance with 10% elimination with a half-time of 2 days and 90% with 100 days. 137 Cs was the nuclide released into the community in the Goiânia accident [149, 150] and was one of the major contributors to dose from environmental contamination at Chernobyl . Tritium taken in as tritiated water or gas is incorporated into the water of the body and so produces a more or less uniform irradiation. Its biological half-life is about 10 days so that, as with caesium, the exposure could be considered as chronic and,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
in practice, a linear dose–response would be expected. In the absence of a specific in vitro dose–effect curve for tritium, an X ray curve around 200–300 kVp will suffice. Prosser et al. have demonstrated an RBE of 1.13 at low doses or dose rates for tritium with respect to 250 kVp X rays. 9.7.5. Examples of dose estimations 9.7.5.1. Acute whole body exposure Brewen et al. and Preston et al. described an accident involving a 60 Co source in which a high dose was received fairly homogeneously over the front of the body. The mean dose to the back was lower, but it too was exposed as the man turned and walked away from the source. The total exposure time was less than one minute. A number of blood samples were taken at intervals ranging from six hours to three years after the event. The aberration yield remained fairly constant over the period of 6 hours to 32 days, during which time 7 blood samples were taken and 300 metaphases analysed from each. When the data for the 7 samples 75 were combined, 478 dicentrics and rings were observed in 2100 cells. These workers used an in vitro , gamma ray dose–response curve, where the dose D was expressed in roentgen (R): Y = 3.93 x 10 –4 D + 8.16 x 10 –6 D2 (21) to estimate a mean whole body exposure of 144 R (1 R = 0.0095 Gy). This agreed well with the physical estimate of 127 R made from a thermoluminescence dosimeter that the man had worn and a reconstruction of the event using a phantom. The general haematological changes noted were also consistent with an exposure of about 150 R. 9.7.5.2. Criticality Accident Consider a criticality accident in which 100 cells are
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
scored and 120 dicentrics observed, i.e. 1.2 dicentrics per cell. The neutron to gamma ratio supplied from physical measurements is 2:3 in absorbed dose. Cytogenetic dose estimates are to be made using calibration curves for 0.7 MeV fission spectrum neutrons and 60 Co gamma rays. The yield equations for these curves are: Neutrons: Y = 0.0005 + 8.32 x 10 –1 D (22) Gamma rays: Y = 0.0005 + 1.64 x 10 –2 D + 4.92 x 10 – 2 D2 (23) Following the steps listed in Section 9.7.4.1: (1) 1.20 dicentrics per cell is equivalent to 1.44 Gy neutrons; (2) 1.44 x 3/2 = 2.16 Gy gamma rays; (3) 2.16 Gy gamma rays are equivalent to 0.266 dicentrics per cell; (4) 1.20 − 0.266 = 0.934, which is the dicentric yield attributable to neutrons; (5) 0.934 dicentrics per cell is equivalent to 1.12 Gy neutrons. Repeating step 2, 1.12 x 3/2 = 1.683 Gy gamma rays, etc. After a few iterations, doses of 1.21 Gy neutrons and 1.82 Gy gamma rays are obtained. The complete sequence is laid out in Table 10. An in vitro validation of this approach has been described where very good estimates of actual neutron and gamma doses were obtained in international exercises to compare criticality accident dosimetry [137, 154] TABLE 10. SEQUENCE OF STEPS USED IN MAKING DOSE ESTIMATES FOR MIXED GAMMA AND NEUTRON IRRADIATION Steps 1 and 5 Step 2 Step 3 Step 4 Neutron dose Gamma ray dose Gamma ray yield Neutron yield (Gy) (Gy) (dicentrics per cell) (dicentrics per cell) 1.44 2.16 0.266 0.934 1.12 1.68 0.167 1.032 76 1.24 1.86 0.201 0.999 1.20 1.80 0.189 1.011 1.21 1.82 0.194 1.006 9.7.5.3. Low dose overexposure A non-destructive testing radiographer, working with 192 Ir sources, returned a monthly thermoluminescence dosimeter which recorded
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
a penetrating radiation exposure of 250 mSv. No colleagues who regularly worked alongside him recorded exposures on their dosimeters. There was no evidence of any systems failure or any other explanation for the overdosed badge. The case was referred for cytogenetic analysis where 1000 metaphases were scored and all were undamaged. This was reported as the best estimate of dose being zero but, using the curve Y = 0.001 + 0.0164D + 0.0492D 2 , zero carried an upper 95% confidence limit of 0.12 Gy. Investigators were doubtful if the man had indeed been irradiated and so in this case it proved useful to present the results in a different way. Using the odds ratio approach, described in Section 9.7.4.2 the odds in favour of zero come out at approximately 1300:1. 9.7.5 4. Acute non-uniform exposure An inhomogeneous irradiation, resulting in highly localized exposure sufficient to cause skin burns, occurred when a non-radiation worker picked up a 250 GBq (6.7 Ci) 192 Ir source and placed it in his pocket . Blood was sampled promptly and one thousand lymphocyte metaphases were examined; 99 of them contained the following unstable aberrations: 86 dicentrics, 2 centric rings and 60 excess acentrics. The distribution of dicentrics was: TABLE 11. DISTRIBUTION OF DICENTRICS AFTER ACUTE NON-UNIFORM EXPOSURE Dicentric per cell 0 1 2 3 4 5No. of cells 932 56 9 1 1 1The investigating laboratory’s in vitro dose–response curves were: 222 10 00 .510 57 .1 DDYdicentrics > −− ×+×= (24) 222 10 90 .310 30 .2 DDYacentrics > −− ×+×= (25) Using the contaminated Poisson method, (Section 9.7.4.3, Method 1) the maximum likelihood estimate for the yield of dicentrics, Y F, in the irradiated cells is given by substituting data from the example into Eq. (12). By iteration, Y F = 0.489
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
dicentrics per irradiated cell, which corresponds on the dose–response curve to 2.97 Gy. The size of the irradiated fraction, f, is given by solving Eq. (13), which, in this example, gives f = 0.176. As this value represents the population of cells which was irradiated and survived, it needs to be adjusted, as described in Eq. (14), in order to take account of selection against the irradiated cells by factors such as interphase death and mitotic delay. There is some experimental evidence indicating that this selection is an exponential 77 function of dose, with D 0 = 2.70 Gy. In the present example, the dose estimate of about 3.0 Gy would imply that only about 0.33 of the irradiated cells (p in Eq. (14) survived to be analysed. The fraction originally exposed, F, is equal to 0.393 and is obtained by solving Eq. (14). In round terms, therefore, the irradiated fraction of the body is about 40%, with an average dose of about 3.0 Gy. In the Qdr method, (Section 9.7.4.3) it should be noted that the investigating laboratory did not normally use the yield of dicentrics plus rings for dose estimation, but rather dicentrics alone, ie. Qd. As rings are rarely observed aberrations, as compared with dicentrics and excess acentrics, this modification has only a trivial effect. Therefore, substituting values into Eq. (16) and omitting centric rings gives 222 10 90 .810 87 .3222110 00 .510 57 .199 86 > DD eDDQd −− ×−×−−−−××+×== (26) The equation can be solved for D by iteration and gives a dose estimate of 3.19 Gy. This is in good agreement with the value of 2.97 Gy derived from the contaminated Poisson method. 9.7.5.5. Delayed blood sampling Below are presented two examples of dose calculation in cases of delayed blood sampling. Adjusting
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the dicentric yield Stephan et al. have reported an accident in which two men were fairly uniformly exposed for about five minutes to a 60 Co gamma ray source. They wore film badges which indicated 470 and 170 mSv and these values agreed very well with physical calculations of the doses. Unfortunately, blood sampling was delayed by 215 days for the more highly exposed man and by 103 days for his colleague. About 1500 metaphases were examined from each man and almost identical yields of 0.47 and 0.46 dicentrics per 100 cells were obtained. These correspond to 0.13 Gy on the dose–response curve: 264 10 00 .510 00 .3 DDY −− ×+×= (27) The authors chose to adjust the dicentric yields by × 3 and × 2, respectively, to account for the delays. This decision was based on the data of Brewen et al. and Preston et al. from the accidental whole body irradiation described in Section 9.7.5.1. The adjusted dicentric yields produced dose estimates of 0.31 and 0.22 Gy. Although this brings the biological estimate of dose for the more highly exposed man closer to the physical estimate, there is still some discrepancy. Had the authors chosen to use the delay data from the study of ankylosing spondylitics , a correction of at most 1.4 might have applied, so that the discrepancy between biological and physical estimates for the first man would have been greater. In view of the spondylitics’ effect lasting up to 20 weeks, the delay of 103 days would require no correction. Using the Qdr method Ishihara et al. have described a serious accident in which an 192 Ir industrial radiography source was taken into a dwelling, irradiating six people. The two most seriously overexposed subjects received partial body irradiation, which was
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
evident from skin burns. This was further reflected in the aberration data, where doses estimated by the Qdr method were 1.95 and 1.50 Gy, substantially higher than the values of 1.52 and 0.54 Gy whole-body 78 dose, respectively, from the yields of dicentrics and rings on a per cell basis. The values of dicentrics and rings per cell varied somewhat in the first two months after exposure, but then became more stable up to six months. By contrast, the Qdr values stayed relatively constant from the beginning up to 400 and 200 days, respectively, when further study ceased. 9.7.5.6. Protracted and fractionated exposure In December 1998 a serious accident occurred in Istanbul, where a former radiotherapy 60 Co source was broken open in a scrap metal yard . Ten persons were irradiated, mostly during one day, with exposure times ranging from 2 to 7 hours . One of the subjects is used here as an example. His exposure was for 7 hours, and the dicentric frequency, from the pooled results of four laboratories, was 157 dicentrics in 688 cells = 0.228 ± 0.18 dic/cell. From the acute dose–response curve 2060 .0003 .0001 .0 DDY ++= (28) the acute dose estimate (±SE) = 1.7 ± 0.1 Gy. The uncertainty here is slightly simplified by ignoring any errors on the dose–response curve. Applying the G function, where 5.3270 === ttx (29) so that G(x) = 0.413, the dose–response curve now becomes 2025 .0003 .0001 .0 DDY ++= (30) The dicentric frequency now corresponds to a 7 hours exposure of 2.5 ± 0.1 Gy. 9.7.5.7. Internal incorporation of radionuclides An accidental inhalation of about 35 GBq (~1 Ci) of tritiated water droplets by a factory worker is described by Lloyd et al. . Removal of tritium from her body was speeded up by
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
forced diuresis. A committed dose to soft tissue was obtained from measuring the concentration and rate of excretion of tritium in her urine. Dicentric yields were measured in blood samples taken at various times after the event, and data from 40–50 days were used for biological dosimetry as by then all committed dose had been received. Dicentric yields were referred to a linear in vitro dose response calibration coefficient, 5.37 x 10 –2 D, producing an estimate of average dose to lymphocytes of 0.58 Gy. This value needed further multiplication by a factor of 0.66. The derivation of this factor takes account of the differing water contents of the whole body, soft tissue and lymphocytes. Aberration yield is calibrated against dose to lymphocytes whereas tritiated water delivers dose principally to the soft tissues of the body. This correction produced a biological dose estimate of 0.38 Gy with 95% confidence limits of 0.48 and 0.28 Gy, and is a more realistic comparison with 0.47 Gy ± 20% obtained from the urine measurements. The conversion of tritium concentration in urine to dose to soft tissue also allowed for the water content of soft tissue . 79 10. TRANSLOCATION ANALYSIS A recognized drawback of the dicentric and cytokinesis-block micronucleus (CBMN) assays is that the damage is unstable and therefore is eliminated from the peripheral blood lymphocyte pool at the rate that cell renewal occurs (see Section 4). It has long been recognized that analysis for more persistent types of damage, e.g. stable translocations, is needed to address biological dosimetry for old or long term exposures. Translocations are detectable by karyotyping, which is, however, too laborious to be applied routinely in biological dosimetry. The introduction of FISH has opened the possibility to detect translocations by an alternative means. The technique employs specific sequences
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
of DNA which can be used as probes to particular part of the genome and then by attachment of various fluorochromes to highlight or ‘paint’ the regions in different colours. Translocations are seen as coloured rearrangements in a fluorescence microscope (shown in Figs 26 and 27). FIG. 26. Human metaphase with coloured painted chromosomes #2 (FITC, green), #4 (Texas Red) and #8 (FITC+Texas Red, yellow), and the rest counterstained with DAPI. An apparently simple translocation, or two-way translocation [t(Ba),t(Ab)] involving chromosome # 2 is observed. > 81 FIG. 27. Human metaphase with monocoloured painted chromosomes #1, #4 and #11 labelled with Cy3 (red), centromeres highlighted with a pancetromeric probe labelled with FITC (green), and the rest counterstained with DAPI. An apparently simple translocation or two-way translocation [t(Ba),t(Ab)] involving chromosome # 1 is observed. FISH has many applications in medicine and in fundamental cytogenetics. In this publication, however, only its application to biological dosimetry will be addressed. A large variety of probes are now available so that one may selectively paint whole or limited regions of each of the human chromosomes. By attaching fluorochromes in varying ratios to specific sites it is possible to highlight different regions concurrently with a wide range of colours. One drawback in using many colours such as the multicolour FISH (mFISH) procedure is that the shade differences may be too subtle for discrimination by the human eye. Therefore electronic systems are required to capture images and display them with applied false colours (Figs 28 and 29). > 82 FIG. 28. View of a metaphase stained by mFISH. A: in RGB colours as taken by the camera. FIG. 29. View of a metaphase stained by mFISH. B: the same metaphase but where a pseudo colour has been associated to each pair of chromosome. 10.1. CELL CULTURE AND
