Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 51
6 Biologic Dosimetry and Biologic Markers THE TERM BIOLOGIC MARKER encompasses many biologic end points. For mutagens or carcinogens, these end points can be characterized according to where they occur in the process leading to an environmentally mediated disease, such as cancer. In order to produce a biologic marker, a harmful agent must interact with cellular macromolecules, including DNA, to induce some chromosome-or DNA-damaging event, such as an aberration, the formation of a micronucleus, a DNA adduct, or a somatic mutation. Biologic markers such as aberrations, micronuclei, DNA adducts, or mutations can be measured in a surrogate tissue or in the specific tissue of interest, that is, the target tissue for the effect. Chromosome- or DNA-damaging events also can be measured, using molecular techniques, in oncogenes and suppressor genes. These end points overlap with manifestations of early frank pathology, such as altered cell structure or function, metaplasia or dysplasia, early in situ carcinoma, and, finally, cancer. A sequential ordering of markers is useful because it can define the stage of pathogenesis by end point. Damage evident in the cancer genes themselves is really more a molecular manifestation of disease than it is a biologic marker. There are three kinds of biologic markers: markers of exposure or dose, markers of effect, and markers of susceptibility. Biologic markers of effect record biologic responses in individuals who have been exposed to a genotoxic agent, but markers of exposure (or dose) do not necessarily indicate effects. Superimposed on this are markers of susceptibility; those that could be used to identify persons who are at increased risk of devel-
OCR for page 52
oping a disease that could be triggered by a given exposure. Included here might be persons whose ability to repair DNA damage is limited. Biologic effects seen after moderate and low doses of ionizing radiation are almost invariably the result of damage to the genetic apparatus. Because of the centrality of the genetic material to the induction of damage, and because it is possible to visualize damage by molecular genetic or cytogenetic means, methods that use markers of effect after exposure to low to moderate doses of radiation have concentrated on somatic genetic effects. MARKERS OF EXPOSURE AND DOSE Cytogenetic Markers A chromosome aberration occurs when cells are irradiated and the chromosomes are broken and can rejoin with time after exposure. The kinetics of induction and repair have been carefully studied. For acute exposure to low linear energy transfer (LET) ionizing radiation there is a linear dose-response relationship for simple terminal deletions. Aberrations that require two independent breaks and the interaction between two chromosomes increase linearly at low doses and as the square of the radiation dose at higher doses. When the dose is from high-LET radiation or if the low-LET radiation is protracted or fractionated, aberrations increase as a linear function of dose. Because relevant calibration curves for aberrations can be obtained using human lymphocytes in vitro, it is possible to use the frequency of aberrations measured in lymphocytes of the exposed individual to estimate radiation dose when actual physical measures of dose to an individual are unavailable. Results of studies of people accidentally exposed to high radiation doses indicate that doses estimated from yields of aberrations coincide well with the measured doses (Gooch and others 1964, Bender and Gooch 1967). Chromosome aberrations induced in Go human lymphocytes have been the system of choice for a biologic dosimeter used to quantify the dose to which an individual has been exposed or to verify or corroborate a suspected exposure for which no physical dose measurements have been available (Dolphin and others 1973, Lloyd and others 1987). These studies used mainly a dicentric aberration, an unstable aberration whose frequency decreases with time after exposure. This frequency depends on cell turnover rate; aberration persistence can be relatively long in nonproliferating cells. The background concentration of dicentric cells in unirradiated persons is low, as little as 1–2 dicentrics per 1,000 cells in T-lymphocytes (Littlefield and others 1990), and there is little variability among individuals, so that small radiation-induced increases can be quan-
OCR for page 53
tified. This system can be used to estimate doses as low as 0.10 Gy (10 rad). Below that, the sample size required for statistically reliable results is so large it is impractical to obtain (Evans and others 1979; Lloyd and others 1980). Classic cytogenetic techniques for estimating a dose can be used not only for measurement of a person's exposure, but also for limited epidemiologic purposes if one keeps in mind the unstable nature of aberrations such as dicentrics and the fact that loss may vary between individuals with time after exposure (Littlefield and others 1990). However, if one wants to estimate the dose to which a population has been exposed, as the dose becomes lower than 0.10 Gy (10 rad) the method becomes labor intensive and impractical. If there is a question about the magnitude of the exposure, such an evaluation can establish upper bounds of exposure and help define the dose to populations. It must be kept in mind that the decay found in this end point makes it useful only for an individual recently exposed to radiation (Wolff 1991). Authors