Chromosomal Analvsis to Assess Radiation Dose

Transcription

1 Chromosomal Analvsis to Assess Radiation Dose J DAVID C. LLOYD National Radiological Protection Board, Chilton, Didcot, Oxon, OX ORQ, UK Key Words. Cytogenetics Radiation. Dosimetry Dicentric. Micronucleus ABSTRACT This brief account of dosimetry based on cytogenetic assays reviews the sensitivity of the method and the speed with which a dose estimate can be obtained. The usefulness of chromosome studies to physicians is also discussed. Medical uses include confirming triage category for highly irradiated subjects and providing evidence for nonuniform exposure or other reasons for the presence of surviving functional stem cells. Such information can assist in medical management decisions, especially regarding transplantation of marrow or stem cells and treatment with growth factors. For subjects exposed to zero or low doses, cytogenetic studies can provide useful information for assisting physicians in counseling. Stem Cells 997;5(supp2):95-2 INTRODUCTION Biological dosimetry using chromosomal damage in peripheral blood lymphocytes after an accidental overexposure to radiation was first performed in 962 on victims of the Recuplex criticality accident in Hanford [l]. This demonstration excited considerable interest, particularly as it involved exposure to neutrons, for which physical methods of dosimetry are difficult. Consequently, decisions were taken in a number of countries to set up laboratories for biological dosimetry, and the method has since been developed into a standardized routine based on the dicentric chromosome aberration. For a detailed account of the radiationinduced aberrations, laboratory procedures, and how doses are calculated from the yields of aberration, the reader is referred to a manual published by the International Atomic Energy Agency ( MA) [2]. Fortunately, it has rarely been necessary to undertake a dose estimation after a criticality accident, but to date several thousand persons have been examined, mostly for exposure to low linear energy transfer radiation. The technique now occupies a valuable niche in the investigation of radiation incidents. It sup plements physical methods of dosimetry, and where these are unavailable, the biological method may be the only means of quantifying the dose. Because of cost the technique is employed only when overexpe sure is suspected or known to have occurred; the method does not replace physical means of routine personal monitoring. Today analysis of chromosomal aberrations is a common feature of the investigation of cases accidentally exposed to ionizing radiation. SENSITIVITY The lower limit of detection of the cytogenetic method is an equivalent whole-body dose of around mgy of y rays, such as from cobalt-6. The technique is possibly a little more sensitive to 25 kvp x-rays and certainly detects doses as low as -2 mgy fission-spectrum neutrons. Dose estimates carry considerable uncertainty because of sampling statistics, particularly at low levels, and are usually expressed as 95% confidence limits. Table illustrates a range of y dose estimates and their limits, showing how the latter depend on the numbers of metaphases scored for dicentrics. Radiation Injury and the Chernobyl Catastrophe. STEM CELLS 997;5(suppl2): AlphaMed Press. All rights reserved.

2 96 Chromosomal Analysis to Assess Radiation Dose Table. The 95% confidence limits on dose estimates from.-3. Gy of gamma rays, showing how the limits are influenced by the numbers of metaphases examined Estimated Dose (GY) Limit Number of cells examined Upper Upper.5 Upper. Upper 2. Upper 3. Upper , I The lower level of dose estimation extends well below the dose at which a physician is likely to consider treatment. By contrast, the upper level of detection is well into the lethal range of exposure, although the measured response tends to flatten off above about 8 Gy y rays because of a saturation phenomenon. Therefore, for exceptionally high exposures, dose estimation is uncertain. This was amply illustrated in the case of a Norwegian fatally exposed in 982 [3]. On the basis of the fmding of five dicentrics per cell, it was interpreted that the dose was well in excess of Gy. It was finally concluded from other methods, such as electron spin resonance of nitroglycerine tablets in the patient s pocket, that the most likely average whole-body dose was approximately 22 Gy. NONUNIFORMITY OF EXPOSURE In practice, accidental irradiations often involve partial or inhomogeneous exposure. It may therefore be necessary to attempt some refinement of the equivalent whole-body dose estimate to indicate the extent of the inhomogeneity. For a uniform irradiation, the distribution of aberrations among the cells is close to Poisson. For nonuniform irradiation, the aberrations are distributed among the cells with a variance greater than that predicted by the Poisson distribution, which can sometimes be demonstrated. By examining the magnitude of this overdispersion it is possible to estimate approximately the proportion of the body exposed and the average dose. The procedure requires a correction factor to take account of in vivo and in vitro selection due to cell death (apoptosis) or mitotic delay which reduces the likelihood of the irradiated fraction of cells coming to metaphase by 48 h in culture. Some simplifying assumptions are also required regarding the distribution of mitogen-responsive lymphocytes within the body. The M A manual [2] gives a more detailed description of the computational steps and some worked examples of this procedure. It may therefore be possible to identify persons who have received a high, life-threatening exposure to much of the body, but with perhaps 5%-% of their tissue being spared or only sublethally irradiated. Such information could be linked to other hematological data collected during the first few days after an accident. These may indicate the likelihood that some areas of active bone marrow may have been spared and, given time and appropriate support, should proceed to natural repopulation. Such information would be of value to physicians in planning the management of bone marrow aplasia. It should be emphasized that the distribution analysis of cytogenetic damage observed in lymphocytes does not help to delineate the actual exposed areas of the body. These may, however, be obvious if doses exceed an erythema threshold.