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
FIXING PROCEDURES The procedures for obtaining blood, culturing the lymphocytes and harvesting fixed cells are similar to those described for the dicentric assay (Sections 9.1 and 9.2). Although translocations are stable through mitosis, it is still good practice to carry out the analysis on M1 metaphases. This is particularly important because the mitotic loss of cells containing unstable aberrations could distort the mean frequency of translocations. Moreover there may be occasions when both stable translocation and unstable dicentric frequencies are required > 83 from the same specimen. For FISH analysis it is better to store refrigerated fixed cell suspensions. Cells dispensed onto microscope slides can be stored. They should be kept at -20°C but, even so, deterioration has sometimes been noted. Fixed cell suspensions are more convenient for transporting to other laboratories. Generally, for biological dosimetry, only a part of the genome (e.g. three pairs of chromosomes) is painted. This leads to the requirement to score more metaphases than would be scored with the dicentric assay. It is therefore helpful and more cost effective to produce slides, each with a large number of scorable quality metaphases. 10.2. PAINTING THE CHROMOSOMES With the range of probes and fluorochrome combinations now commercially available it is possible to highlight all chromosomes by the method known as multicolour FISH (mFISH) . This permits full karyotyping and thus scoring all inter-chromosomal translocations. The centromeres and telomeres of all chromosomes can be separately highlighted too. Intrachromosomal exchanges such as pericentric inversions may be detected by selectively painting the p and q arms of a chromosome in different colours , and rearrangements within a single arm may be detected by mBAND where multi-coloured banding is produced along a chromosome [164, 165]. An increased frequency of intrachanges with respect to interchanges has been proposed as a ‘fingerprint’ of
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the effect of high LET radiations and therefore these methods have particular applications for researching radiation quality effects. For most retrospective biological dosimetry applications it is sufficient to detect just inter-chromosomal translocations and ideally mFISH can provide the maximum information from each metaphase. It can also be extended so that individual chromosome arms can be highlighted in different colours (pq-mFISH) but is an expensive, time consuming procedure and highly demanding of sophisticated image capture and manipulation systems. Therefore the practice has evolved for painting a limited number of chromosome pairs either with the same or in separate colours and counterstaining the remaining chromosomes. Application of a pan-centromeric probe simultaneously with whole chromosome paints is recommended to distinguish between dicentrics and translocations more accurately (Fig. 30). > 84 FIG. 30. Human metaphase with monocoloured painted chromosomes #1, #4 and #11 labelled with Cy3 (red), centromeres highlighted with a pancetromeric probe labelled with FITC (green), and the rest counterstained with DAPI. An apparently simple dicentric [dic(BA),ace(ab)] involving chromosome # 1 is observed. A dicentric plus an acentric involving the counterstained chromosomes is also present. Generally, painting three of the larger chromosomes (i.e. #1 to #12 — see Fig. 7), representing about 20% of the genome (see Tables 2 and 3), leads to about 33% efficiency in detecting translocations when a single colour is used. The percentage of the genome that each cocktail ‘paints’ relative to the total genome is estimated from the physical lengths of chromosomes . The total genomic translocation frequencies may be estimated according to a standard formula proposed by Lucas et al. , which applies with the assumption of simple pair-wise exchanges. It is advisable not to include chromosomes 7 or 14 in probe combinations as translocations and other aberrations involving these chromosomes can arise in
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
vivo during immunological development and may thus confound the quantification of a radiation effect [171, 172]. For retrospective biological dosimetry a single colour FISH for a triple cocktail of target chromosomes appears to be sufficient. Multiple colour painting of the triplet increases the detection efficiency (if chromosomes #1, #4 and #12 are highlighted from about 31% to about 34%) and gives a better detection of complex translocations that can be encountered following high dose recent exposures. The equations, given below in Section 10.5 for converting to full genome equivalence can be applied to both single and multiple colour painting. > 85 10.3. SCORING CRITERIA > 10.3.1. Selection of scorable cells Although there is no firm consensus on which metaphases should be scored, well-spread metaphase cells are considered suitable for scoring if the cells appear to be intact, the centromeres are morphologically detectable and present in all the painted chromosomes, and the fluorochrome labelling is sufficiently bright to detect exchanges between chromosomes labelled in different colours . Routinely, completeness of the counterstained chromosomes is not considered but most researchers would reject a metaphase if it is obviously missing several chromosomes, e.g. <40 objects. Some researchers consider, that all the painted material present should be scored although this involves a certain degree of judgement because the limits of resolution with current FISH technology are about 11–15 Mbp . In consequence, some symmetrical translocations look like apparently incomplete exchanges, but investigations with telomere probes have shown, that a large portion of apparently incomplete translocations are complete ones . Those cells that are obviously deficient in a large portion of painted material or labelled centromeres should be excluded from the scoring. For retrospective dosimetry, it was shown, that the frequencies of translocations in the stable cells, defined as cells without dicentrics, centric rings or
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
acentrics, are constant with time [64, 174]. Therefore it is recommended to record whether each translocation occurs in a stable or unstable cell. > 10.3.2. Nomenclature and recording data To describe the chromosome aberrations detected by painting two specific nomenclature systems were developed independently, and descriptions based on the conventional terminology of routine cytogenetic scoring were also used [173, 175–178]. The nomenclature systems were introduced because, with partial genome analysis, the conventional terminology proved inadequate as many patterns revealed by FISH appeared to be more complicated than expected. (i) A system with the acronym PAINT was developed to be purely descriptive of each aberrant painted object in the metaphase . Each is therefore described individually without cross-reference to other aberrant objects in the cell. Each colour is designated by a letter, starting alphabetically with the counterstain. A capital letter designates the component that bears a centromere. Thus, with single colour painting, t(Ab) is a bicoloured object consisting of a centromeric piece of a counterstained chromosome and a non-centromeric piece of a painted chromosome. Conversely, t(Ba) is an object where the centromere is on the painted component. Multiple coloured painting is accommodated by including further letters in the nomenclature. The reader is referred to Tucker et al. for full descriptions of all the abbreviations used in the system. An additional suggestion made in that paper of counting colour junctions as an index of damage relatable to dose has no practical application to retrospective dosimetry. (ii) Savage and Simpson (S&S) [176, 177] proposed a terminology comprising numerals and letters describing each exchange in its entirety. The numerals refer to the number of objects containing painted material, and the alphabetical ordering of letters reflects how common the patterns are expected to be. This so-called S&S system applies only to single paint patterns.
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
However, it can be used with dual and triple paint patterns but each painted chromosome has to be scored in isolation irrespective of the colours of > 86 partners . This nomenclature has considerable uses in mechanistic studies, particularly, for example, in understanding complex rearrangements. A more conventional terminology may be employed that names translocations as reciprocal, terminal or interstitial . The first two have also been called complete or two-way and incomplete or one-way translocations, respectively. The third includes inversions and insertions. Complete/incomplete or reciprocal/terminal involve mechanistic concepts. For biological dosimetry purposes they are probably best referred to as two- or one-way, purely on the basis of their visual appearance and with no mechanistic implications. Indeed mechanistic studies have shown that one-way patterns do not provide a reliable estimate of exchange incompleteness . An insertion is one of many types of complex rearrangement which are formally defined as arising from three or more breaks on two or more chromosomes . The nomenclatures described above are not mutually exclusive but rather complementary and comparisons between nomenclatures have been applied to a common data set . Nowadays the most widely used method for recording data is to describe each abnormal metaphase as a unit using the PAINT nomenclature but in a slightly modified way that considers the underlying mechanisms of the formation of aberrations. The abbreviations of the PAINT system are used but a note is made of the associations between objects in the metaphase, thereby incorporating aspects of the conventional terminology too. Chromosome aberrations are classified as simples or complexes, the latter ones considered when three or more breaks in two or more chromosomes are needed to produce the observed abnormality. Aberrations are considered complete when all broken pieces are rejoined and as incomplete when one or more
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
pieces appeared unrejoined. For example, t(Ba) seen with t(Ab) is regarded as a simple complete or two-way translocation, and either pattern alone is regarded as a simple incomplete or one-way translocation when seen alone in a metaphase, sometimes with an associated painted acentric, t(Ba) plus ace(b). Complexes are recorded as such and described either as insertions, e.g. ins(Aba), or as the more complicated rearrangements like a t(Ba) with ace(ab) or dic(BA) with a t(Ab). Note that painting a concrete set of chromosomes exchange aberrations like a t(Ba) plus a t(Ab) are considered as ‘apparently’ simple aberrations because they can arise from undetectable complex aberrations, that are only detectable with mFISH [182, 182]. 10.4. DATA HANDLING Lucas et al. derived the equations for calculating genome equivalence, and these have been further summarized by Lucas and Deng . The genomic translocation frequency is usually calculated by using the formula for the painted fractions of the genome as follows: )1(05 .2 ppPG ffFF −= (31) where: FG is the full genome aberration frequency, Fp is the translocation frequency detected by FISH, and f p is the fraction of genome hybridized , taking into account the gender of the subjects. 87 This is more fully elaborated below in Section 10.5.1. The conversion of data to full genome equivalence is a recommended procedure to use when data are to be combined or interlaboratory comparisons are to be made between results from various studies where different combinations of whole chromosome painting probes have been used. The assumption, sometimes referred to as the Lucas formula, is that the probability of the involvement of a particular painted chromosome in an aberration is proportional to its DNA content. This issue has been intensively investigated [184–186] and in essence it is accepted that this assumption gives a reasonable
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
approximation. However, there is a consensus that using the (DNA content) in the Lucas formula, larger chromosomes may tend to be overestimated in their participation in simple exchange aberrations compared to the smaller ones [187, 188]. Therefore, the use in the Lucas formula of the (DNA content) 2/3 rather than the (DNA content) gives more accurate results. Some authors have argued that this kind of proportionality could be symptomatic of interchanges involving primarily chromatin near the boundary of chromosome territories [186–188]. The best data on relative DNA contents of the human chromosomes are given by Morton and the values shown in Tables 2 and 3 have been calculated from data given in his Table 4, column 2. 10.4.1. Single colour painting A fraction, f, of the genome is painted (green) and the remainder, 1 – f, is counterstained (blue). Note: fp in Eq. (31) above has, for simplicity, here been shortened to f. There will be f2 green–green exchanges (1 – f) 2 blue–blue exchanges 2f(1 – f) blue–green exchanges Total 1.00 However, this total includes exchanges within the same chromosome, e.g. inversions. The total number of interchromosomal exchanges is 0.974, using the same assumption of DNA proportionality (see calculations in Lucas et al. ). Hence, the fraction of all translocations that are blue–green translocations is given by Eq. (32): )1(05 .2974 .0)1(2 ffffFFGP −=−= (32) where: FP and FG are, respectively, the translocation frequency measured by FISH and the whole genome translocation frequency. The same formula would apply to blue–green dicentrics. Example Suppose that chromosome pairs 1, 2 and 4 are painted. Their respective DNA contents (male) from Table 2 are 0.0828, 0.0804 and 0.0639. Therefore, f = 0.2271, so that F P/F G = 0.360. 88 This combination of chromosomes painted is 36% efficient in measuring bicoloured