have reported values for the average disappearance half-time of lymphocytes containing dicentric and centric rings ranging from 130 days (Ramalho and Nascimento 1991) to 3 years (Lloyd and others 1980). Evaluation of the frequency of stable chromosome aberrations (those that do not decrease with time) has been made possible by techniques that measure translocations between chromosomes. This is done by evaluating banded chromosome preparations or by using a less accurate but more rapid size-grouping method. Such techniques were useful in measuring aberrations in the survivors of the atomic bombing of Hiroshima and Nagasaki at long times after radiation exposure. Fluorescent in situ hybridization (FISH) can be used to further define stable chromosome aberrations (Pinkel and others 1988). It is a promising cytogenetic method for determining the dose of radiation to an individual, and consequently, to a population, particularly for those receiving protracted exposures or for those exposed a long time ago (Straume and Lucas 1995). The technique uses chromosome-specific fluorescent probes to "paint" specific chromosomes so that exchanges involving that chromosome can be identified rapidly (Lucas and others 1989a; Gray and others 1991). When combined with further development of DNA probes specific for centromeres (Lucas and others 1992a), this technology can now efficiently and accurately detect reciprocal translocations in human cells. In addition, full genomic translocation frequencies can be obtained from FISH translocation frequencies after staining only a small fraction of the genome (Lucas and others 1989b). This finding permits scaling to the full genome from only a few painted chromosomes. Fluorescent in situ hybridization has been used to demonstrate that dicentrics and reciprocal translocations are induced with identical frequency (Straume and Lucas
OCR for page 54
1993; Nakano 1993), making this technique as sensitive as the use of dicentrics. The sensitivity of this method depends on the background level, the ability to score large numbers of cells rapidly, and the ability to detect even small translocations. Theoretically, it could detect lower doses than is possible using conventional cytogenetic methods. Additional measurements of the background number of translocations as a function of age, stability with time after exposure, and the amount of variability from one person to another are needed before the full promise of this assay can be realized. Some limited preliminary information exists about using fluorescent in situ hybridization to determine reciprocal translocations in unexposed persons. This suggests that background frequencies are in the range of 4–8 per 1,000 lymphocytes and that they increase with age, possibly because of their persistence (Lucas and others 1992a, 1994). Another cytogenetic end point that has been used to estimate exposures is the measurement of micronuclei in populations of exposed cells. Micronuclei are formed when cells with broken chromosomes divide and the acentric pieces do not proceed to the poles at anaphase. These fragments, or lagging chromosomes, are not included in the daughter nuclei and they form recognizable, diminutive chromatin bodies or micronuclei in the daughter cells. The evaluation of micronuclei is much easier to perform than is chromosome analysis. It does require that the cells divide after the insult for the expression of micronuclei. To ensure that only dividing cells are scored, cells are treated with cytochalasin B, which blocks cytokinesis and results in binucleated cells. Only the binucleated cells are evaluated for the formation of micronuclei. The problems with this procedure are that at very high doses there is evidence that micronuclei can join to decrease frequency and that there is a high and variable background frequency in human lymphocytes. It has been estimated that the high variability and the high background limit the sensitivity to doses of radiation of 0.3 Gy (30 rad) or more (Prosser and others 1989). The rapid nature of the test, the possibility of automation, and the response to multiple environmental insults do make this useful for identifying markers of exposure for some experimental questions and exposure conditions. Genetic or Molecular Markers One method of detecting exposure uses fluorescent-labeled monoclonal antibodies to detect mutations or losses of the alleles on chromosome 4 that code for glycophorin A, a glycoprotein responsible for the M and N blood types found on nonnucleated erythrocytes (Langlois and others 1993; Grant and Bigbee 1993). Because the gene has no known function, it is now thought to be neutral, and thus free from biases caused
OCR for page 55