3 ~ Lloyd 97 Table 2. Lymphocyte chromocome damage obwted in Chemobyl victims pnor to bone marrow transplantation Unique patient number 2 4 5a 6a 9 la 6a 28a 29% Number of cells scored Number of dicentrics scored Number of dicentricfree cells 'Patients who showed own marrow recovery [4]. dh.8 t.4.85 t.7.76 f.4.8 f t t.6.94 t.4.7 t.4.89 t.4.68 f.4 Total number of aberrations scored , Number of aberration-free cells 5 Dose estimates (Gy) from dicentrics Provided that the whole-body dose is not too high, undamaged lymphocytes may also be present in a blood sample taken from a more or less uniformly irradiated person because of the inherent Poisson distrib ution of the damage. This is illustrated by the analysis of samples from some Chernobyl victims. Table 2 shows data on chromosome aberrations from of the 3 seriously irradiated individuals sampled, prior to their treatment by bone marrow transplantation [4]. The remaining three cases are not included because for two of them the lymphocyte cultures yielded no or too few metaphases and, in the third, the cultures were not initiated until after extensive blood transfusions. The statistical test for the Poisson distribution showed values of variance to mean (dy, column 5) for dicentrics to be all somewhat underdispersed, which is a tendency of high uniform exposure [5]. Despite no indications of overdispersion, three cases [5, and 28 contained cells free from chromosomal damage (columns 4 and 7). These three are among the six patients who, before their deaths, showed evidence of grdt rejection, indicating some recovery of their own marrow. Based on this experience, it is to be hoped that in the future such chromosome data would be taken into account in medical management decisions. INTERNALLY INCORPORATED RADIONUCLIDES Cytogenetic analysis is mainly suited to estimating doses after an exposure to penetrating radiation from an external source. Experience has shown that this is what occurs most often in radiation accidents. Dose estimations when one or more radionuclides have been incorporated into the body are less certain [6], mainly because of the localization of radionuclides in particular organs or tissues. This represents another kind of partial-body irradiation, with the added complications of protracted exposure at variable dose rates as the material is lost from tissues by physical decay, biological turnover or excretion. Overdispersed chromosomal aberrations in excess of background have been observed in lymphocytes from persons who have incorporated nuclides, e.g., radioiodine [7] and actinide [8]. An assessment of the dose to the pool of circulating lymphocytes can be made from the yield of chromosomal aberrations. However, this would tend to give a misleading estimate of the whole-body dose, and it generally does not assist in determining dose to the principal tissues of interest, e.g., the thyroid or liver and bone surfaces. Exceptions to this limitation of dose assessment are those chemical forms of nuclides that tend to disperse fairly evenly throughout the body. Tritium, absorbed as tritiated water, is a recognized hazard in some parts of the nuclear industry, and it leads to a fairly homogeneous irradiation because it equilibrates quite rapidly with body water. An industrial accident with tritium [9], which fortunately led only to modest personal doses,