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
translocations. Therefore, to obtain the full genome translocation yield the observed yield is divided by 0.36. 10.4.2. Two colour painting Suppose a fraction, f 1 , is painted red, another fraction, f 2 , is painted green and (1 – f 1 – f 2 )= f 3 is counterstained blue. There will be: f12 red–red exchanges f22 green–green exchanges f32 blue–blue exchanges 2f 1 f2 red–green exchanges 2f 1 f3 red–blue exchanges 2f 2 f3 green–blue exchanges Again, the total interchromosomal exchanges are 0.974, and hence the fraction of all bicoloured translocations is given by # [ ])1()1()1(05 .2974 .0)(2 332211323121 ffffffffffff −+−+−−++ > . (33) Example Suppose that chromosome pairs 1, 2 and 4 are painted red; pairs 3, 5 and 6 are painted green and the rest is counterstained blue. The fractions from Table 2 are f 1 = 0.227 and f 2 =0.186: 582 .0)284 .0(05 .2)042 .0151 .0175 .0(05 .2 ==−+= > GP FF (34) This combination is 58% efficient in detecting translocations. It should be noted that where a two-way exchange between two differently coloured painted chromosomes occurs it is still only counted as a single event. 10.4.3. More than two colours The calculations can be extended to multicolour FISH painting. For many colours the equation becomes ⎥⎦⎤⎢⎣⎡ −−= ∑∑ < jijiiiiGP ffffFF )1(05 .2 (35) All the calculations detailed in this section are available as part of the Dose Estimate software, mentioned in Section 8.3. 89 10.5 THE CONTROL LEVEL OF TRANSLOCATIONS Control levels of translocations are higher than for dicentrics, and to some extent this is due to the former being a persisting type of aberration. It is therefore important to take the translocation background into account, particularly after low doses, when attempting retrospective biological dosimetry. Of course, a pre-exposure control blood
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
sample from the accidentally irradiated subject or from a population study group is unavailable, and therefore an assumed value based on generic survey data has to be used. Ideally a laboratory should develop its own control database but this is an extensive undertaking given that it would have to cover a number of confounders and especially a wide span of age groups. A comprehensive meta-analysis published by Sigurdson et al. currently provides the best international database, broken down by age, gender, race and smoking habits. It incorporates data from an earlier study that combined results from some European laboratories . From both studies, it appears clearly that age is the major factor that determines the background frequency of translocations which rises substantially above the age of 60 years (Fig. 31). 00,5 11,5 22,5 01-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 60-64 65-69 70-74 75-79 80+ > Age (years) Yield of translocations for 1000 cells > Age (years) FIG. 31. Number of translocations as a function of age . It is important to take the background into account and to subtract from the number of translocations observed in an individual's lymphocytes the translocation rate expected given his or her age. In addition to a confirmed laboratory effect, the study by Sigurdson et al. showed a significant variation between the four principal geographic areas. On the other hand, no effect was observed for ethnicity or gender. It thus appears essential to compare only data from the same laboratory for studying factors that influence the translocation rate. Among the few studies that report the effect of gender on the translocation rate [107, 189, 190], only that by Whitehouse et al. shows a higher translocation rate in men than women for the 20–29-year-age group, significantly different
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
from the 30–39-year-age group (p 31 ppm) Kim et al. Smith et al. (< 31 ppm) Zhang et al. Industrial pollution Beskid et al. Beskid et al. Sram et al. Heavy metals Maeng et al. (Chromium, smokers only) Dulout et al. (Arsenic) Doherty et al. (Chromium) Maeng et al. (Chromium, Non-smokers only) Only alcohol tended to create an excess of translocations in both of the studies that considered it. This trend was not observed for smoking (8/11 studies) or pesticides (1/2 studies). The case of benzene is particular because the analyses were performed for the chromosomes involved in diseases specifically related to exposure. This means that only exchanges between two chromosomes (# 8 and 21 or # 14 and 18) were recorded in
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
some studies [202–204]. All three reported a significant effect when they studied only the t (8; 21) and t (14;18) translocation rates. However, 3 studies were identified in which the translocations between those chromosomes and all the others were examined, and none showed significant results. Substances that are used and abused by choice, such as tobacco, drugs, and alcohol, affect translocations only when their consumption is heavy and chronic. Nonetheless, they often 91 induce a significant increase in translocation rates when they are combined or associated with other types of agents (metals and mutagenic chemicals) . The synergistic effects of abusable substances reported in some studies suggest that smoking increases the translocation rate generated by an occupational exposure. This is the case for people exposed to ionizing radiation , to pesticides , and to chromium . The synergistic effect of smoking is all the more evident in that most studies of the effects of smoking alone do not show any significant increase in translocation rates. On the other hand, for alcoholism, the synergistic effect is more difficult to show because alcoholics rarely have only a single addiction, or at least because the number who do not have other addictions is too low for a comparative analysis with a control group. A study of the impact of alcoholism alone on the translocation rate would thus be instructive. The effect of toxic agents used in the workplace (pesticides, benzene, and metals) on the number of translocations is often proportional to dose and duration of exposure. It has also been shown that exposure to a mixture of products is more harmful than exposure to a single chemical element. Finally, the efficacy of individual protection (gloves, masks, jumpsuits/coveralls) was shown by the diminution in the translocation rate in the population exposed to these types
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
of mutagenic substances. Fig. 32 illustrates the impact of the factors studied on the rate of translocations as a function of the type of factor, of the exposure, and of the studies. > 0510 15 20 25 30 35 40 > Burim, 2004 Badr, 1982 Ramsey, 1995 Pressl, 1999 Bothwell, 2000 Tucker, 2003 Whitehouse, 2005 Beskid, 2007 Van Diemen, 1995 Tawn, 1989 Sigurdson, 2008 Steenland, 1997 non ajusted for age (appliers) Steenland, 1997 non ajusted for age (landowners) Steenland, 1997 ajusted for age (appliers) Steenland, 1997 ajusted for age (landowners) Tucker 2003 (smokers) Tucker 2003 (non smokers) Beskid, 2007 (Prague) Beskid, 2007 (Kosice) Beskid, 2007 (Sofia) Beskid, 2006 (benzene éthyl) Maeng, 2004 (Chrome smokers) Maeng, 2004 (Chrome, non smokers) Doherty, 2001 (alliage chrome-cobalt) LDB 0,2 Gy LDB 0,5 Gy LDB 0,7 Gy > Translocation rate per 1000 cells Translocation rate for control population Translocation rate for exposed population * * *** * ** ******** LDB : Translocation rate obtained in the Laboratoire de Dosimetrie Biologique for given radiation dose (Cobalt exposure, dose rate of 0.5 Gy/min). * Studies where the difference between the exposed and the control group is significant Alcohol Tobacco Pesticides Benzene and Pollution Ionising Radiation exposure Heavy Metals FIG. 32. Comparison between the translocation rates generated by different agents and the translocation rate generated by in vitro irradiation. For each study the translocation rate per 1000 cells in the control population and in the exposed population are indicated. In addition, the translocation rate obtained after in vitro irradiation of blood samples to doses of 0.2, 0.5, and 0.7 Gy are also presented in this figure, 92 to compare the induction rates. However, it is necessary to note that the irradiations were performed in a short time period and were acute exposures whereas, exposures to alcohol, smoking, pesticides,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
or benzene are chronic. Exposures are designated as chronic when the individual is exposed to the genotoxic agent for all or a significant part of their life span (lifestyle or occupational exposure). Fig. 32 also shows that the translocation rates encountered in populations exposed to alcohol, smoking, and pesticides are considerably lower than the rates measured for benzene and especially chromium exposure. Moreover, the values for the control groups are relatively similar to those observed in the groups exposed to alcohol, smoking, and pesticides. In comparison with a group exposed to ionizing radiation at a dose of 0.5 Gy, the translocation rate for people exposed to alcohol, smoking, and pesticides is low. One can therefore conclude that if a high translocation rate is observed it can be attributed to the irradiation. On the other hand, during a retrospective study of exposure to ionizing radiation among people highly exposed to chromium or benzene, it will not be possible to differentiate the radiological from the environmental translocation rates. Nevertheless, exposure to benzene or chromium is not common and it should be possible to identify it by an appropriate questionnaire covering past and present occupational activities. 10.6. PERSISTENCE OF TRANSLOCATIONS The greatest disadvantage of the dicentric method is that the aberration yield in exposed people decreases with time after exposure. Dicentrics induced in the peripheral circulating lymphocyte pool will be removed by cell death and diluted by renewal of blood cells. They are mitotically unstable aberrations unable to pass through repeated cell divisions. By contrast, translocations are generally mitotically stable and, provided that the cell as a whole is stable (C S), translocations induced in stem cells can pass through to mature circulating lymphocytes. Initially translocations are induced at a frequency similar to that of dicentrics but it is their post-irradiation stability that
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
makes them more suited for retrospective biological dosimetry. The persistence of the translocation frequency has long been a matter of discussion. Several years after the Goiânia accident in Brazil, the frequency of translocations was found to be lower than the frequency of dicentrics observed just after the exposure [209, 210]. However, the initial translocation yield was not available in either of these studies. In a retrospective study of Mayak (Southern Urals) nuclear-industrial personnel, the estimated doses were mainly lower than would be predicted by the calibration curve for transmissible apparently simple translocations . Other follow-up studies of accidentally irradiated persons showed that the frequency of translocations remained relatively constant with time. No substantial change in the translocation frequencies was observed in Chernobyl irradiated subjects from 5 to 8 years after the accident . In whole-body irradiated individuals from an accident in Estonia, the frequencies of translocations remained fairly constant for two years, except for one individual exposed to protracted whole-body but also to non-uniform irradiation . However, four years after the same accident, the frequency of all translocations in cells containing only simple rearrangements fell on average to 65% of their initial value, but two-way translocations were slightly more persistent than total translocations . The reduction in the frequency of translocations with time was attributed to the partial-body irradiation and agrees with the idea of a coincident distribution of dicentrics and translocations in such exposures [215, 216]. From this it would follow that, at long times after partial-body exposures, the estimation of the whole-body dose will tend to be lower as the dose increases . 93 A more marked decrease with time post-irradiation in the frequency of translocations as the dose increased has been described in other cases of accidental exposures of humans; the frequency of translocations persisted with time
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
for doses below 1–3 Gy but a decrease was observed after higher doses [217, 218]. In cancer patients treated with radiotherapy, a clear reduction of the yield of translocations with time post irradiation was observed, which was more pronounced in persons with a higher initial frequency of translocations. [219–222]. A possible explanation for these observations in radiotherapy patients is the lethal dose to the stem cells and the repopulation of this area by unirradiated cells from outside the irradiated part of the body. Overall, these results indicate that at high doses the initial yield of translocations may not always be constant with time post-irradiation. Completeness and complexity are factors that can influence the disappearance of translocations with time post-irradiation. The majority of complex aberrations are not transmissible, and as a consequence cells carrying such aberrations will disappear with time post irradiation [213, 215, 223–226]. Another factor contributing to the persistence of translocations is the co-occurrence of translocations and unstable aberrations in the same cell. The procedure to consider stable cells instead of total cells for retrospective dose-estimation has therefore been proposed and is under consideration. In a follow-up study on victims of the Estonia accident the initial yield of translocations, when all cells are considered, decreased to about 70% after 2 years . However, a further study in the same group, in which digitalized images of damaged cells were re-analysed to select those without unstable aberrations, indicated that after 7 years the yield of translocations was similar to that in the first 2 years. This study however was limited because it had to rely on cell images that were retained because they contained damage in the painted chromosomes. No correction was possible for those cells originally considered as ‘normal’ regarding the painted material because they had not been digitized. Nevertheless,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