by selection. The technique uses flow cytometry to separate the cells, and its use is limited to the half of the population that is heterozygous MN. Erythrocytes that have lost either M or the n allele, and are thus M0 or N0, can be detected, as can cells that are homozygous MM or NN (from somatic recombination). The M0 and N0 variants increase with radiation dose, and this assay could be used as an marker of exposure or dose for individuals. The observation that the grouped data from the survivors of the atomic bombing of Hiroshima and Nagasaki showed a dose-related increase in M0 and N0 variants many years after the exposure indicates that the phenomenon has an extremely long half-life. A positive correlation has been found between increases in M0 variant cells and chromosome aberrations in the survivors many years after the exposure (Kyoizumi and others 1989). Both of these observations suggest that M and N alleles are markers of exposure if the population size is large enough and if the dose is high enough. Because there is no in vitro calibration that can be used to determine what response is expected from exposure to radiation, the assay has not been validated and further work is needed. One of the most extensively studied human mutation systems involves the induction of mutations in the X-linked hypoxanthine phosphoribosyl transferase (hprt) locus in human cells. This locus is involved in the salvage pathway whereby hypoxanthine or guanine is phosphorylated so that it can be incorporated into DNA. Loss of the locus or its gene activity conveys resistance to the purine 6-thioguanine, which kills wild-type cells. Detection of cells that will grow to form colonies in the presence of 6-thioguanine allows analysis of the molecular nature underlying the mutations (Albertini and others 1982; Morley and others 1983). In vitro experiments have demonstrated a linear increase of hprt mutants as a function of dose, which suggests that hprt could be useful for biologic dosimetry. However, data on hprt mutants in survivors of the atomic bombing of Hiroshima and Nagasaki show extreme scatter and decreased yield in people several years after their exposure, suggesting that a loss of mutants could have occurred by negative selection. A further complication exists with the use of such loci that are detected by selection, in that these mutants could represent a cluster that has been expressly selected. This could inflate quantification of the effects of the inducing agent (Nicklas and others 1986). As more is learned about the quantitative relationships of in vivo induction of hprt mutations in humans their use for biodosimetry could be validated. Other gene loci have been tested for utility as biologic markers of dose. Among those is the autosomal HLA-A locus. Wild-type human T-lymphocytes are killed by exposure to complement plus monoclonal antibodies against HLA-A2 or HLA-A3. Mutant cells that do not express the
OCR for page 56
surface antigens survive (Janatipour and others 1988). Mutants can be induced by radiation, but the data that show such effects have been obtained only at high doses. Akiyama and colleagues (1992) studied survivors in Hiroshima and Nagasaki and failed to find a radiation effect at this locus. Because of the loss of mutants with time the technique will not be a useful lifetime dosimeter. The absence of any information regarding the half-life of the mutants and the reasons for their loss with time makes its usefulness as a biologic marker of dose problematic. Lymphocyte T-cell antigen receptor (tcr) assays also have been investigated (Akiyama and others 1992). Tcr alpha and tcr beta chains form a heterodimer involved in the surface expression of CD3 complexes. If a mutation occurs in either of these, the CD3 complex cannot be expressed on the cell surface. Such mutants are detected as CD3- cells as determined by the use of monoclonal CD3 antibodies. When Hiroshima and Nagasaki survivors were tested, no increase in mutants was found over the entire dose range, although female patients with cancer of the reproductive system who received very high doses of radiation showed a radiation-induced increase in mutants at this locus. The mutant frequency declined with a half-life of about 2 years, which might have contributed to the lack of an effect observed in the survivors. It should be noted, however, that the standard cytogenetic study of unstable chromosome dicentrics still shows an increase with dose in the survivors, even though that end point has a half-life of anywhere from 2 to 5 years. It thus appears that in the absence of a known dose-response curve, including effects at low doses, and in the absence of studies indicating why the effect was negative in the survivors, effects measured with this locus are not a promising as are those that use the hprt locus. Attempts also have been made to study mutations at the beta-globin locus. The method used automated techniques to pick up rare red blood cells that contain mutant hemoglobin. Mitigating against the use of betaglobin as a marker of dose for radiation is the fact that the method of assay is extremely cumbersome and expensive, that the mutations observed are very rare, and that the mutations consist of only a single base change—they are true point mutations, which are rarely induced by ionizing radiation. The committee recognizes that if biologic markers of dose are to fulfill their promise more research will be required to validate the response of specific markers to radiation dose, to identify new markers, and to better characterize the limitations and sensitivities of known markers. Combined Biologic-Maker Assays Perhaps the most useful biodosimetry analysis that can be performed
OCR for page 57