4 98 Chromosomal Analysis to Assess Radiation Dose illustrated that cytogenetic analysis is feasible and produces estimates of dose in good agreement with those calculated from measurements of tritium excreted in urine. Another important exception is the incorporation of isotopes of cesium, as these also tend to give a fairly uniform distribution of dose. The Goihia incident [ can be cited as an exceptionally vivid illustration where members of the general public were exposed externally and/or internally to 37 CsC. Cytogenetic analyses, together with y-ray body monitoring, were the main techniques employed for the assessment of quite large numbers of potentially exposed subjects. SPEED OF OBTAINING A DOSE ESTIMATE Obtaining a dose estimate from chromosomal analysis is time-consuming. Upon receipt of a blood specimen in the laboratory, the normal procedure is to culture the lymphocytes for 48 h and then fix and stain them. Microscope analysis can commence at about 5 h. To score 5 metaphases requires approximately two person-days using a conventional light microscope. However, given first priority and with several persons collaborating by scoring replicate slides, a result can be available in about 55 h. In practice, with routine service, a physician should expect to receive a dose estimate within three working days of obtaining a blood specimen. The need to examine large numbers of cells (Table ) for estimating doses below Gy requires considerable time and is tedious work. Recent developments in computer-driven fluorescence microscopes now enable automatic and rapid location of metaphases [. Pattern-recognition algorithms to enable automatic scoring of dicentrics has not been achieved [2], but the possibility of specific labeling of centromeres opens up the prospect of accurate and rapid dicentric hunting using fluorescence. The main delay, however, in obtaining the biological estimate of doses arises from the need to incubate the cells to allow them to reach metaphase. The technique of premature chromosome condensation (PCC) may reduce this delay considerably so that a dose estimate could be made within a few hours of receipt of a blood specimen. Polyethylene glycol is used to fuse the human lymphocytes with mitotic rodent cells. This causes the single-stranded Go human chromosomes to condense so that breaks in them can be discerned. At present the technique is not used routinely and has not been calibrated extensively. If it can be determined that obtaining a rapid biological estimate of dose would greatly aid the management of patients, then more research effort should be directed into the PCC technique. This might arise, for example, if it is determined that the optimum time for administration of some hemopoietic growth factors is within the first 48 h after irradiation. Because the PCC method bypasses the need for culturing cells, it also bypasses the selective elimination by interphase apoptosis or mitotic delay of the fraction of irradiated cells failing to reach metaphase. Thus it has been shown with in vitro simulation of partialbody exposure [ 3 that the fraction of aberration-bearing prematurely condensed clusters of chromosomes is the same as the fraction of damaged cells in the blood sample. This very early indication of the presence of spared cells, and by implication spared marrow, would be potentially valuable in deciding on early intervention with growth factors. Another recent development in biological dosimetry is examination of lymphocytes for micronuclei induced by radiation. Micronuclei comprise a mixture of chromosome and/or chromatid fragments and also complete chromosomes that for a variety of reasons fail to separate correctly at mitosis and move into the daughter nuclei. These remain isolated and become encapsulated within a separate envelope of nuclear membrane and may be seen as nuclear-staining inclusions within the cytoplasm. Quantification of the dose response for micronuclei became much more reliable with the introduction of the cytochalasin B blocking method [4]. This permits the micronuclei to be scored in cells that are guaranteed to have completed only one in vitro cell cycle. The images of micronuclei are far simpler than chromosomal aberrations, and scoring of them is therefore faster. It has been proposed that this assay would be especially useful if large numbers of casualties were involved [ 5. However, a disadvantage is that with current techniques the cells need 72 h of incubation before micronucleus preparations can be made.