this study taken with the theoretical likelihood that after a long period of time only stable cells will remain, suggests that to consider the yield of translocations in stable cells is more appropriate than the yield of translocations in all cells. This can only be finally resolved by good follow-up studies with FISH and unstable aberration analyses run in parallel on irradiated accident victims starting promptly after their exposure. 10.7. CALIBRATION CURVES For dose estimations with translocations detected by FISH, each laboratory needs to establish its own curves. The mathematics of curve fitting is just the same as for dicentrics which has been described earlier (Section 8.3). The curve should be made with the same FISH probe cocktail that is routinely used for case investigations. Doing this prevents the need to convert to genome equivalence which could introduce some extra uncertainty. For low-LET radiations, when calibration curves for translocations have been constructed taking into account stable or total cells, there were no differences in the fitted coefficients if only apparently simple translocations were considered [227, 228]. However, during microscope analysis, it is recommended to score all aberrations detected in the entire chromosome set, not only those affecting the painted material . This will give the opportunity to establish for certain if the restriction to stable cells only gives more realistic dose-estimations. FISH dose estimations generally will be undertaken for cases where doses were high, but protracted or after low radiation exposure a long time ago, revealing no medical symptoms. In contrast to acute exposure dosimetry, where the linear quadratic curve will be used, here the linear α term of the dose response curve is crucially important. The F-test indicated in Section 8.3 can be used to ascertain the reliability of the linear coefficient. Few, if any, > 94 published calibration
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
curves have enough scoring in the low dose range to have obtained a sufficiently reliable linear coefficient with a small confidence range . To construct a linear dose relationship, several dose points below 1 Gy each with a large number of cells need to be scored and this comprises a considerable workload. In the interim, one may make a number of reasonable assumptions to resolve this problem. The ratio of formation of dicentrics and translocations is about 1:1 [229, 230] and therefore similar dose response coefficients can be expected. Evidence from various published data from dicentrics suggest a linear calibration coefficient of about 15–20 translocations per 1000 genome equivalent cells per Gy for high energy gamma rays . It is further known, that the linear term of the dose effect curves is mainly influenced by the linear energy transfer of the radiation quality, whereas the curvature due to the β coefficient is dose rate dependent (as described for the G-term correction in Section 9.7.4.5). There are indications that the linear coefficient obtained using in vitro chronic exposure, provided it is made at body temperature, is not different from that reported for the linear term of the acute dose response . Thus until reliable linear coefficients for FISH translocations have been obtained by in vitro low dose-response calibration experiments, it is reasonable to use the linear term obtained with the same radiation quality for acute dicentric calibration curves. With dose reconstruction a long time after exposure, the dose estimation will be based on the assumption that lymphocytes irradiated in vitro and blood stem cells irradiated in vivo will show similar frequencies of translocations. It is not sure whether the radiosensitivity of stem cells and of mature lymphocytes are identical or whether there might be an impact of the intervening cell divisions
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
where cells containing unstable aberrations will be eliminated. Retrospective dosimetry assumes that these are not major confounders and the recent literature suggests that this is of no practical importance . In conclusion, dose reconstruction on the basis of translocations in stable cells is an established method but has limitations. It seems to be a good tool after protracted and low dose exposure where the linear term of the calibration curve predominates. As an interim measure the term from dicentric dose-response curves could be assumed. After high, especially acute, dose exposure, restricting to stable cells may underestimate the dose because the number of cells with translocations in unstable cells will increase together with the number of complex aberrations. Moreover there is a limit on the upper dose to which one may calibrate as cells free from unstable damage become increasingly rare. 10.8. EXAMPLES OF FISH BEING USED FOR RETROSPECTIVE BIOLOGICAL DOSIMETRY These studies were designed to investigate the feasibility of the FISH translocations assay for retrospective dosimetry in (1) populations with no prior biological and physical dosimetry investigation; (2) populations with known physical dosimetry estimates; and (3) populations with known biological dosimetry estimates using conventional dicentric analysis immediately following exposure. The data from the last group are considered to be the most reliable ones for comparison with translocation frequencies in order to define the stability of translocations. Four study groups selected were composed of (1) nuclear power plant workers; (2) populations living in contaminated areas; (3) Chernobyl cleanup workers; and (4) individuals or groups of persons accidentally exposed. > 10.8.1. Retrospective biological dosimetry in population groups without prior personal dosimetry In order to perform retrospective estimations of radiation doses, the frequency of chromosomal aberrations was determined in 15 individuals known to be severely exposed as a > 95 result of
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Chernobyl accident, and all were treated for symptoms of the delayed stage of the cutaneous radiation syndrome. These studies began in 1991 and follow-ups were performed until 1994 [212, 233, 234]. In 1991, biological dose equivalent estimates were determined, either by measuring the frequency of dicentric and ring chromosomes using the Qdr method or by measuring the frequency of stable translocations using FISH with composite whole chromosome specific DNA libraries and a pan-centromeric DNA probe. With both methods, fairly comparable individual dose estimates between 1.1 and 5.8 Gy were obtained for 12 out of 15 individuals, whilst three of them showed no elevated aberration frequencies . For the follow-up studies the frequencies of translocations were examined in the same donors during a period of three years from September 1991 until July 1994, when, in 11 out of 12 cases, they remained fairly constant. This permitted comparable dose estimates from the various sampling times to be made . From these studies a direct conclusion on the stability of translocations cannot be made because there are no reference data immediately following exposure (i.e. biological and physical dosimetry). However, the follow-up studies indicate that translocations can remain constant from five years post-exposure time and at different dose levels. > 10.8.2. Retrospective biological dosimetry in population or occupational exposure groups with physical dose estimates Several studies designed primarily to estimate absorbed doses have been carried out on the frequencies of chromosome aberrations in the lymphocytes, e.g. of the atom bomb victims of Hiroshima and Nagasaki (Japan) or in Chernobyl cleanup workers. The frequencies of translocations recorded in atom bomb survivors seemed close to expected values derived from the individuals’ DS86 (Dosimetry System 1986) estimated doses compared with an in vitro dose–response curve . These studies, therefore, supported the idea of long term persistence of
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
translocations. However, by contrast, a significant difference was found for four workers irradiated during the Oak Ridge (USA) Y-12 accident in 1959, where some years later the translocation frequencies were substantially below expected values . A pilot study carried out in 1994 of about 60 personnel recruited from Estonia for cleanup work in Chernobyl in 1986 or 1987 with registered doses ranging from 0 to 300 mSv was undertaken to determine whether both dicentric and translocation analyses might verify their recorded doses . In another set of investigations, 52 cleanup workers were studied with FISH painting . The dicentric estimates were no longer valid but translocations could be used to verify early dosimetry carried out on only the more highly irradiated persons. For the vast majority of lesser exposed subjects, FISH was found to be impractical as an individual dosimeter. However, it has been suggested as having some value for discriminating groups of subjects exposed to different doses , and this is supported by the study of the Estonian cleanup workers [92, 237]. There is another data set on 75 Mayak workers for whom physical dosimetry was available and who had received their main exposure between 1948 and 1963 . Cumulative external doses between 0.02 and 9.91 Sv and plutonium burdens ranging between 0.26 and 18.5 kBq were reported. At 35 to 40 years after protracted exposure using whole chromosome painting probes for chromosomes 1, 4 and 12 in combination with a pan-centromeric probe, the translocation frequencies were determined. The results showed a higher frequency of translocations in the Mayak workers in comparison with a matched control group. However, the range of translocation yields was generally lower than expected from the registered personal doses and calibration curves [235, 238]. > 96 FISH painting was carried out on metaphase preparations obtained
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
from 73 radiation exposed residents from settlements along the Techa river. The study group comprised two subgroups living in settlements either 7 to 60 km or 78 to 148 km downstream from the facility. Both were distinguished from controls, and significantly higher mean translocation frequencies were observed . Biological dosimetry studies of radiation workers at the Sellafield nuclear site with accumulated lifetime whole body doses ranging from 173 to 1108 mSv, all but three being >500 mSv, were carried out in the period 1991 to 1994. When the workers were divided into dose range cohorts the groups’ mean translocation frequencies showed a significant increase with dose categories. However, by contrast, the cumulative lifetime doses were unrelated to dicentric frequencies . In Hiroshima atom bomb survivors a good correlation was found between electron spin resonance dosimetry and cytogenetic dosimetry using translocation frequencies from lymphocytes of 40 survivors who lived close (approximately 2 km) to the hypocentre, and who were at least 10 years old at the time of bombing . The Hiroshima atom bomb survivor studies indicate the persistence of stable translocations. However, from the other studies mentioned above, it can be concluded that some fraction of translocations seems to decrease with post-exposure time. 10.8.3. Retrospective biological dosimetry in persons with known biological dose estimates made shortly after accidents, using conventional dicentric analyses Tritium accident The accidental overexposure to tritiated water described earlier (Section 9.7.5.7) was also examined retrospectively by FISH . Initially, dicentrics had indicated an average dose of 0.38 Gy, which compared well with 0.47 Gy obtained by measuring tritium in urine. These values are average doses to soft tissues of the body as tritium incorporates into body water, delivering a more or less uniform exposure to all the soft tissues. Subsequent blood samplings showed an expected reduction in
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
dicentric yields consistent with a disappearance half-time of 3.3 years. Six and eleven years after the accident, FISH dosimetry was attempted using the combined yields of one- and two-way translocations. On the first occasion, one laboratory made the analysis by single colour painting chromosomes #1, 2 and 4, and for the second analysis the work was shared with a second laboratory that painted chromosomes #2, 3 and 5. Dose estimates were made by reference to an in vitro calibration made with tritium in one of the laboratories, which yielded a linear dose–response curve for full genome corrected total translocations of Y = C + (5.26 ± 0.49) x 10 –2 D. The combined data from all the FISH scoring produced a dose estimate of 0.48 Gy. Goiânia accident In the Goiânia radiation accident (Brazil, 1987) a large number of persons were exposed when a spent 137 Cs radiotherapy source was broken open [149, 150]. These persons provided a good cohort for a follow-up study. Immediately after detection of the accident, lymphocytes from 129 affected individuals were analysed for the frequencies of dicentrics and rings. Twenty-nine persons had an estimated dose in the range of 0.3 to 5.9 Gy . Although most of the individuals received an inhomogeneous exposure, suggested by the presence of localized skin lesions, all cases except six showed a Poisson distribution of aberrations. Some of these victims were followed up over the years by examining the frequencies of dicentrics (analyses started immediately) as well as translocations using FISH (started after five years) for retrospective radiation dosimetry . 97 Data on translocation frequencies (using various probe cocktails, covering about 80% of the genome) obtained by FISH could be directly compared with the baseline original frequencies of dicentrics from the same persons . The frequencies of translocations observed years