on people exposed to low doses of ionizing radiation is to measure multiple end points using several of the various assays simultaneously. This provides complementary information from each of the assays, such as persistence of stem cell effects from the glycophorin A assay, mutation spectra from the hprt assay, and good dose sensitivity from the chromosome aberration assays. In addition, each assay measures a somewhat different kind of genotoxic effect caused by radiation exposure, such as clastogenicity (chromosome aberrations) as compared with loss of enzyme activity (hprt) or loss of allelic expression (glycophorin A). Confounders would be somewhat different for each assay, so the precision of dose reconstruction and the prediction of subsequent health effects from radiation would be more precisely defined. The disadvantages of this approach are its cost and the need for a sufficiently large blood sample from each person in a large population. The committee suggests that when biologic markers are used to assess dose, multiple end points should be measured using various assays simultaneously so that the abilities and value of the assays can be maximized. MARKERS OF EFFECT Occasionally, the same end point can be a marker of both dose and effect, depending on the type of damage and the time when the marker is studied. For example, radiation exposure can produce a well-characterized number of chromosome aberrations per unit of dose, and aberrations are good markers of exposure during the early stages of cell proliferation in tissue. As the cells divide, form hyperplastic nodules, progress to benign neoplasms, and finally form radiation-induced malignancies, the cells lose unstable aberrations but might retain stable chromosome aberrations or other genetic lesions that survive cell division and can be classified as markers of disease or effect. It is essential to conduct studies that will relate markers of dose to markers of effect and explain how both relate to the induction of a specific disease or to the incidence of cancer. There are unique chromosome aberrations that are characteristic of defined cancer types. For example, the Philadelphia chromosome is a biologic marker for chronic myelogenous leukemia. The presence of this marker in bone marrow cells predicts the development of the disease. There are other examples for other cancers, such as the loss of part of chromosome 17 during the induction of colon cancer and the rearrangement in either chromosome 1 or 16 or both during the development of breast cancer. These genetic changes are biologic markers of effect. Other molecular markers suggest increased risk for the development of radiation-induced cancer. For example, individuals who are heterozygous for
OCR for page 58
the Rb gene (retinoblastoma) are at an increased risk for both spontaneous retinoblastoma and radiation-induced cancer of the skeleton. MARKERS OF SUSCEPTIBILITY Heterogeneity with respect to responses to insults like radiation raises concern over potentially susceptible subsets of human populations. Susceptible subpopulations would contain individuals that have different levels of protection from or sensitivity to the genotoxic effects of ionizing radiation. Susceptible groups, such as persons with defective DNA repair (ataxia telangiectasia) or those who are heterozygous for the Rb gene and are sensitive to radiation-induced cancer, are known to exist. Knowledge of the marker responses of such individuals is needed. Differences in response in a population could just as well result from differences in individual susceptibility as differences in individual exposure or dose. It must be asked whether there are specific biologic marker responses that could be identified in such populations. Such differences could lead to confusion in evaluation biologic markers of exposure and effect. More research is needed to identify the responses of susceptible subsets of human populations. MARKERS IN RETROSPECTIVE DOSIMETRY Which biologic markers are most appropriate for use in monitoring a human population? The answer will depend on a study's goals. Markers of exposure would be identified first where exposure assessment is the primary concern, with the proviso that the sensitivity of the marker is adequate to the task at hand. By contrast, marker effects would be studied if an exposure has resulted in identifiable genetic damage. Markers of susceptibility are needed to interpret inter-individual differences in response to radiation or genotoxic chemical exposure and, in some cases, to select those persons at greatest risk. The important questions are whether biologic markers of exposure can be useful in dose reconstruction and, if so, how? For retrospective dose reconstruction, it is generally agreed that markers of exposure are not useful below an acute dose of 0.1 Gy (10 rad), at least with current technology. Some of the problems include the large degree of individual variability in background for individual markers. Also, there is the problem of decay or loss of some markers with time after exposure. This is probably less true for markers in stem cell lines than it is for those in short-lived cells. The greatest difference between the use of biologic markers in future events or accidents and using them to reconstruct accidents that occurred
OCR for page 59
in the past is that the kinetics of measuring schemes can be used to advantage. For example, among the most sensitive markers for radiation exposure is the enumeration of unstable chromosome aberrations. If this marker is measured within a year after an acute exposure there will be little decay and the sensitivity will allow it to serve as a good dosimeter. Of course, chronic exposure will induce an accumulation of unstable chromosomal aberrations, and the combination of new aberrations, with the decay of the older lesions, would confuse the evaluation of the total dose. In a similar manner, the sensitivity of hprt and tcr assays shows decay with time after exposure. The disadvantages associated with unstable markers are avoided when stable markers, such as reciprocal translocations measured by fluorescent in situ hybridization, are used. One area in which the use of biologic markers could be important is in the identification of persons exposed to high doses of radiation. In contrast to concerns about evaluating low doses in large populations, biologic markers can effectively identify persons exposed to radiation levels above 0.1 Gy (10 rad). This is just as valuable a public health finding—to demonstrate whether specific persons did or did not receive a high dose—as is documenting the overall range of doses for epidemiologic purposes. MARKERS IN EPIDEMIOLOGY Biologic markers of effect could be useful as epidemiologic end points, although the validity of specific markers as surrogates or predictors of disease must be demonstrated. Knowledge of the relationships between specific markers and specific disease end points is required if the use of markers is to have epidemiologic value in an exposed populace. There must be more than a simple correlation between a marker and an effect in a population; that is, there must be studies that correlate the marker with the effect (such as cancer) in specific persons. There are human populations, primarily those given therapy for cancers such as Hodgkin's disease, in which the secondary cancer (acute leukemia) rate is 7–8% after 7 years. Later comparison of biologic marker status in individuals who do or do not develop disease could be compared in a nested case-control study that could shed light on the relationship of the marker to the subsequent development of disease and whether the worker might be used in broader studies as a true surrogate for the disease. Therefore, the committee suggests careful consideration be given to creating a cryopreserved repository of tissue, blood, or both, from persons at high risk for cancer after a known exposure to radiation (or other agent). Markers such a hprt have been used in cases of human exposure to chemicals including ethylene oxide, butadiene, and polycyclic aromatic
OCR for page 60
hydrocarbons. In these cases, the markers were correlated either with DNA adducts or with urinary excretion products. The question asked in these cases was not whether the persons had been exposed—this was already known—but whether there was a demonstrable effect of the exposure. Even in these studies, with a demonstrated genotoxic effect, there could be no clear answer about the relationship of the marker and a specific disease end point. SUMMARY AND RECOMMENDATIONS The use of biologic markers for dose reconstruction or epidemiologic studies associated with dose reconstruction involves extensive effort and expense, and many of the techniques for finding markers carry much uncertainty (Albertini and others 1990; Grant and Bigbee 1993; Grant and Jensen 1993). Background variability is a major problem with markers other than dicentrics measured by cytogenetic methods. A promising advance is in the scoring of stable chromosome translocations using fluorescent in situ hybridization. Validation measurements made by fluorescent in situ hybridization have shown that the frequency of reciprocal translocations in whole-body-exposed individuals is constant with time after exposure. The validations have provided dose reconstruction results that are consistent with independent dosimetry methods (Straume and others 1992; Lucas and others 1992b, 1993, 1994). The glycophorin A assay also detects a long-lived change in M and N blood cells that can be related to the dose of radiation and that might be useful as a test for a marker of exposure. The use of multiple assays could help to reduce the uncertainties that arise from interindividual and intraindividual variability. In any case, the most sensitive methods can reliably detect only those markers that indicate acute doses greater than 0.10 Gy (10 rad). At lower doses and dose rates, the use of currently known markers is unlikely to help with dose reconstruction. For additional information in this area, the reader is referred to a recent review of this subject by Mendelsohn and colleagues (Biomarkers and Occupational Health: Progress and Perspectives, 1995). The committee makes the following four recommendations: For biologic markers to be useful in dose reconstruction, research will be necessary to measure the stability of persistent biologic markers, define ''calibration curves" for low to moderate and chronic exposures, determine the frequency of specific markers in unexposed populations,
OCR for page 61
define the sources of inter-individual variability for various markers, and develop better definitions of marker responses after partial-body (external) or specific organ (internal emitter) exposure. New assays should be developed to address the problems with individual variability in background, with identification of differences in individual susceptibility to radiation genotoxicity, and with the lack of sensitivity for quantifying low radiation exposures so that acute doses greater than 0.1 Gy (10 rad) can be reconstructed. Biologic markers of effect should be used as epidemiologic end points. However, until clear connections are established between the marker and the disease, their use could be misleading rather than illuminating. As the utility of biologic markers becomes established and accepted, the committee recommends that the CDC develop procedural strategies for conducting field studies for both specimen collection and laboratory analyses in the event of an acute release of activity.
Representative terms from entire chapter: