Fundamental and Mokular Mechanisms of Mutagenesis
Information requirements and regulatory approaches for heritable genetic risk assessment and risk communication *
Mutation Research 330 (1995) 35-40
Kerry L. Dearfield * U.S. Environmental Protection Agency, Office of Pesticide Programs (75090, 401 M Street, S. W., Washington, DC 20460, USA
Accepted 6 February 1995
With the evolution of genetic toxicology as a scientific discipline and the formation of the Environmental Mutagen Society (EMS), much thought was given to the study of chemicals in the human environment for their mutagenic effects. The Societys goal was to promote scientific investigation and dissemination of information related to genetic toxicology. Subsequently, the concern for chemically induced genetic damage in human germ cells and its potential impact on genetic diseases was detailed in the Committee 17 Report (1975). With new information on the involvement of genetic alterations in disease and on the ramifications of possible effects of exposures to environmental mutagens, it is becoming increasingly necessary to again focus our attention on the assessment of heritable genetic effects. Clearly, strategies for communication of genetic hazard/risk assessments to exposed individuals and to those charged with regulating environmental agents need to be developed.
Keywords: Heritable genetic risk; Risk communication
1. Regulatory approach
The United States regulatory agencies, includ- ing the U.S. Environmental Protection Agency (USEPA) and the U.S. Food and Drug Adminis- tration (USFDA), have recognized the need to
*This manuscript has been reviewed by the Office of Prevention, Pesticides and Toxic Substances, U.S. Environ- mental Protection Agency and approved for publication. Ap- proval does not signify that the contents necessarily reflect the views or policies of the Agency, nor does mention of trade names or commercial products constitute endorsement or recommendation for use.
* Corresponding author. Tel. 703 305 6780; Fax 703 305
screen chemicals for mutagenic hazards (USEPA, 1991; USFDA, 1982) as have similar agencies in other countries. Formal guidelines for assessing mutagenicity risk to germ cells in humans were published by the USEPA in 1986 (USEPA, 1986).
Mutagenicity assays have proliferated and, with considerable data and experience, the field has become competent in identifying the mutagenic activity of chemicals. Certain assays have been well validated and have become incorporated into the standard screening batteries (Dearfield et al., 1991). Results from these screening tests are used by hazard/risk assessors to decide if further testing should be performed with a chemical. Mutagenicity testing has, in fact, become one of
5453. the primary means to identify hazardous chemi-
0027-5107/95/$09.50 0 1995 Elsevier Science B.V. All rights resewed SSDI 0027-5107(95)00034-S
36 KL. Dearjield /Mutation Research 330 (1995) 35-40
cals during the development phase and remove them from further consideration before they en- ter the marketplace. This practice thus mitigates human exposure and contributes significantly to pollution prevention.
Information from mutagenicity screening tests is first used in a qualitative assessment of muta- genic hazard. In the absence of activity, the agent is deemed to be of low concern and further evaluation may not be necessary. However, if mutagenic activity is detected, the information may be used in several ways. The most visible application over the past 20 years has been as part of the weight of evidence for carcinogenicity and the prediction of potential carcinogenicity. While carcinogen identification is important, mu- tagenicity testing is not simply a component of the assessment of carcinogenicity. Results from mutagenicity tests assist the hazard/risk assessor in many ways, including delineating modes of cancer induction (e.g., in discriminating genotoxic versus nongenotoxic carcinogens) and in deciding how to perform high- to low-dose extrapolations. Nevertheless, there are many aspects of muta- genicity testing per se that are equally important and deserve attention.
2. Role in toxicology
Genetic toxicology is extremely useful, crucial in some instances, in the assessment of health effects other than cancer. Unfortunately, perhaps because of its focus on short-term screening tests, it has tended to divorce itself from the main- stream of toxicology. The field of genetic toxicol- ogy has advanced rapidly with new molecular technology and has much to contribute, particu- larly to reproductive, developmental, neurobehav- ioral, and immunotoxicology areas. Clearly, ge- netic toxicology has much to offer in the interpre- tation of the broad range of toxic effects of muta- genic chemicals.
However, there is a danger that genetic toxi- cology could be subsumed into other specialties as, for example, reproductive toxicology. Genetic toxicologys unique focus on DNA enables it to stand independently as a specialty. But genetic
toxicologists must interact at all levels with other toxicologists to derive the best possible assess- ment of potentially adverse health outcomes due to exposures to environmental mutagens.
Why is it important to interact with all of toxicology, and why is the assessment of germ cell effects crucial? Because of the importance of transmissible genetic damage. From a recent ef- fort sponsored by the United Nations Environ- ment Programme (UNEP) and the International Commission for Protection against Environmen- tal Mutagens and Carcinogens (ICPEMC), an extensive list was assembled of examples of hu- man diseases and conditions caused by mutations in germ cells (UNEP, 1992). The list includes various transmissible propensities for cancer (e.g., retinoblastoma), as well as other human disorders such as manic depression, fragile-X syndrome, dyslexia, cystic fibrosis, hemophilia, muscular dys- trophy, multiple sclerosis, and sickle-cell anemia. Thus, genetic toxicology research can play a role in the elucidation of potential mechanistic rela- tionships as well as environmental causes of these human illnesses.
Some individuals have asserted that carcino- genicity risk assessment can be used as a surro- gate to protect against any mutagenic (germ cell) risk. This persuasion argues that genotoxicity test- ing is ancillary to carcinogenicity testing. It is clear that genotoxic events are part of the car- cinogenic process; however, it is equally clear there are many other toxic outcomes in which genotoxic events play a major role. To consider genotoxicity risk solely within the confines of carcinogenicity does not provide a risk assessor with adequate knowledge of a primary informa- tion component that impacts a broad range of adverse health outcomes.
3. Qualitative and quantitative assessments
In the qualitative assessment of mutagenicity data, it is extremely important to consider all relevant data. Available information from other toxicity testing such as chronic, subchronic and acute studies, reproduction and developmental studies, neurotoxicity studies, metabolism and
I2 L. Dearfield / Mutation Research 330 (199.5) 35-40 31
distribution studies, chemical reactivity, and structure-activity relationship (SAR) considera- tions must be incorporated. These data in con- junction with mutagenicity results provide insight into the possible biological consequences of ge- netic damage in somatic cells as well as in germ cells. Evidence for damage to germ cells becomes a crucial piece of information for heritable ge- netic risk considerations.
Some of the tests used in the regulatory scheme for germ cell effects include the dominant lethal assay, cytogenetic and DNA damage and repair assays employing germ cells. Germ cells in trans- genie animal models may be targets for gene mutations following chemical exposures. Valida- tion of these models as germ cell assays should be strongly encouraged. Advances in molecular tech- niques and mathematical modeling also may pro- vide useful insight into the assessment of germ (and somatic) cell effects.
Having obtained evidence that a test agent is mutagenic and can interact with germ cells, fur- ther testing is designed to provide information for a quantitative assessment of heritable risk. Quan- titative methods utilized by the USEPA use data from the specific locus test (either visible or bio- chemical) and/or the heritable translocation test. Other tests may be used for quantitation, but these are most appropriate tests currently used for quantitative risk analysis. If a case can be made for dominant skeletal mutations, or some other test, there is a possibility for its use in the testing scheme. Once test results are available, a quantitative risk assessment may be performed using the Agencys mutagenicity risk assessment guidelines (USEPA, 1986).
Germ cell assays and their interpretation are familiar to most genetic toxicologists. However, quantitative risk analysis is less familiar. Fre- quently, genetic toxicologists have limited experi- ence in performing this type of assessment, and have even less experience in extrapolating the information to potentially adverse human health outcomes (see Czeizel, 1989; Ehling, 1989; So- bels, 1989).
Much of the information about the potential for genetic damage in humans is assessed indi- rectly (extrapolated) via results from in vitro and
in vivo animal studies and in vitro studies with human cells; for example, in the parallelogram concept (Sobels, 1989). In fact, many of the tenets of the parallelogram approach are applied in the regulatory assessment of potentially adverse hu- man health outcomes based upon mutagenicity and other toxicity data.