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
after radiation exposure (from 1992 onwards) at higher doses (>1 Gy) were two to three times lower than the initial dicentrics determined in 1987. For exposure levels estimated to be <0.9 Gy small differences were found between the frequencies of translocation and the initial dicentric yields. The accuracy of these dose estimates might be increased by scoring more cells. However, factors such as the persistence of translocation carrying lymphocytes, translocation levels not proportional to chromosome size, and interindividual variation reduce the precision of these estimates. No decline in one- and two-way translocations during the follow-up was found , which is similar to the Chernobyl studies. Straume et al. also evaluated two victims of the Goiânia accident one year after their exposure using FISH. When the data were compared with dicentrics frequencies obtained immediately after the accident, lower translocation frequencies were observed. German and Estonian accidents By contrast, in another study undertaken 11 years after an accident involving three radiation workers, FISH, using chromosomes #2, 4 and 8 and pan-centromere probe, gave stable translocation frequencies that were not significantly different from the mean dicentric frequencies determined by conventional FPG staining shortly after detection of the accident. About 75% of the translocations were identified as two-way types. Following a radiation accident in Estonia in 1994, chromosomal analyses were carried out after one month and subsequently 2, 6, 10, 12, 17, 22 and 24 months after exposure of five individuals assessed to have received approximately 1 to 3 Gy [213, 246]. In the follow-up studies, two-way translocations remained relatively stable in all five subjects, and in one person a significant decrease in one-way translocations was observed. Dicentrics decreased in all subjects to about 50% of the initial frequencies by 12 months post-exposure . A further follow-up study spanning the 7 years
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
after the accident indicated that scoring translocations in stable cells appeared to abolish the decline in translocations observed in all cells. In stable cells, the yield of translocations was independent of time during the first years of follow-up . Istanbul accident In Section 9.7.5.6, a case is described where several persons were irradiated by an unshielded 60 Co source mixed with scrap metal. One month elapsed between the accident and recognition by the authorities that exposures to radiation had occurred. The patients had considerably depressed blood cell counts. For the five most seriously exposed persons, dicentric analysis indicated doses ranging from 2.2 to 3.1 Gy. This includes using the G function adjustment to the dose–response curve to take account of their exposures being protracted over several hours. In Section 9.7.4.4, it is noted that when exposures are sufficiently large to cause deterministic effects, such as lowered blood counts, the dicentric yields may decline appreciably over a period of a few weeks. FISH dosimetry was also performed with the same blood specimens as were used for the dicentric assay. The analyses were performed in three laboratories , and the resultant dose estimates were based on the combined yields of one- and two-way translocations pooled from the laboratories. The FISH dose estimates, which also include the G function adjustment, were 20 to 30% higher than the values derived from dicentrics. FISH is usually considered for deployment as a retrospective dosimeter where blood sampling occurs on a timescale of years after irradiation or where long term exposures have occurred, for example, from environmental contamination with radionuclides. This case has, however, illustrated rather well that FISH also has a role in cases where high doses are 98 received, with moderately delayed sampling, on a timescale where, for lesser doses, the dicentric assay is usually
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
considered to be quite adequate. Georgian accident Eleven young frontier guards were accidentally exposed to one or several 137 Cs sources with activity not exceeding 150 GBq, at the Lilo military training centre. The sources were intended for training and instrument calibration purposes. The victims were irradiated for approximately one year, from mid-1996 to April 1997 . Four most exposed persons were hospitalized in France where cytogenetics was done in November 1997 (Table 13). TABLE 13. NUMBERS AND, IN BRACKETS, FREQUENCIES PER CELL OF UNSTABLE CHROMOSOME ABERRATIONS aPatient Scored cells Dic Rc Ace u test Dose Dic Dose, FISH Gy [95% CI] Gy [95% CI] 1 500 14 (0.03) 0 (0.000) 11 (0.022) -0,43 0.4 [0.2–0.6] 0.7[0.4–1.0] 2 500 19 (0.04) 1 (0.002) 15 (0.030) -0,59 0.5[0.4–0.7] 0.4[0.1–0.7] 3 502 55 (0.11) 4 (0.008) 24 (0.048) 4,68 1.1[0.9–1.3] 0.8[0.6–1.1] 4 518 80 (0.15) 4 (0.008) 25 (0.048) 3,61 1.3[1.1–1.5] 1.7[1.4–1.9] a The u-test indicates significant overdispersion, u >1.96; or underdispersion, u 1.96) suggesting a partial-body exposure. This is in agreement with the circumstances of exposure as reconstructed by physical dosimetry. All four patients probably suffered from lymphopenia before their arrival in France and therefore, based on unstable aberrations (Section 9.7.7.4), the averaged whole-body dose estimates may be underestimates. Therefore
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
the FISH translocations assay was also carried out, considering all cells and not confined to stable cells only. Three pairs of chromosome (# 2, 4 and 12) painted together with a pan-centromeric probe. For person 2 no difference was seen in the doses estimated by using dicentrics or translocations yields (Table 13). For persons 1 and 4, the FISH values were higher than for dicentrics, but the differences were not statistically significant. For person 3 however, a higher dose was estimated using dicentrics. These differences can probably be explained by the heterogeneity and fractionation of exposures, which differed markedly from one patient to another, and therefore modified the distributions of translocations in unstable cells and consequently the relative disappearance of dicentrics compared to translocations. Further follow-up cytogenetics were undertaken (Figs 33 and 34) although samples were not available for each person on each occasion. 99 0 0,4 0,6 0,8 11,2 1,4 1,6 1,8 21 2 3 4 Patient oct-97 + 2 month + 2 years + 3 years Dose (Gy) 0 0.5 1 1.5 2 2.5 1 2 3 4 Dose (Gy) oct-97 + 2 month + 2 years + 3 years Patient FIG. 33. Changes in dicentric derived dose estimates with time after irradiation. FIG. 34. Changes in two-way translocation derived dose estimates with time after irradiation. 100 As expected, a decrease in the dicentrics yield was observed over time in all patients. By contrast, no decline in the frequency of two-way translocations was observed for three of the four patients. For person 1 the frequency of translocations decreased two months after the first blood sample was taken but, given the uncertainties, is not statistically significant. The general stability of the later translocations data probably reflects the rapid lymphocyte turnover associated with their lymphopenia and rapid elimination of unstable
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
aberrations. The later FISH data are possibly indicating the dose received by the bone marrow stem cells. > 101 11. PREMATURE CHROMOSOME CONDENSATION (PCC) ANALYSIS Biological dosimetry is generally performed by analysing dicentrics and/or translocations at the first mitosis following in vitro PHA blastic transformation. These assays have several recognized problems, namely radiation induced mitotic delay and cell death during the two day assay culture that operate especially after high doses, which can cause considerable underestimation of the radiation exposure dose . This section describes techniques for inducing the chromosomes to condense prematurely, i.e. at some time before the first mitosis and so reduce or eliminate the culture time and hence the opportunity for mitotic delay or death to occur. 11.1. PCC BY MITOTIC FUSION The induction of PCC by fusing human lymphocytes with Chinese hamster ovary (CHO) mitotic cells in the presence of a fusing agent, polyethylene glycol (PEG), enables one to measure the chromosomal aberrations immediately following irradiation without the need for any mitogen stimulation or culturing . The use of this PCC method, in combination with conventional techniques such as C banding or FISH with chromosome specific DNA libraries with or without a pan-centromeric probe, permits the detection of breaks, dicentrics and rings as well as translocations. This assay has been proposed as a biodosimetric tool by analysing the frequencies of chromosomal aberrations, i.e. excess of breaks, dicentrics and translocations [67, 72, 75, 250]. The assay is useful to determine exposure to low doses as well as following life threatening high acute doses of low and high LET radiation. Moreover, it can discriminate accurately between total and partial body exposures . Since with this PCC assay the number of normal cells reflects more accurately the proportion of unirradiated lymphocytes, this method is efficient for detecting even a
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
small spared fraction (as low as 5%). The assay similarly could also be better able to quantify small localized burns from partial body exposures. > 11.1.1. Cell culture and cell fusion conditions 11.1.1.1. Using CHO mitotic cells Chinese hamster ovary (CHO) mitotic cells should be prepared before performing the PCC analysis. CHO cell cultures are typically set up from stock lines. These are readily available, very easy to handle and have a short cell cycle of approximately 12 hours. CHO cells can be grown in 750 mL tissue culture flasks or roller bottles, in complete medium (composed of F-10, 15% newborn calf serum and antibiotics). Mitotic cells can be obtained by adding Colcemid (at a final concentration of 0.1 μg/mL) for 4 to 6 hours when cultures are half full, followed by mitotic shake-off. Mitotic cells from one flask or roller bottle can be isolated several times per day; therefore, after each isolation, fresh medium supplemented with Colcemid should be replaced in the flasks containing the remaining cells. The mitotic cells can be prepared in large quantities in advance and kept in a freezer at -80 to -110°C before use. 11.1.1.2. Isolating lymphocytes Generally, Ficoll Hypaque should be used for isolating lymphocytes as described in the earlier Section 9.1.5.2. This has the advantage that when enough lymphocytes are isolated, a part can be used immediately and the rest frozen at -80 to -120°C for future use, if found necessary. 11.1.1.3. Fusing agent Generally, polyethylene glycol (PEG) of molecular weight 1450 should be used, and the desired concentration for fusion is 40% to 50% w/v (in F-10 medium without serum, PBS, or preferably in RPMI-1640 medium with HEPES without serum). > 103 11.1.1.4. Fusion and chromosome condensation processes To induce prematurely condensed chromosomes, lymphocytes should be fused with mitotic CHO cells
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
(ratio 5:1) that possess a mitotic promoting factor in the presence of PEG. The fusion process takes only 4 min (1 min in PEG alone, and then wash medium F-10 should be added gradually). This is followed by a one hour incubation in complete medium with Colcemid at 37°C [67, 72, 251]. 11.1.1.5. Fixation procedures In principle, this is similar to the method used for metaphases (Section 9.2) but the optimum timings and concentrations vary slightly. Lymphocytes should be treated with a hypotonic solution of KCl (0.075M) and kept in a prewarmed water bath (at 37°C) for 3 to 4 min, and, following centrifugation, cells can be fixed in a mixture of acetic acid:methanol (1:3). Slide preparation is performed by using the standard technique that is similar to other assays described earlier (see Section 9.2). 11.1.1.6. Staining procedures The choice of staining technique depends on the biological end point to be analysed, as follows: (1) Standard chromosome breaks analysis. For the purpose of analysing chromosomal aberrations as radiation induced chromosome breaks, slides can be stained with conventional Giemsa (Gurr improved R66) or the FPG technique as was already described in Section 9.3.2 (Fig. 35) [67, 252]. The protocol using FPG was developed for PCC with cells where the two fused chromosome complements are completely intermingled. It is probably not necessary for the lymphocyte technique described here because the two sets of chromosomes tend to remain in two groups as shown in Fig. 35 where it can be seen that the single stranded human chromosomes are clearly distinguishable. A disadvantage of FPG staining is that chromosomes tend to swell and this may hinder accurate scoring of PCC fragments as small adjacent swollen objects could touch and appear to be a single structure. > 104 FIG. 35. Human G 0 PCC with
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
some fragments arrowed produced by the mitotic fusion method. (2) Dicentric analysis using C banding. In addition to chromosome breaks, dicentrics can also be visualized. For this, slides should be pretreated with barium hydroxide and salt solution, (Section 9.3.3) followed by Giemsa staining, which highlights the centromeric region of all chromosomes so that dicentric chromosomes can be easily distinguished from monocentrics (Fig. 36) [70, 79]. 105 FIG. 36. PCC stained by the C-banding method showing two dicentrics (d). (3) Translocation and dicentric analysis using the chromosome painting assay. The frequency of radiation induced translocations and dicentrics can also be determined with the PCC preparations by employing the FISH technique, using either chromosome painting probes alone or in combination with a pan-centromeric probe (Fig. 37). The latter gives more accurate discrimination between translocations and dicentrics . > 106 FIG. 37. PCC with double coloured FISH, the combination of chromosome paint (#8) and a pan-centromeric probe for whole genome. In unirradiated control (A) normal PCC. In irradi-ated cells, arrows indicate (B) excess of break in PCC, ace(b); (C) dicentric, dic (BA), bicoloured fragment, ace(ba); (D) ring, r(B); (E), (F) terminal translocation, t(BA) and t(AB) . > 11.1.2. Analysis Criteria for analysis of slides are similar in part to those described in Section 9.4 (i.e. coding slides, scanning parameters, etc.). PCC spreads can be located manually or by use of automated metaphase finder systems that are in more general use (Section 13.3) . It is advisable to facilitate scoring by using a recording system that permits marking each chromosome piece on a drawing or image of the PCC spread. A microscope attachment (camera lucida) can be used to visualize PCC spreads on a much larger scale and record markings on a drawing. Some metaphase finding systems are equipped with specialized applications that allow
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
annotation of digitized images. The microscope stage coordinates of PCC spreads on slides should be recorded and the selection method of candidate PCC spreads for scoring should not introduce bias to distort aberration yields. > 107 Analysis involves counting the number of chromosome elements, which appear as single chromatids and can be discriminated easily from the CHO mitotic chromosomes in the human interphase PCC spreads following Giemsa staining. When the FPG technique is used the human chromosomes appear darkly stained while the CHO cells, which were grown for more than two cell cycles in medium supplemented with BrdU, display the harlequin effect and appear very lightly stained (see Fig. 22). When the FISH assay is used, cot-1 hamster DNA can be used to mask all signals in the CHO chromosomes so that only the appropriate human PCC are highlighted (Fig. 37). 11.1.3. Scoring criteria The appearance of the PCC can be used to define easily the cell cycle position of the lymphocytes at the time of their treatment. Cells that were in G 1 , S and G 2 appear as single chromatid, pulverized chromosomes and having two chromatids, respectively. For biological dosimetry with Giemsa stained preparations, one scores only the spreads comprising single chromatids, i.e. human lymphocytes that were treated in G 0 /G 1 , and each element represents one human chromosome (Fig. 35). Therefore, in unirradiated lymphocytes 46 elements will be scored. The number of chromosome elements in the exposed samples is recorded, and the induced frequency is estimated by simply subtracting the value obtained in untreated samples. In cases of suspect partial body exposures an alternative analysis method, Qpcc, which involves the analysis of the yield of excess PCC fragments in damaged cells (containing excess PCC fragments) has been introduced . This method is identical in
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
concept to the Qdr method introduced by Sasaki and Miyata (see Section 9.7.4.3). Following C banding or using a pan-centromeric probe and chromosome specific DNA libraries employing the FISH technique, slides can be scored for the presence of dicentrics and/or translocations (see Fig. 37), recorded and analysed, as described in Sections 9.4 and 10.4. 11.2. PCC BY CHEMICAL INDUCTION 11.2.1. The rapid interphase chromosome assay (RICA) This assay also removes the need for prolonged culturing. Lymphocytes isolated from blood by the Ficoll Hypaque method (Section 9.1.5.2) are placed in culture medium containing a phosphotase inhibitor such as okadaic acid or calyculin A, adenosine triphosphate and p34 cdc2 /cyclin B kinase and incubated at 37 oC for just 3 hours. The full protocol is described by Prasanna et al. . Fixation and spread preparation (hypotonic potassium chloride; 3:1 methanol: acetic acid; dropping onto cleaned slides) follows the normal procedures used for metaphases. Radiation induced damage is then analysed after in situ hybridization and chromosome painting by fluorescence microscopy (Fig. 38). 108 FIG. 38. Photomicrographs showing FISH painted human chromosome 1 (red) and 2 (green) in interphase lymphocytes irradiated by 60 Co γ-rays and visualized by the RICA assay . Normal cell producing two red spots and two green spots (a & b), aberrant chromosome 1 producing more than two red spots (c), aberrant chromosome 2 producing more than two green spots (d), cells with more than two green and red spots (e & f) (courtesy Pathak and Prasanna, AFRRI, USA). Normal cells display two fluorescent spots per chromosome, whereas cells with structural aberrations (breaks and exchanges) involving specific chromosome(s) corresponding to painted probe(s), may show more than two spots. Using a single large chromosome probe is adequate for biological dosimetry . However, using more than one chromosome probe improves sensitivity
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
. 11.2.2. The PCC ring assay Among the chemically-induced PCC methods for biological dosimetry, a simple and useful procedure is the scoring of rings in Giemsa-stained chromosomes. This technique still requires the lymphocytes to be cultured and the method, described by Kanda et al. recommends 48 hours cultures. It therefore does not save time but the PCC-rings assay is particularly applicable to high overdoses in the range where the dose response for the conventional dicentric assay shows signs of saturation. It has been calibrated and used for doses up to 20 Gy equivalent to X rays. At such a dose the number of induced dicentrics and fragments is too large for reliable scoring. However in lymphocytes rings are induced at a much lower frequency, often ~10% of that of dicentrics, and this makes ring scoring a feasible endpoint after a very high dose. 109 Preparation of chemicals Inhibitors of DNA phosphorylation such as okadaic acid or calyculin A should be prepared. These chemicals are carcinogenic and therefore should be handled with appropriate safety precautions. Calyculin A can induce PCC about 20 times more effectively than okadaic acid, although their mechanisms of PCC induction are probably similar as judged by the dose-dependence and the resulting chromosome morphology. Okadaic acid or calyculin A is dissolved in dimethylsulphoxide (DMSO), diluted with medium and stored at -20°C as a stock solution (e.g. 5–10 μM). Culture Chemically-induced PCC in lymphocytes generally requires the cells to be cycling. Therefore the procedure is to PHA stimulate and culture the cells for 48 hours using a method similar to that described in Section 9.1 for obtaining metaphases. Although PCC can be induced in whole blood cultures, using isolating lymphocytes produces cleaner preparations with a lot of cells (described in Section 9.1.5.2). Thus, especially in the case of
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
very high-dose exposure, isolating lymphocytes is strongly recommended. The standard protocol for PCC induction is that okadaic acid (500 nM) or calyculin A (20–50 nM) is added to the cultures during the final hour and this will therefore produce a mixture of PCC cells in all stages of the first cell cycle. However, the effectiveness of the chemicals might be dependent on the culture condition and drug quality. The concentration and period of optimal treatment should be determined based on the incidence of cells exhibiting PCC and the quality of chromosome morphology in each laboratory. Insufficient treatment results in a lack of analysable cells, whereas overtreatment results in fuzzy and too condensed chromosomes. The fixation, slide preparation and Giemsa staining procedures are similar to the methods used for metaphases. 11.2.2.2. Scoring criteria In 48 hours cultures of highly irradiated lymphocytes most analysable cells are between the late G 2 phase and metaphase. With low-dose exposures there may be contamination with cells at anaphase. Compared with the appearance of ring chromosomes in metaphase spreads (Fig. 11), PCC rings in late G 2 and anaphase (Fig. 39A) cells are narrow which makes their identification easy. Therefore, these cells are preferred for scoring PCC rings. Cells at the late G2 phase and those at anaphase can be distinguished by having attached or separated sister chromatids, respectively (Fig. 39B). 11.2.2.1. Cell culture, chemical treatment and slide preparation This is described in a step by step detailed protocol in Annex III. 110 FIG. 39. Examples of okadaic acid induced PCC of irradiated lymphocytes at different cell cycle phases. (A) G2/M-PCC cells, (B) M/A-PCC cell exhibiting separated sister chromatids. Arrows indicate ring chromosomes . The frequencies of PCC rings are not significantly different between late G 2 cells and anaphase cells, and these data can be
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
pooled. A circular shaped chromosome is scored as a PCC ring. The centromeres are not clearly visualized in PCC cells stained just with Giemsa so that PCC rings are not classified into centric or acentric forms. Just as with dicentrics, (Section 9.7.4.3) the analysis of the intercellular distribution of PCC rings compared with the Poisson distribution provides some information regarding uniformity of exposure to low LET radiation or the radiation quality involved in an accident. In situations e.g. delayed discovery cases, where there is a time gap between irradiation and blood sampling it should be possible to use a half-life calculation to adjust the observed yield to obtain an estimate of the original PCC ring frequency. At present there are little firm data to support this. However the cytogenetic follow-up study of one survivor of the Tokai-mura accident reported a half-life of about 8.7 months . 11.3. A RADIATION ACCIDENT INVESTIGATED BY THE PCC RINGS METHOD Soon after the PCC rings technique, calibrated in vitro with 200 kV X rays, was published , the opportunity arose to examine a serious radiation accident using the method where PCC was induced by okadaic acid. Biological dosimetry was performed on three seriously exposed victims of the Tokai-mura criticality accident in Japan in 1999 . The frequencies of PCC rings per 100 cells from samples obtained 9 hours after the accident were 150, 77 and 24, which, respectively, led to dose estimates of >20, 7.4 (95%C.I. 6.5–8.2) and 2.3 (1.8–2.8) Gy-Eq. One should bear in mind that the exposures were to a mixed field of gamma and neutron radiation, and the equivalent dose measured in Sv (Section 2) is inappropriate to use at such high doses because it is based on the judged risks of stochastic effects at low doses. The organ RBE-weighted dose
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
was specially defined for characterizing high dose exposure as a product of absorbed organ dose and RBE in order to evaluate onset of deterministic health effects . The RBE of 200 kV X rays is set to 1. The unit of RBE-weighted dose is J kg -1 and is called in the gray-equivalent (Gy-Eq). For the most highly irradiated person the dose could only be approximated to >20 Gy-Eq because the 111 published in vitro calibration , showed a levelling off (saturation) in the linear quadratic dose–response for whole body RBE-weighted dose approaching 20 Gy-Eq of 200 kV X rays (or 20 Gy of whole body absorbed dose since RBE is 1). Parallel analyses of the blood samples were also made by conventionally scoring dicentrics and rings (dic+rc) in metaphases. Because such high exposures had occurred, with a consequential rapid fall in peripheral lymphocyte counts, the cells were cultured by a method to maximize the likelihood of obtaining metaphases . This method concentrates the lymphocytes using a Ficoll Hypaque column and is similar to that described in Section 9.1.5.2. This yielded, for the most heavily irradiated patient, 715 dicentrics and 188 centric + acentric rings in 78 cells where every metaphase was damaged. The corresponding yields for the other two persons were 479 dicentrics and 55 rings in 175 cells and 191 dic + rc in 300 cells. Table 14 taken from summarizes the resultant estimates of doses by the cytogenetic methods and also by physical measurements using sodium activation analysis. TABLE 14. COMPARISON OF THE DOSES ESTIMATED BY VARIOUS INDICATORS Patient Estimated Whole Body RBE-weighted Dose (Gy-Eq) a by PCC-Ring Dic Dic+R/Rc 24 Na bA >20 22.6 24.5 17–24 B 7.4(6.5–8.2) 8.3 8.3 8.7–13 C 2.3(1.8–2.8) - 3.0(2.8–3.2) 2.5–3.6 a RBE set 1 for X- (patients A
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
and B) or γ-rays (patient C). b Ishigure et al . , where the neutron RBE is assessed at 1.5–2.0. 112 12. THE CYTOKINESIS-BLOCK MICRONUCLEUS (CBMN) ASSAY 12.1. BACKGROUND Ionizing radiation induces the formation of acentric chromosome fragments and to a small extent malsegregation of whole chromosomes. Acentric chromosome fragments and whole chromosomes that are unable to interact with the spindle lag behind at anaphase, and as a result they are not included in the main daughter nuclei. A lagging chromosome fragment or whole chromosome forms into a small separate nucleus; hence the term micronucleus. The peripheral blood lymphocyte MN assay based on MN expression in short term culture of lymphocytes was first described by Countryman and Heddle . However, in this original method no attempt was made to determine whether the cells scored had actually completed nuclear division in vitro which made the assay unreliable because chromosome damage in cells can only be expressed as micronuclei if cells divide. A more reliable approach was eventually developed based on the use of the cytokinesis inhibitor, cytochalasin-B. Using cytochalasin-B, Fenech and Morley were able to demonstrate in 1985 [83, 84] that cells that had completed one nuclear division could be accumulated and recognized as binucleated (BN) cells. MN could then be specifically and efficiently scored in these BN cells while excluding non-dividing mononuclear cells that were unable to express MN in vitro (Fig. 19). Consequently, the results obtained with the MN assay are not confounded by interindividual and interexperimental variation in the frequency of dividing cells, which has been shown to have a profound effect on the observed MN frequency [84, 256, 258]. The resulting cytokinesis-block MN (CBMN) assay has since become the standard method for measuring MN in cultured lymphocytes. Lymphocytes collected in a blood sample are themselves the result
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
of cell divisions occurring in vivo . One might therefore expect that some may already contain MN. Thus, it has been shown that scoring MN in mononucleated lymphocytes in conventional blood smears could be particularly useful for monitoring genetic damage in chronically exposed populations [259–263]. Furthermore, scoring of MN in mononucleated cells could also be used as an interesting additional parameter in the CBMN assay [262, 263]. In the 1990s the CBMN–centromere assay was developed, using FISH and a pan-centromeric probe to visualize centromeres. This method allows discrimination between MN containing acentric fragments and whole chromosomes [69, 85, 263–267]. Applying this method, the sensitivity of the CBMN assay can be substantially increased in the low dose range [85, 266, 267] (see Section 12.4.2). More recently a more comprehensive version of the CBMN assay known as the Cytokinesis-Block Micronucleus Cytome (CBMN Cyt) assay has been developed and validated which, apart from MN in binucleated and mononucleated cells, also includes measurement of nucleoplasmic bridges (NPB, Fig. 19C) and nuclear buds in binucleated cells which are biomarkers of dicentric chromosomes and gene amplification respectively. Furthermore, in the CBMN Cyt assay the proportion of mono-nucleated, binucleated and multinucleated cells as well as necrotic and apoptotic cells is scored which provides measures of cellular proliferation and cell death that can also be informative in biological dosimetry [88, 268]. It is also possible to score micronuclei in erythrocytes as a biomarker of chromosome damage noting that the method has an upper limit of detection of 1 Gy and samples need to be collected as soon as possible after exposure due to inhibition of erythropoiesis. Recently the flow cytometric in vivo MN assay in immature mouse erythrocytes has been adapted for use in humans, by restricting MN scoring to the transferring receptor positive reticulocytes
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
> 113 (Tf-Ret; CD71) . Evaluation of the reticulocyte assay in patients treated with radioiodine for thyroid cancer shows that the method may be of use for monitoring individuals after suspected accidental radiation exposure [270, 271]. 12.2. LYMPHOCITE CULTURE FOR CBMN ASSAY The lymphocyte culture method is similar to that described in Section 9.1 for obtaining metaphases. The main differences, however, are that (i) Cyt-B is added to the cultures at 24 or 44 hours (24 hours is preferable for radiation biological dosimetry to ensure only first division cells are captured), (ii) bromodeoxyuridine and Colcemid are not used, (iii) the culture time is extended to 72 hours, and hypotonic treatment, fixation and centrifugation are modified to preserve the cell cytoplasm so that binucleated cells are easily identified. The preparations are either conventionally stained with Giemsa for light microscopy or with a fluorescent dye such as acridine orange for fluorescence microscopy. The preparations can also be further processed to highlight centromeres using FISH and a pan-centromeric FISH probe. Detailed protocols are given in Annex IV. 12.3. CBMN ASSAY SCORING CRITERIA Detailed scoring criteria for all the biomarkers in the CBMN Cyt assay have been published . In this section only scoring criteria for MN and NPB in binucleated cells are provided because these are the best validated biomarkers for biological dosimetry of ionizing radiation exposure. > 12.3.1. Criteria for selecting binucleated cells which can be scored for micronucleus frequency The cytokinesis-block cells that may be scored for MN frequency should have the following characteristics (Fig. 19): (a) The cells should be binucleated (BN). (b) The two nuclei in a BN cell should have intact nuclear membranes and be situated within the same cytoplasmic boundary. (c) The two nuclei in a BN cell should be approximately equal in size, staining pattern and
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
staining intensity. (d) The two nuclei within a BN cell may be unconnected or may be attached by one or more fine nucleoplasmic bridges, which are no wider than 1/4th of the nuclear diameter. (e) The two main nuclei in a BN cell may touch but ideally should not overlap each other. A cell with two overlapping nuclei can be scored only if the nuclear boundaries of either nucleus are distinguishable. (f) The cytoplasmic boundary or membrane of a BN cell should be intact and clearly distinguishable from the cytoplasmic boundaries of adjacent cells. > 12.3.2. Criteria for scoring micronuclei MN are morphologically identical to but smaller than the main nuclei (Fig. 19). They also have the following characteristics: (a) The diameter of MN in human lymphocytes usually varies between 1/16th and 1/3rd of the mean diameter of the main nuclei, which corresponds to 1/256th and 1/9th of the area of one of the main nuclei in a BN cell, respectively. > 114 (b) MN are non-refractile and can therefore be readily distinguished from artefacts such as staining particles. (c) MN are not linked or connected to the main nuclei. (d) MN may touch but not overlap the main nuclei and the micronuclear boundary should be distinguishable from the nuclear boundary. (e) MN usually have the same staining intensity as the main nuclei but occasionally staining may be more intense. Table 15 illustrates a simple layout for a data sheet for recording MN. TABLE 15. LAYOUT OF A MICRONUCLEUS SCORING RESULTS SHEET FOR DUPLICATE CULTURES (1 & 2) FROM A SINGLE BLOOD SAMPLE Sample No: Scorer: Date: Slide No. Micronucleus distribution in BN cells Total No. of BN cells Total No. of Micronuclei 0 MN 1 MN 2 MN 3 MN 4 MN 5 MN > 5 MN 1 500
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
2 500 1 + 2 1000 Remarks: 12.3.3. Criteria for scoring nucleoplasmic bridges A nucleoplasmic bridge (NPB) is a continuous DNA-containing structure linking the nuclei in a binucleated cell. NPB originate from dicentric chromosomes (resulting from misrepaired DNA breaks or telomere end fusions) in which the centromeres are pulled to opposite poles during anaphase (Fig. 19A and C). They have the following characteristics: (a) The width of a NPB may vary considerably but usually does not exceed 1/4th of the diameter of the nuclei within the cell. (b) NPB should also have the same staining characteristics as the main nuclei. (c) On rare occasions more than one NPB may be observed within one binucleated cell. (d) A binucleated cell with a NPB may contain one or more MN. (e) BN cells with one or more NPB and no MN may also be observed. Table 16 illustrates a simple layout for a data sheet for recording NPB. 115 TABLE 16. LAYOUT OF A NUCLEOPLASMIC BRIDGE SCORING RESULT SHEET FOR DUPLICATE CULTURES (1 & 2) FROM A SINGLE BLOOD SAMPLE Sample No: Scorer: Date: Slide No. NPB distribution in BN cells Total No. of BN cells Total No. of NPB 0 NPB 1 NPB 2 NPB 3 NPB 4 NPB 5 NPB > 5 NPB 1 500 2 500 1 + 2 1000 Remarks: It may be more difficult to score NPB in BN cells with touching nuclei, and it is therefore reasonable to specify whether NPB were scored in all BN cells regardless of proximity of nuclei within a BN cell or whether they were scored separately in those BN cells in which nuclei were clearly separated and those BN cells with touching nuclei. There is not enough evidence yet to recommend scoring NPB only in BN cells in which nuclei
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
do not touch. 12.4. CBMN ASSAY DATA HANDLING 12.4.1. Dose–response The procedures for producing in vitro dose–response calibration curves are as previously described in Section 8. Many studies have shown that the number of radiation induced micronuclei is strongly correlated with radiation dose and quality [87, 272–275]. As there are however interlaboratory differences in MN dose response, due to the use of different protocols, scoring criteria, etc., just as for the other assays described in this publication, any laboratory intending to carry out biological dosimetry, should make its own in vitro dose response calibration curves. Ideally, at least 8 doses should be used in the range up to 5 Gy. Curve fitting by linear (high LET) and linear-quadratic (low LET) models follow the procedures described in Section 8. A typical example of a MN dose response curve for low LET radiation ( 60 Co γ-rays, dose rate 0.5 Gy/min) is shown in Fig. 40. 116 020 40 60 80 100 0 1 2 3 4 ## Dose, Gy M icronuclei per 100 cells FIG. 40. Typical linear-quadratic MN dose response curve for 60 Co γ-rays. Solid curve: pooled data from 47 donors; broken curves: the upper and lower 95% confidence intervals. 12.4.2. Background frequency The background frequency of MN is reported to be quite variable; values ranging from 0 to 40 per 1000 BN cells have been recorded [257–286]. The two most important factors influencing MN background frequency, besides dietary factors and exposure to a wide range of environmental clastogens and aneugens, are age and gender [84, 277]. Large scale biomonitoring studies have shown that the spontaneous micronucleus yield increases systematically with age. For a male control population values of 0.35 MN/1000 BN cells/year and 0.44 MN/1000/year were obtained respectively in a study of nuclear power plant and hospital
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
workers [278–280]. These values are in agreement with the large scale study of Fenech of variables influencing baseline micronucleus frequencies: 0.31 MN/1000/year. For a female control population a more prominent increase of 0.58 MN/1000/year was found , again in agreement with Fenech : 0.52 MN//1000/year. Analysis of the MN for the presence of centromeres, by using a pan-centromeric FISH probe (Fig. 41), showed that the age increase of baseline MN frequencies can be attributed almost totally to centromere-positive MN, reflecting an increased chromosome loss with age [266, 279, 280]. 117 FIG. 41. Binucleated cells showing a centromere negative MN (a) and a centromere positive MN (b). Centromeres are stained with a pan-centromeric probe (spectrum orange) and nuclei and MN are counterstained with DAPI. The X-chromosome is almost completely responsible for this spontaneously occurring chromosome loss [281, 282]. This explains also the gender difference in spontaneous MN frequencies where for a population with mean age 41.4 and 41.8 years the mean spontaneous MN frequencies were 16.4 for males and 23.5 per 1000 BN cells for females respectively; by contrast, the difference in centromere-negative MN is not significant: 6.7 versus 7.7 . This background variability clearly poses limitations on using MN as a biological dosimeter for low doses, where pre-existing individual background frequencies are not known. Estimates have been made, suggesting that the CBMN assay in its basic form could only detect in vivo exposures in excess of 0.2–0.3 Gy X rays [87, 266, 283]. As it has been shown that most of the radiation induced MN originate primarily from acentric fragments while spontaneous MN contain especially whole chromosomes [85, 264– 267] the use of the CBMN-centromere assay substantially increases the sensitivity of the CBMN assay in the low dose range [85, 266]. In both studies [85, 266], using a pan-centromeric
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
probe, the majority of spontaneous MN were centromere positive (MNCM +ve )(respectively 73 and 71%) while most radiation induced MN were centromere negative (MNCM -ve ). The number of MNCM +ve only showed a very small increase with dose (respectively 3.7 and 5.3 MNCM +ve per Gy per 1000 BN cells). By manual scoring of MNCM -ve in 2000 BN cells a detection limit, at the 95% confidence limit, of 0.1 Gy can be achieved [266, 267]. 12.4.3. Nuclear division index (NDI )When scoring cytokinesis-block (CB) lymphocyte preparations one observes cells with 1, 2, 3, etc., main nuclei. The relative frequencies of the cells may be used to define cell cycle pro-gression of the lymphocytes after mitogenic stimulation. This is referred to as the NDI . The index is in itself not sufficiently robust for direct application as a biodosimeter. Nevertheless the assay is frequently employed as a useful research tool for understanding the cell cycling kinetics of the cultures. It will indicate perturbations that may be caused by exposure to a mutagen such as radiation. The data arise > (a) (b) 118 directly from the CBMN assay, without any additional laboratory effort, and therefore the method is included in this publication. 12.4.3.1. Criteria for scoring viable mono-nucleated, binucleated and multinucleated cells These cell types have the following characteristics: • Mono-, bi- and multi-nucleated cells are viable cells with an intact cytoplasm and normal nucleus morphology containing one, two, three or more nuclei respectively. • They may or may not contain one or more MN or nuclear buds (NBUD) and in the case of bi- and multi-nucleated cells they may or may not contain one or more NPB. Necrotic and apoptotic cells should not be included amongst the viable cells scored. On rare occasions multinucleated cells with more than four
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
nuclei are observed if the cell cycle time is much shorter than normal or the cytokinesis-blocking time is too long. 12.4.3.2. Calculating the NDI Five hundred viable cells are scored to determine the frequency of cells with 1, 2, 3 or 4 nuclei and the NDI is calculated by using the formula (36): N MMMMNDI 4321 432 +++= (36) where: M1 to M4 represent the number of cells with one to four nuclei, and N is the total number of viable cells scored. The published methods for calculating NDI did not consider its uncertainty. Indeed, a method for deriving the uncertainty does not seem to have been published subsequently. It is therefore described here and due to its complexity, a fully worked example is shown in Annex IV-4. As the values M 1 to M 4 are correlated, the uncertainties on the NDI cannot be calculated using standard error analysis. Instead, the covariance, which measures how the variables are dependent on each other, must be taken into account. The values M 1 to M 4 in the NDI can be assumed to form a multinomial distribution, which means that there is a fixed number (three or more) of possible outcomes for numbers of nuclei in a cell — i.e. 1, 2, 3 or 4 in this case. The variance (var) and covariance (covar) of each variable, M 1 to M 4 can then be calculated using Eqs (37) to (39): )p(np `) (M iii −= 1var (37) > jiji pnp MMar −=`) `, (cov (38) for i and j = 1, 2, 3 or 4 where: M1`, M 2`, M 3` or M 4` are the values of 1 x M 1 , 2 x M 2 , 3 x M 3 and 4 x M 4 , 119
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
n is the sum of the total number of cells times their respective numbers of micronuclei (equivalent to the numerator of the NDI equation), and pi and pj are the probabilities of Mi` and Mj` which are equal to Mi ` or Mj` divided by n.Using the definition of covariance, it can then be shown that the variance of NDI is dependent on the variances and covariances calculated using Eqs 37 and 38: # ∑ ∑ ∑= = +=+= 4141412 `) ``cov( `2)var( `)var( i i ij jijiii MMMMMMNDI (39) This calculation is relatively complex but can be easily carried out using one of the widely available statistical packages or the Dose Estimate program mentioned in Section 8.3. 12.5. APPLICATION OF THE CBMN ASSAY FOR BIOLOGICAL DOSIMETRY 12.5.1. Patient studies To verify the applicability of the CBMN method in biological dosimetry, MN yields were measured in peripheral blood lymphocytes of 1) different groups of cancer patients receiving fractionated partial body radiotherapy, e.g. prostate, cervix, Hodgkin’s disease [285–288] and 2) thyroid cancer patients undergoing radioiodine treatment [289–291]. These studies showed that the doses estimated by MN agreed quite well with averaged whole body doses calculated from the radiation treatment plans plus cumulative dose–volume histograms [285–287, 292]. A meta-analysis, focused on thyroid cancer patients , showed that the post-radiation MN induction increased more than three times in comparison with the pre-irradiation frequency, demonstrating that the CBMN assay is sensitive enough to detect the genetic damage in circulating lymphocytes from exposure to low averaged whole body dose from internally incorporated radiation sources. 12.5.1.1. Radioiodine case study The CBMN test was used to study the response of lymphocytes of a 34-year male following treatment with 131 I ablative radiation therapy after a total thyroidectomy for cancer . Fortuitously, several months before diagnosis the patient
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
had volunteered a blood sample for an in vitro study of micronucleus expression following external exposure to graded doses of X rays (198 mGy/min). The background frequency (pre-treatment baseline) in the unexposed culture showed a mean frequency of 6.0 MN per 1,000 binucleated (BN) cells while mean values of 18.5, 29.0, 41.0, 61.0 and 75.5 MN per 1,000 BN cells were found following X ray doses of 50, 100, 150, 200 and 250 mGy, respectively. The data were found to fit a non-threshold, linear dose-response function (Y = 3.714 +2.783D; r=0.99) as shown in Fig. 42. 120 FIG. 42. In vitro MN dose response for low doses of X rays to the patient’s lymphocytes before diagnosis and after 131 I therapy (courtesy Livingston, REAC/TS, USA). Blood was taken 11 days after the first in vivo 131 I treatment with 48 mCi (1.78 GBq) and at monthly intervals thereafter and eventually at quarterly intervals out to five years. The first post-treatment sample showed 35.5 MN per 1,000 BN cells and the six-fold increase above the pre-treatment baseline suggests a dose to the peripheral blood of about 110 mGy. Twenty-six months after the first 131 I treatment a second treatment of 390 mCi (14.46 GBq) was administered to the patient which resulted in a further increase in micronuclei. The micronuclei count fluctuated widely over time and was about 10-fold higher than the pre-treatment baseline value after 5 years of follow-up (Fig. 43). 121 FIG. 43. A five year MN follow up of the patient before, during and after 131 I therapy treatments (courtesy Livingston, REAC/TS, USA). More than 15 years after the second treatment the patient was cancer-free and healthy. Results of this study support the conclusion that the CBMN test is a rapid, sensitive and quantitative biomarker of radiation exposure. However
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
such studies are not able to determine local dose to the target tissue which in this case was any residual thyroid cells plus metastases of thyroidal origin. 12.5.2. Biomonitoring studies After having been validated as an in vivo biomonitor in several patient studies, the CBMN assay as well as the CBMN-centromere assay have been applied for large scale biomonitoring of occupationally exposed radiation workers, e.g. nuclear power plant and hospital staff [266, 278–280, 293–295]. These biomonitoring studies showed the dependence of MN on the accumulated dose received over the years preceding the venipuncture. In the study of Thierens et al. a linear regression of the individual micronucleus frequencies, corrected for the age effect (see Section 12.4.2.), showed an increase of 0.0175 MN per 1000 BN cells/mSv with a Pearson correlation coefficient value of 0.10. Application of the CBMN-centromere assay in a second study of radiation workers by Thierens et al. resulted in almost the same increase of MN with dose, 0.025 MN per 1000 BN cells/mGy and demonstrated that this dose dependence is completely due to MNCM -ve pointing to the clastogenic action of ionizing radiation. Dose dependence of MN in an occupational exposure setting was also found in a study by Vaglenov et al. . They reported an increase of 0.03 MN per 1000 BN cells/mGy. Large scale biomonitoring studies show that the micronucleus assay is able to demonstrate genetic damage at the population level for accumulated doses received occupationally exceeding 50 mGy. 122 12.5.3. Accident studies 12.5.3.1. Chernobyl accidental The CBMN assay has also been used successfully for assessing the protracted exposure due to incorporation of long lived radionuclides by residents in the vicinity of the Chernobyl nuclear power plant. Eighty individuals who were located between 100–200 km from Chernobyl at the time of the accident
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
in 1986 were tested for their MN frequency in BN lymphocytes between 1989 and 1991 . In this study whole body counts for 134 Cs and 137 Cs were performed, so that the MN frequency could be related to body dose. Multiple regression analysis of the data from the 80 subjects showed that (a) the MN frequencies were significantly associated with the radiocaesium activity level (p = 0.004) and (b) the estimated internal absorbed dose (which ranged from 0.6 to 9.2 mGy) was significantly and positively correlated with MN frequency (R = 0.71). 12.5.3.2. The Istanbul accident For accidents involving a few subjects and where speed in obtaining results has not been so vital, most laboratories have chosen to use the dicentric assay. Thus there are few published accounts of MN being used as a biological dosimeter soon after an accident. One example, however, is the accident in Istanbul [158, 159] previously described in Sections 9.7.5.6 and 10.9.3 where ten scrap metal workers were irradiated by an unshielded former radiotherapy 60 Co source. Lymphocytes sampled ~1 month after the exposures were assayed for MN with the CBMN assay as well as for dicentrics and FISH translocations. Using data pooled from two laboratories, MN derived dose estimates were made for eight of the subjects and gave values in the range 0.7–2.7 Gy, in excellent agreement with doses obtained from dicentrics. It was noted in Section 10.9.3 that the dose estimates from FISH were about 20 to 30% higher than those based on the dicentric yields and this was probably due to the subjects’ severely depressed blood cell counts. The same tendency to underestimate doses in such a situation would also apply with the MN assay because this too is a class of damage that has a limited in vivo persistence, especially
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
after high doses. 12.5.3.3. Semipalatinsk Nuclear Test Site The Semipalatinsk nuclear test site area has been highly contaminated with radioactive fallout during 40 years of continual weapons testing (1949–1989). Individuals living near the site have been exposed to both internal and external radiation. Dicentric and MN analysis was performed in people living in different contaminated villages and one control village. A higher incidence of dicentrics as well as micronuclei was found in residents of the contaminated areas and this higher incidence seems to be mainly caused by their internally incorporated radionuclides . 12.5.3.4. Accident with a 50 kV contact radiotherapy X ray device In 2003, the CBMN assay was applied for retrospective assessment of the dose received by a hospital worker, who was exposed accidentally by a 50 kV contact radiotherapy X ray device during maintenance . A dose estimate of 0.73 Gy was obtained with 95 % confidence limits of 0.54–0.96 Gy. Dicentric scoring resulted in a dose estimate of 0.62 Gy (range 0.45–0.90 Gy), in very good agreement with the CBMN assay dose. A skin injury on the back of the worker indicated that the overexposure was a partial body irradiation. From the overdispersion of the dicentrics data it was deduced that a fraction of 49 % of the body was irradiated. It was not possible to apply this type of analysis to the MN data as MN invariably exhibit overdispersion, even in the case of a total body irradiation. A second blood sample, taken 1 year later, showed that the MN yield decreased with time post-exposure. The 123 disappearance half-time was 342 days; very close to a value of 377 days obtained from dicentrics. This result is in agreement with the decline in the micronucleus frequency with post-irradiation time down to about 60 % at 1y post-treatment,
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
observed in radiotherapy patients . 12.5.3.5. Large scale radiation accidents In case of large scale radiation accidents, when hundreds of people may be exposed, it is important to distinguish the severely exposed individuals ( ≥ 1 Gy), who require early medical treatment, from those less exposed. For this purpose, a rapid biological dosimetry assay is needed. In a recent study the efficacy of automated MN scoring has been confirmed for fast population triage in a multicentre setting. More detailed information is given in Section 13.3.3. > 124 13. AUTOMATION OF CHROMOSOMAL ASSAYS For efficient preparedness for response to radiation events involving mass casualties, it has become imperative to automate cytogenetic dose assessment methods to increase throughput as they are time-consuming and laborious. Moreover, automation also improves quality control and assurance. Furthermore, it also enhances safety of laboratory personnel as the protocol involves processing of blood, which is an occupational biohazard. Cytogenetic laboratory automation involves: (i) automation of sample preparation, (ii) automation of analysis, and (iii) laboratory information management system for sample tracking and data handling . 13.1. AUTOMATED SAMPLE PROCESSING Automated sample-processing in a cytogenetic laboratory may consist of any or all of the following equipment stations: (i) a robotic blood handler, (ii) a biosafety hood, (iii) incubators, (iv) metaphase harvester, (v) metaphase spreader, and (vi) slide stainer. > 13.1.1. Robotic blood handler A customized automated liquid-handling robot for high-throughput processing of blood samples and isolation of lymphocytes from peripheral whole blood can eliminate an important rate-limiting bottleneck in sample processing for cytogenetic dose assessment . The commercially available liquid-handling robots capable of dispensing, diluting, and aspirating blood samples, specifically for blood banking applications may be customized and used for the desired purpose. These systems are precise, accurate, and do not cross contaminate specimens . A customized robotic
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
blood handling station may be equipped with a large customized work deck, a bar-code reader for maintaining chain-of-custody of samples, robotic arms for liquid-handling and transporting of vacutainers, centrifuge tubes, and a wash station for pipette tips. The robot can also be integrated with both a cell viability analyser for correcting for lymphocyte density while setting up cultures and a swinging-bucket automated centrifuge for density gradient isolation of lymphocytes for setting up isolated lymphocyte cultures. However, all equipment must be enclosed in an engineered Biosafety Level 2 environment to ensure sterility of the samples and occupational safety of laboratory personnel. The system must provide a positive chain-of-custody . > 13.1.2. Metaphase harvester To obtain consistently and reliably high quality metaphase spreads, customized commercially available metaphase harvesters can be used for harvesting spreads from a blood culture. These devices eliminate the labour-intensive process by performing repetitious tasks involved in metaphase harvesting from cultures such as centrifugation of cell suspensions, aspiration and safe disposal of supernatant, treatment with hypotonic solution, and fixation of cells with acetic acid:methanol. These steps are carried out under controlled environmental conditions in a one-step protocol without user interaction thereby enhancing the quality and reproducibility of the process . > 13.1.3. Metaphase spreader Metaphase spreading onto glass slides is influenced by temperature and humidity . An automated system provides optimal environmental conditions of temperature and humidity during spreading of the cell suspension onto glass slides with a greater throughput than can be achieved manually. The spreader may be fitted with a microprocessor to precisely balance and control temperature, humidity and drying time. These controls, in conjunction with its functionally derived shape allow different users to obtain consistent results for both human and animal cells. A built-in sealable pipette guide provides consistent sample spreading and helps in preventing sample
|
{
"page_id": null,
"source": 7334,
"title": "from dpo"
}
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.