5 Lloyd 99 Table 3. A comparison of dose estimates and their confidence limits based on scoring 25 metaphases for dicenuics or 5 cytokinesis-blocked cells for micronuclei End point and dose response equation Observation Dose (Gy) 95% confidence limits (Gy) Dicentrics Y =, t.4d t.65d2 Micronuclei Y =.3 t.7d t.87d2 Upper I Prosser et al. [6] have considered the logistic implications of using chromosomal or micronucleus methods for dosimetry in an accident where many persons may have been irradiated. They concluded that during the 24 h when chromosomal preparations are available, but before the micronucleus slides are ready, chromosome analysis on a sample of up to 25 metaphases per patient would provide initial dose estimates (Table 3). Although provisional, these would be accurate enough to confirm the triage group into which each patient had been placed, based on acute symptoms, and to identify confounding symptoms such as vomiting due to hysteria. Reliable identification would also be made of those persons who had received doses likely to require treatment for severe marrow failure. This group would then be given priority for improving the estimation by increasing the number of cells analyzed. Whether this is done by scoring micronuclei or continuing with the chromosomal aberrations depends on the number of samples awaiting analysis and how many skilled microscopists are available. DELAYED BLOOD SAMPLING The dicentric is sometimes referred to as an unstable aberration. This means that its persistence in the peripheral blood is determined by the life span of the blood lymphocytes. Replacement of cells by stem cell division will dilute the frequency of dicentrics in the blood because even if dicentrics are induced in stem cells they will be eliminated at mitosis. This imposes a limit on the time after an accident for which it is still possible to provide a dose estimate by analyzing for dicentrics. The replacement rate of lymph@ cytes is certainly variable between persons. A time of about three years for the half-life of the cells has generally been assumed, although both longer and shorter times have been suggested. Certainly after high exposures, sufficient to cause deterministic reactions to the radiation, there is good evidence to suggest that the replacement rate of lymphocytes is much faster. Another class of aberrations, termed stable translocations, will usually pass successfully through divisions. They should persist in the body for much longer and so may be used for retrospective dosimetry, perhaps even decades after the exposure. Analysis for such translocations is very time-consuming, requiring G-band karyotyping or the more newly developed method of fluorescence in situ hybridization (FISH) [7]. Much research is currently under way to develop FISH translocations as an alternative end point for estimating dose. However, it is envisaged that the method will be used only for analysis long after exposure (> year), and in practice most patients are referred to physicians much sooner than this.

6 2 Chromosomal Analysis to Assess Radiation Dose ESTIMATE WHAT IS THE USE TO A PHYSICIAN OF A BIOLOGICAL OF DOSE? After any serious incident involving radiation there is a requirement for radiological protection professionals to investigate the circumstances. For incidents that might have involved exposure of humans, this would include attempts to assess doses. Biological dosimetry can make a useful contribution to this by providing data independent of the physical monitoring, calculations, or reconstmctions employed by health physicists. The results of inquiries, and particularly the lessons learned, are disseminated within the scientific community and to the public. Such investigations would also be required for medical-legal proceedings that may arise from an accident. However, discussion is confined here to the usefulness of biological dose estimates to the physician who is concerned with the management of patients who have been exposed to radiation. In the event of an accident that is quickly identified, such as that at Chemobyl, patients will be observed over the first few days for prodromal symptoms and erythema; blood will be taken frequently to determine changes in a number of biological parameters. The absence or severity of prodromal symptoms will permit an early triage of casualties and indicate whether further developments might be expected, possibly lifethreatening, during the ensuing weeks or months. Reports from Chemobyl indicate that this preliminary sorting of casualties, which in a sense could be considered as employing a crude biological dosimeter, was very effective and was corroborated when estimates based on cytogenetic assays became available. If suitable personal monitors are worn, which incidentally was not the case at Chemobyl or Goihia, information from these should ideally become known within the first 24 h. Evaluations of doses by cytogenetic analysis available from about the third day (or sooner if done by PCC) will serve as a means of confiiing the initial patient assessment. As discussed above, if there are a large number of patients possibly stretching the resources of the cytogenetics laboratories, then sufficiently accurate dose estimates could be made from a preliminary examination of up to 25 metaphases per patient, with the option of firming up the data later. At this point the cytogenetics should be able to identify anomalous situations of patients exhibiting prodromal-like symptoms that may be due to nonradiological causes such as hysteria or infection. Physicians base their management decisions on the symptoms displayed by their patients rather than on dosimeters. Initial