Once a quantitative risk assessment is called for, how is this accomplished? Because few publi- cations are available on the performance of quan- titative heritable genetic risk assessments, im- proved guidance and principles upon which to base credible assessments of risk for heritable genetic effects are required. In addition, proce- dures for the extrapolation of results of such assessments to adverse human health outcomes are crucially important.
Why have these methodologies not been pur- sued in depth before now; i.e., why has the risk of chemically induced germ cell effects in humans been so difficult to assess? Animal models demonstrate that specific chemicals can cause specific genetic alterations in germ cells. Also, specific genetic alterations are associated with well-defined human diseases and conditions. However, the definite linkage of a chemically induced genetic lesion in parental germ cells to a specific human disease state in offspring has not been demonstrated (i.e., there is no experimental evidence for human germ cell mutagens). Much commitment of genetic toxicologists is tied to the belief that such a linkage exists. Experimental evidence for human germ cell mutagenesis em- anates from demonstrable chemically induced ge- netic lesions in somatic cells; for example, the specific induction of mutations in rus genes lead- ing to human liver tumors in exposed individuals.
Why has it been difficult to assess the poten- tial for human germ cell genetic alterations? The UNEP/ICPEMC review (UNEP, 1992) presents three major reasons: (i) rarity of situations in which enough mutations are induced to be de- tected in an epidemiological study; (ii) rarity of the individual diseases, particularly of marker genes identified with the genetic disease (more of these markers are being detected); and (iii) diffi- culty in identifying and studying suitable popula- tions exposed to substantial levels of mutagens.
38 RL. Dearfield /Mutation Research 330 (1995) 35-40
This last difficulty may be mitigated as cohorts of environmentally exposed populations are being identified, for instance in Eastern Europe and China. The numbers problem is highlighted when studying atomic bomb survivors in Japan where no significant sequelae from probable germ cell effects have been identified.
4. The ethylene oxide example and risk eommuni- cation
An approach to quantitative risk assessment was presented by the USEPA using ethylene ox- ide as an example (Rhomberg et al., 1990). This case study was offered to stimulate thinking about rigorous methods for quantitative risk assessment and to gain an understanding of the information requirements for such assessments. Equally im- portant is the crystallization of thinking about what the quantitative numbers mean in relation to potentially adverse human health outcomes. And finally, the paper points to the need for effective, careful (and compassionate) risk com- munication.
To state the example briefly, a male parent is exposed via inhalation to 0.5 ppm of ethylene oxide (Et01 for 8 h/day, 5 days/week, for 3 weeks. A 3-week exposure period was selected based on the known postmeiotic specific effects of Et0 (i.e., it is the most sensitive time frame for exposure of humans). A mouse inhalation model was used to extrapolate the germ cell effects measured by induction of heritable translocations (Generoso et al., 1990). For the human situation, an induced risk of 2.8 or (rounded to) 3 addi- tional offspring with heritable translocations per 10000 live births is predicted (see Rhomberg et al., 1990, for details). This figure is in addition to the background of 19 heritable translocations per 10000 live births for humans, estimated by an ICPEMC working group (Lyon et al., 1983).
What does this increased number of three induced heritable translocations over a back- ground of 19 per 10000 mean in terms of genetic risk, or more pointedly, in terms of potentially adverse human outcomes? Indeed, the genetic event (heritable translocation, gene mutation,
etc.) in and of itself is not an adverse outcome. This is unlike the case of cancer, where the tumor is itself the adverse outcome. The genotypic event must be phenotypically expressed to cause the adverse outcome; thus another layer of complex- ity in the risk assessment for mutagenicity must be considered. At one extreme, the three addi- tional heritable translocations may not be mani- fest as an identifiable adverse outcome. Also, the question must be raised as to whether these three additional events would be detectable within the variability of the background of 19? Does the level of exposure (0.5 ppm Et01 that causes this increase raise concern, or not? The field of ge- netic toxicology has not carefully considered how to communicate such information.
Suppose the Et0 exposure was raised to 100 ppm; the projected heritable translocation rate is now double the background rate. Is a doubling of the background a level at which one should have real concern and communicate that concern to an exposed population? These are difficult questions and should not be taken lightly. They should be answered as honestly as possible within the con- straints of...