biological dose estimates therefore need not be very precise. Highly irradiated persons will already have been identified and hematological, microbial and organ function monitoring will already be in progress. For those persons needing management for impending bone marrow failure, some assessment of the likelihood of regeneration of the patient s own marrow, perhaps assisted by growth factors, will be necessary. This will require information on the heterogeneity of the exposure, which can be supplemented by further cytogenetic data using the distribution analysis and presence of undamaged cells described earlier. Priority would therefore be given to improving the chromosomal scoring for these patients to provide firmer data for the statistical analysis. For patients irradiated with doses below the threshold for producing a prodromal response, and subsequently confirmed by cytogenetics, physicians would usually observe and treat symptomatically. An important function of the physician is in the counseling of persons who have been involved in a radiation accident. Indeed, for all patients, reassurance and psychological support are essential from an early stage. This is particularly so for those who remain free from symptoms of overexposure during the first week after an accident and are thus unlikely to be at risk of early deterministic effects. These people will seek information on the likely longer term consequences of their exposure. Clearly the demonstration of no or very few chromosomal aberrations in lymphocytes, implying that the dose was minimal at most, is valuable. In addition, persons whose possibility of induced malignancy is thought not to be minor might prefer a quantitative appraisal of their dose. This may then be used as a basis for calculating risk. CONCLUSIONS After a radiation accident there is a need for biological dosimetry to supplement physical methods. Currently, the best available method is that of cytogenetics, and there is a well-developed pool of experience with this assay. The technique provides an estimate of equivalent acute whole-body dose, but this may

7 Lloyd 2 be modified to take account of protracted or inhomogeneous exposure, depending on the circumstances of the accident. For treatment of overexposed persons, a precise biological dose estimate is initially not necessary, since the physician manages the clinical conditions. The cytogenetic technique, using lymphocytes, provides a biological assessment of dose. This can be used indirectly to make an assessment of the health consequences of the exposure, since dose can be related to risk. For high doses, deterministic effects, especially on important stem cell populations such as in the marrow, are of immediate concern. At low doses, the risks of induced stochastic effects need to be considered. ACKNOWLEDGMENTS I am indebted to Drs. A. Guskova and V. U. Nugis of the Institute of Biophysics, Moscow, for access to the original data sheets that enabled the information in Table 2 to be compiled. REFERENCES Bender MA, Gooch PC. Somatic chromosome aberrations induced by human whole-body irradiation: the Recuplex criticality accident. Radiat Res 966;29: IAEA, Biological Dosimetry: Chromosomal Aberration Analysis for Dose Assessment. Technical Report 26, International Atomic Energy Agency, Vienna, 986. Stavem P, Brggger A, Devik F et al. Lethal acute gamma radiation accident at Kjeller, Norway. Acta Radio Oncol 985;24:6-63. Baranov A, Gale RP, Guskova A et al. Bone marrow transplantation after the Chernobyl nuclear accident. N Engl J Med 989;32: Lloyd DC, Edwards AA, Prosser JS. Chromosome aberrations induced in human lymphocytes by in vitro acute X and gamma radiation. Radiat Prot Dosim 986;5: Lloyd DC. The problems of interpreting aberration yields induced by in vivo irradiation of lymphocytes. In: Evans HJ, Lloyd DC, eds. Mutation- Induced Chromosome Damage in Man. Edinburgh: Edinburgh University Press, 978: Lloyd DC, Purrott RJ, Dolphin GW et al. A comparison of physical and cytogenetic estimates of radiation dose in patients treated with iodine-3 for thyroid carcinoma. Int J Radiat Biol 976;3: Littlefield LG, Joiner EE, DuFrain RJ et al. Cytogenetic dose estimates from in vivo samples from persons involved in real or suspected radation exposures. In: Hiibner KF, Fry SA, eds. The Medical Basis for Radiation Accident Preparedness. Amsterdam: Elsevier, 98: Lloyd DC, Edwards AA, Prosser JS et al. Accidental intake of tritiated water: a report of two cases. Radiat Prot Dosim 986; 5: IAEA, The Radiological Accident in Goilnia. International Atomic Energy Agency, Vienna, 988. Fantes JA, Green DK, Hill W et al. Application of automation to the detection of radiation damage using FISH technology. Int J Radiat Biol 995;68: Finnon P, Lloyd DC, Edwards AA. Progress in automatic dicentric hunting. In: Obe G, Natarajan AT, eds. Chromosomal Alterations. Heidelberg: Springer, 994: Lloyd DC, Edwards AA, Natarajan AT et al. Biological Dosimetry Applied to In Vitro Simulated Partial Body Irradiation. Report No. EUR2558. Commission of the European Communities: Luxembourg, Fenech M, Morley AA. Measurement of micronuclei in lymphocytes. Mutat Res 985;47: Almassy Z, Krepinsky AB, Bianco A et al. The present state and perspectives of micronucleus assay in radiation protection. A review. Appl Radiat Isot 987;38: Prosser JS, Lloyd DC, Edwards AA. A comparison of chromosomal and micronuclear methods for radiation accident dosimetry. In Proceedings, 4th International Symposium on Radiation Protection, Malvem, Institute of Physics, Bristol, 989: Gray JW, Pinkel D, Brown JM. Fluorescence in situ hybridization in cancer and radiation biology. Radiat Res 994;37: