Tobacco Documents Online

In Press: Reoroductive Toxicolostv CRITERIA FOR IDENTIFYING AND LISTING SUBSTANCES KNOWN TO CAUSE REPRODUCTIVE TOXICITY UNDER CALIFORNIA'S PROPOSITION 65* 1 2 3 4 5 6 7 Donald R. Mattisonl, Peter K. Working2, William F. Blazak3, Claude L. Hughes, Jr.4, Joanne M. Killinger5, David L. Olive6 and K.S. RaoT Division of Reproductive Pharmacology and Toxicology, Department of Obstetrics and Gynecology, Slot 518, University of Arkansas for Medical Sciences, 4301 Markham Street, Little Rock, AR 72205 and Division of Human Risk Assessment, National Center for Toxicological Research, Jefferson, AR, 72079. Department of Pharmacological Sciences, Genentech, Inc., 460 Pt. San Bruno, South San Francisco, CA 94080. Sterling Winthrop Research Institute, 81 Columbia Turnpike, Rensselaer, NY 12144-3493. Reproductive Hormone Lab, Box 3418, Duke University, Durham, NC 27710. Battelle Columbus Laboratories, 505 King Avenue, Room 6214A, Columbus, OH 43201. Department of Gynecology, University of Texas Health Sciences Center, San Antonio, 7703 Floyd Curl Drive, San Antonio, TX 78284. Toxicology Research Laboratory, The Dow Chemical Company, 1803 Building, Midland, MI 48674. * The authors are members of The Expert Committee on Criteria for Listing Reproductive Toxicants convened by the International Life Sciences Institute-Nutrition Foundation. The Expert Committee wishes to extend its appreciation to the ILSI-NF Proposition 65 Task Force, Dr. Catherine St. Hilaire, ENVIRON Corporation and the ILSI Risk Science Institute staff for their assistance in the preparation of this document, in particular, Dr. Carol J. Henry, Dr. R.J. Dutton, Ms. Gretchen Bretsch, Ms. Stephanie D. Carter, and Ms. Elizabeth Jean-Paul. Correspondence and reprint requests should be directed to Dr. Mattison, at the University of Arkansas for Medical Sciences.

For example, substances identified as reproductive toxicants may produce effects that are reversible upon discontinuation of exposure, or that are pertinent only to specifically designated age groups. Provision of this type of information would be very beneficial to exposed individuals who must determine an appropriate response to a warning. Based on the above considerations, the Expert Committee suggests that the state Health and Welfare Agency actively develop education plans for the public and for health care practitioners that will precede and continue throughout the institution of warning requirements. The state should consider developing guidelines for management strategies to assist health care professionals in dealing with these issues. The Expert Committee further recommends that, when listing compounds, the SAP explicitly identify the type or types of toxicity caused. It is also recommended that the SAP, where possible, should attempt to identify the gender affected, the presumed magnitude of the effect (dose-response), the age susceptibility, the window of vulnerability, and reversibility of the effect. The Expert Committee suggests that the SAP request that this information be included in the official list. B. Determinina Acceptable Exposure Levels for Renroductive Toxicants After a substance has been identified as a reproductive toxicant and listed as such, an exposure level that complies with the requirements of the law must be developed. There are two components to the development of the exposure level: (1) identification of the no-observed-effect level, or the no-observed-adverse-effect level (NOEL or NOAEL) and (2) extrapolation of the NOEL (or NOAEL) to an acceptable human exposure level. 1. Selection of a NOEL or NOAEL It is customary to conduct animal experiments at dosage levels exceeding estimated levels of human exposure to increase the likelihood that a weak reproductive toxicant will produce a detectable effect and to compensate for the relatively small numbers of animals used in the test. This results in the necessity for extrapolation

suggestive of such. Confirmation requires a study design (preferably cohort in type) constructed around a prior hypothesis, adjusting meticulously for confounding variables and resulting in a statistical validation of the study hypothesis. Epidemiological studies or clinical fertility trials relating exposure to toxicants or indirect endpoints to direct measures of reproductive function pose unique dilemmas in study design and analysis. Well-constructed randomized clinical trials are preferable, but rare in this context. Prospective cohort trials are preferable also, but careful correction for bias should be carried out. The value of retrospective case control and cohort studies is limited by the choice of the control population. Care must be taken when using concurrent, nonrandomized controls or historical controls to adjust for all known confounding variables. The choice of statistical procedures to be applied to particular study designs can sometimes influence interpretation, especially if the procedures are based on inappropriate biologic models or theories. Simple statements of point estimates of relative risk or odds ratios are meaningless without concomitant calculations of confidence intervals or significance levels. Furthermore, calculations of statistical significance must always be considered in light of the number of comparisons made and the biological plausibility of associations noted. Regarding analysis of fertility trials, it is important to correct for time-dependent variation in follow-up of the population at risk for pregnancy. This may be done with either life table methodology or appropriate modeling techniques. As length of follow-up is never uniform in such studies, simple pregnancy rates are inappropriate for statistical comparisons. Negative studies should always be evaluated in light of the study's power to detect true associations with sufficient statistical confidence. A well-constructed and well-analyzed study of sufficient power demonstrating no association should take evidentiary precedence over a poorly designed study implicating a chemical as a reproductive toxicant. Multiple negative studies of sufficient design and power

endpoints of reproductive toxicity, and 4) consideration of the overall weight-of-evidence for reproductive toxicity. A. Definitions The primary goal of Proposition 65 is to protect the public health by restricting exposures to carcinogens and reproductive toxicants. Efforts to protect human reproductive health would focus on pre- and/or postconception exposures that alter fertility in the couple or fecundity in the male or female and/or that produce subsequent developmental toxicity in the offspring. Because postconception exposures and developmental toxicity have been addressed in an earlier document (16), the Expert Committee focused in this document on preconception exposures altering fertility and fecundity. The Expert Committee began its deliberations by developing a definition of successful human reproduction: The essential components of successful human reproduction include the ability to conceive at the appropriate time in the life cycle of the couple (this includes consideration of the optimum time for conception), the continuation of pregnancy to term, and the formation of a structurally and functionally normal offspring. Under this broad definition, substances that decrease contraceptive efficacy, as well as substances that impair reproductive ability, could be considered to be reproductive toxicants. However, in the Expert Committee's opinion, interference in contraception is beyond the scope of Proposition 65 and is not considered further in this document. Therefore, the Expert Committee has defined a reproductive toxicant as a substance that has been demonstrated to impair successful reproduction.

I. BACKGROUND The International Life Sciences Institute-Nutrition Foundation (ILSI-NF) convened a group of experts in reproductive toxicology to develop criteria for listing substances as reproductive toxicants under the provisions of California's Safe Drinking Water and Toxics Enforcement Act of 1986 ("Proposition 65"). The ILSI-NF Expert Committee on Reproductive Toxicity developed this document in coordination with the ILSI Risk Science Institute to provide additional guidance to the Governor's Scientific Advisory Panel (SAP) as it considers criteria for listing substances "known to cause reproductive toxicity." In developing this document, the Expert Committee focused its efforts on establishing criteria for reproductive toxicity endpoints. A previous document focusing on developmental toxicity was prepared in a similar manner by another expert committee convened by ILSI-NF and submitted to the SAP (16). The Expert Committee on Reproductive Toxicity began its deliberations by reviewing the draft criteria, "Identification of Chemicals as Reproductive Toxicants", proposed by the Reproductive Toxicity Subpanel of the SAP (21) and other guidelines and criteria, most notably the U.S. Environmental Protection Agency's proposed guidelines for assessing male and female reproductive risk (23, 24). Guidelines and criteria developed for making weight-of-evidence determinations of human health risks are bound by the context in which they were developed and the point in time at which they were developed. - The Expert Committee based the criteria recommended in this document on existing criteria and guidelines, making only those modifications that were deemed necessary to meet the unique requirements and objectives of Proposition 65. The Expert Committee would like to emphasize that most modifications were made in response to a fundamental difference between the listing process under Proposition 65 and other "listing" activities aimed at identifying potential toxicants for other purposes. Proposition 65 combines virtually all elements of the risk assessment and risk management process into a single step --the listing

pro,;ess. Thic iz a major difference from approaches used by federal and other state regulatory agencies, which use "listing" activities to identify substances for further evaluation before implementing risk management activities. The Expert Committee cautions that the criteria developed in this document, or any other set of criteria adopted for a particular use, must not be adopted for use in other contexts without a thorough assessment of their relevance. In addition, all criteria, including those developed herein, must be updated periodically to ensure that they reflect current scientific understanding. The criteria recommended by the Expert Committee are meant to provide guidance based on the collective experience of the Committee; they are not meant to be a rigid set of rules. II. GENERAL OBSERVATIONS AND RECOMMENDATIONS In the course of its deliberations, the Expert Committee developed observations and recommendations related to (1) the public health effects of Proposition 65 specifically related to the listing process and (2) approaches for determining acceptable levels of human exposure to potential reproductive toxicants under Proposition 65. A. Public Health Considerations Proposition 65 was adopted by California voters to protect themselves against, and to be informed about, chemicals that cause cancer, birth defects, or other reproductive harm. When the issue is viewed from a public health perspective, the state has a clear obligation to protect and promote the public health in accord with these desires. Reproductive toxicology is an extremely complex and evolving science which focuses on the effects of toxicants on adult reproductive function and development of the offspring. These effects may be produced through alterations in a wide range of processes in either the female or the male. Within each of the processes of normal reproductive function, the various events which may be altered and lead to a toxic response can represent a continuum, in the sense that observed alterations may or may not affect reproductive capacity. Therefore, it is

important that this continuum is recognized in establishing and assessing reproductive toxicity in order to appreciate the complexity of a particular response and better understand its applicability in assessing the risk of reproductive impairment. There are two unique issues of concern to public health that are related to adverse effects on reproduction. First is the concept that alteration of fertility is an adverse effect upon the couple rather than simply individuals. Thus, an adverse outcome is inflicted on some individuals in the absence of exposure; i.e., when one partner in a couple has been exposed to, and affected by, a toxicant, both partners are affected. Second, adverse effects can, by definition, be noted only in those couples actively attempting to conceive. Due to variation in age, marital status, or reproductive desires, the consequences of exposure may not be reflected in impaired conception. The Committee believes that any impairment of reproduction by a substance even in couples not attempting to conceive is an adverse health effect. However, it should be realized that all of the above factors are dynamic rather than static determinants. Thus, risk communication should be made available to all segments of the population regardless of perceived risk. Promotion of the public health includes addressing not only the need for accurate information on risks stemming from exposure, but also the facilitation of care and appropriate responses from both exposed persons and health care providers. From this perspective, it is insufficient to provide information on possible hazards (i.e., "warn") while failing to provide a means to manage the concerns engendered. The Expert Committee believes that providing information on possible hazards can have both beneficial and untoward public health effects. An obvious benefit is that it does provide a warning of a possible hazard of exposure about which individuals may not have been aware. Based on the information provided, individuals may have an opportunity to decide whether or not to limit or avoid exposure based on individual assessment of personal risks and benefits of the exposure. However, insufficient or ineffective communication programs which fail to adequately specify the type of toxicity and the expected consequences of such toxicity could result in unnecessary concern and inappropriate health care utilization.

span, of modification in gamete production and/or interference with the fertilization process and subsequent survival of the conceptus. o Subfertility "Subfertility" is defined as a statistically significant decrease from the fertility pattern of a control population. In comparisons of the test and control populations, adjustments must be made for known confounding variables such as age, anatomic or functional reproductive abnormalities, and exposure to other known reproductive toxicants. Two methods can be used to demonstrate such a statistical deviation: (1) Time to Preanancv (TTP) This is defined as the mean length of time required for pregnancy to occur. TTP may also be designated as its inverse value, referred to as the Cycle Fecundity Rate (CFR). This latter term is similar to a variety of indices including Cycle Fertility Rate and Monthly Fecundity Rate (19). (2) Alteration in the Cumulative Fertility Curve While measures of central tendencies, such as mean time to pregnancy, are satisfactory for continuous or similarly variant effects, more subtle alterations in reproductive ability may fail to alter these values. Analysis of cumulative pregnancy as a function of time, months, menstrual cycles, or cycles attempting conception may allow identification of more subtle alterations in reproductive performance.

exposure and some important genital tract factors (e.g., tubal occlusion and intraperitoneal adhesions). Thus, these endpoints are considered insufficient to list as reproductive toxicants. o Perturbations of Female Sexual Behavior While coital frequency and human sexuality in general are most appropriately considered in the context of the couple, female gender-specific sexual behavior changes are possible reproductive outcomes. However, current means to objectively measure alterations in female sexual behavior and current understanding of the association of behavioral changes with reproductive outcome are not sufficient at the present time to support these endpoints as being sufficient for listing as a reproductive toxicant. 2. Animal Endnoints Reproductive toxicity in animals is defined as adverse effects of chemicals on the adult or maturing organism and includes, but is not limited to, deleterious effects on gonadal structure and function, alterations in fertility (e.g., infertility or subfertility), and impaired gamete function. Although exposure occurs prior to the time of conception, reproductive toxicity may also become evident during fertilization, the embryonic or fetal periods, or even postnatally. Gestational and postnatal deficits have been addressed in a previous document on developmental toxicity (16) and are not further addressed here. Reversibility of an effect was not considered in the categorization of endpoints. A number of chemicals cause reversible reproductive effects in the adult male and female or developing offspring. This concept is important in communication of reproductive hazard warnings. Since the effects are reversible and cease after exposure stops and the chemical is cleared from the system, exposures leading to effects in this category are likely to be of lower risk to human reproduction than those that cause permanent damage. However, because

(3) Standardized Fertility Ratio The Standardized Fertility Ratio (SFR) is a measure of the deficit in birth rates (total or stratum-specific) in an exposed group versus some comparison group. o Reoroductive Loss The endpoint "reproductive loss" is defined in a test population as a statistically significant decrease in the rate of ongoing pregnancies resulting from conception when compared to the outcome of a control population. In comparisons of the test and control populations, adjustment must be made for known confounding or functional reproductive abnormalities and exposure to other known reproductive toxicants. Significant effects may be termed valid regardless of the stage of embryogenesis or fetal development at which the study is terminated, provided that the stage at which the studies ended was uniform throughout the study populations. o Other Endnoints Several couple-specific endpoints centered around sexual behavior provide indirect evidence of reproductive effects, including such endpoints as libido and coital frequency. Additionally, couple-specific functional assessments such as the postcoital test (PCT) for motile sperm provide indirect measures of reproductive function. None , of these couple-specific endpoints providing indirect evidence of reproductive effects have sufficient data available at the present time on the relationship between the endpoint and reproductive success to support a decision to list a substance under Proposition 65.

b. Male-Soecific Endooints Male-specific endpoints are summarized in Table 2 and are described below. o Male Gamete Number The most frequently used clinical tool for evaluation of male reproductive function is quantitation of the number of motile sperm per ejaculate. Thus both sperm number and sperm motility are assessed, and significant decreases in both (i.e., motile sperm) are considered to be sufficient evidence to list a substance as a reproductive toxicant. Although this is presumed to be related to male fecundity (17), graded fertility has not been demonstrated in a human population. The association of decreasing fertility with decreasing numbers of sperm ejaculated has been demonstrated in experimental animals (1). Unfortunately, the threshold number of sperm required to assure fertility is not known and probably varies among couples. The Expert Committee believes that substances that significantly decrease the percentage of progressively motile sperm in the ejaculate are reproductive toxicants and should be listed as such. o Snerm Moroholoav Morphological changes (e.g., a statistically significant reduction in morphologically normal sperm) would provide evidence sufficient to list; however, less severe changes would not be judged to be sufficient to list as reproductive toxicants. o Male Gamete Function There are several assays which measure functional characteristics of the male gamete. This area is also under active investigation and changing rapidly. Those presently available include human sperm/hamster egg

strengthen the conclusion that no adverse effects are associated with the measured exposures to the substance. 2. Animal Studies In the interpretation of data from animal reproductive toxicology studies, the quality of the study, design, conduct, and statistical analyses must be taken into consideration. Studies must be of high quality and designed so that the animals are exposed to the test compound by an appropriate route of administration (i.e., relevant to the human route of exposure). Other routes may be relied upon by taking into consideration pharmacokinetic information. Also, exposures should be at the proper time, and for the proper duration so as to maximize detection of an effect. Details of study design are beyond the scope of this report but can be found elsewhere (4, 23, 24). The study design must include identification of reproductive endpoints suitable for defining an adverse effect (see the following section for a discussion of endpoints). An important consideration is whether the substance is exerting a selective adverse effect on reproductive function. For a substance to be identified as a reproductive hazard, adverse reproductive effects should occur at doses that do not cause other types of toxicity that could interfere with mating ability or frequency, especially other significant systemic toxicity. When reproductive effects are seen in the presence of systemic toxicity, scientific judgment concerning the probability of reproductive toxicity in the absence of other toxicities (and at lower doses) is needed to determine whether an adverse reproductive effect has occurred. Another important consideration in evaluating animal data is a determination of the power of the study, which is the probability that the study will demonstrate a true effect. It is dependent on the sample size, as well as the background incidence and variability of the endpoint(s) examined. The apparent lack of an effect may be due to a true lack of activity or the inability of the study to identify an effect because of small sample size. Conversely, some statistically significant effects may

of results from experimental dosage levels to the normally lower levels of human exposure. An important step in characterizing the dose-response relationship in these studies is to determine the "no-observed-effect level," i.e., the highest exposure level at which no morphological, physiological, or functional modification is detectable under the test conditions. Another widely used concept in toxicology is the "no-observed-adverse-effect level" (NOAEL), i.e., the highest dose level at which no biologically adverse effects occur. In many cases, the NOEL and NOAEL both refer to the same exposure level. The Expert Committee suggests that, when the NOEL and NOAEL differ, the NOAEL, rather than NOEL, provide the basis for establishing permissible levels for human exposure, since it is possible for a substance to have a non-adverse effect at a low dose level and an adverse effect at a higher dose. Depending upon the sensitivity of the endpoint monitored and the test species utilized, different NOAELs can be derived for the same chemical. The Expert Committee recommends that the same endpoint used as a basis for listing the substance be evaluated to determine the NOAEL. If multiple endpoints provide the basis for listing, then the most sensitive one (i.e., the effect occurring at the lowest exposure) should be used. In the selection of the appropriate NOAEL, the study selected should use an exposure route that is relevant to the human exposure whenever possible. However, data from other routes should also be evaluated and taken into consideration, especially if supported by pharmacokinetic information. If sufficient data do not exist to determine the NOAEL for an endpoint, then the lowest-observed-adverse-effect level (LOAEL) and an additional safety factor should be used. If good quantitative human dose-response data exist, the NOAEL should be determined from the human data. It must be emphasized that human studies can be of widely varying quality in both design and analysis. The Expert Committee believes that if it is determined that a compound is a reproductive toxicant based on sufficient human evidence, the NOAEL should be determined from the study that resulted in the lowest NOAEL, if that study was adequately designed, conducted, and analyzed. -6-

arise by chance, especially if a large number of endpoints are analyzed; the use of appropriate historical control data may prevent a false assumption of biological relevance in such cases. The Expert Committee encourages appropriate statistical evaluation of studies and inclusion of biological meaningfulness and relevance in the final interpretation of study results. Negative findings from animal reproductive toxicology studies deserve special scrutiny regarding study design and conduct. Such studies must include sufficient numbers of animals to detect an adverse effect, appropriate dose levels and exposure routes must be used, and the data must be evaluated using appropriate statistical methods. Negative studies should also indicate the power to define an adverse effect or the confidence interval on the null hypothesis. C. Evaluation of Endpoints If positive studies are identified that satisfy all of the above criteria concerning the adequacy of study design, execution, and analysis, the Expert Committee recommends that the endpoints observed be evaluated in terms of their relevance to adverse human reproductive outcome. Only those endpoints derived from adequate studies that are judged to be directly relevant to human reproductive success are considered by the Expert Committee to be sufficient to identify a substance as a reproductive toxicant. Endpoints considered by the Expert Committee to be only suggestive of a reproductive hazard should be interpreted as signs of potential reproductive toxicity, but deemed insufficient to list a substance as a reproductive toxicant. Significant adverse effects elicited by a test agent in the latter endpoints should lead to additional investigations into the potential for this agent to induce adverse reproductive effects. In this section, the Expert Committee has categorized endpoints measured in humans and in laboratory animals according to their relevance to adverse reproductive outcomes. The Committee recognizes that such an evaluation of endpoints has not been undertaken heretofore and emphasizes that the following discussion is based on the collective judgment

of the Committee and reflects the Committee's understanding of the scientific knowledge in this area at this time. As new information is developed, the Committee's analysis should be reviewed and updated. 1. Human Endpoints Specific reproductive endpoints from studies in humans that the Expert Committee considered relevant to the assessment of the potential reproductive hazard of an agent are highlighted in this section. All endpoints have been evaluated in terms of their relation to any of several adverse involuntary reproductive outcomes. Only those endpoints associated with adverse reproductive outcomes are considered to provide a sufficient basis for identifying (and listing) a substance as a reproductive toxicant under Proposition 65. The Expert Committee notes that even when data are available which show that alterations have occurred in two or more endpoints that are only suggestive of reproductive hazard, sufficient evidence does not exist to cause the listing of the substance. The Expert Committee has developed a listing of human endpoints based on couple-specific measures (Table 1), male-specific measures (Table 2), and female-specific measures (Table 3). a. Couole-Snecific Endooints Couple-specific endpoints are summarized in Table 1 and are described below. o Infertility As an endpoint, "infertility" is defined in a test population as the complete absence of reproductive capability. This outcome can be achieved by meeting either of two criteria. The first requires demonstration of an absence of fertility throughout the complete reproductive life span. The second calls for evidence, prior to the completion of the reproductive life

The Expert Committee acknowledges that although the "correctness" of the flexible uncertainty factor approach cannot be objectively assessed, this approach has been adopted by all regulatory and advisory groups that have addressed reproductive toxicity risk assessment, including the U.S. Environmental Protection Agency (EPA), U.S. Food and Drug Administration, the World Health Organization, and the National Academy of Sciences. The appropriate magnitude of an uncertainty factor is influenced by many of the factors described above, all of which should be considered. In this regard, the Expert Committee believes that an approach such as that taken by the EPA in its guidelines for male reproductive toxicity risk assessment (24) is preferable to an across-the-board imposition of a "standard fixed factor" of any magnitude. As described in the EPA document Currently, uncertainty factors are applied to a NOAEL or LOAEL to estimate an exposure level for humans at or below which there should be no adverse reproductive effects (i.e., the reference dose). The total uncertainty factor usually ranges from 10 to 1,000 depending on the number of adjustments needed. Uncertainty factors are used for (1) situations in which the LOAEL must be used because a NOAEL was not established, (2) interspecies extrapolation, and (3) intraspecies adjustment for variable sensitivity among individuals. In addition, adjustments may be appropriate for length of exposure (acute to subchronic) and/or to correct for inadequacy of the NOAEL or LOAEL (including consideration of background variability in the measurements, insensitivity of the endpoint, and protection against effects of more prolonged exposure). III. SCIENTIFIC BASIS FOR THE EXPERT COMMITTEE'S RECOMMENDED CRITERIA Development of adequate criteria requires 1) definition of successful reproduction and other commonly used terms, 2) understanding of adequate test methodologies to detect reproductive toxicity, 3) interpretation and assessment of the importance of the various -8-

suggestive of such. Confirmation requires a study design (preferably cohort in type) constructed around a prior hypothesis, adjusting meticulously for confounding variables and resulting in a statistical validation of the study hypothesis. Epidemiologic studies or clinical fertility trials relating exposure to toxicants or indirect endpoints to direct measures of reproductive function pose unique dilemmas in study design and analysis. Well-constructed randomized clinical trials are preferable, but rare in this context. Prospective cohort trials are preferable also, but careful correction for bias should be carried out. The value of retrospective case control and cohort studies is limited by the choice of the control population. Care must be taken when using concurrent, nonrandomized controls or historical controls to adjust for all known confounding variables. The choice of statistical procedures to be applied to particular study designs can sometimes influence interpretation, especially if the procedures are based on inappropriate biologic models or theories. Simple statements of point estimates of relative risk or odds ratios are meaningless without concomitant calculations of confidence intervals or significance levels. Furthermore, calculations of statistical significance must always be considered in light of the number of comparisons made and the biological plausibility of associations noted. Regarding analysis of fertility trials, it is important to correct for time-dependent variation in follow-up of the population at risk for pregnancy. This may be done with either life table methodology or appropriate modeling techniques. As length of follow-up is never uniform in such studies, simple pregnancy rates are inappropriate for statistical comparisons. Negative studies should always be evaluated in light of the study's power to detect true associations with sufficient statistical confidence. A well-constructed and well-analyzed study of sufficient power demonstrating no association should take evidentiary precedence over a poorly designed study implicating a chemical as a reproductive toxicant. Multiple negative studies of sufficient design and power

If adequate human data are not available, animal studies must be used. If data from several species/strains are available, the most appropriate species should be used in determining the NOAEL. The Expert Committee recommends that the most sensitive species be used in determining the NOAEL, unless there is evidence that data from that species are not relevant to the human. In that case, a more relevant species should be selected. A determination of relevance is based on the effect measured and comparable anatomical, physiological, pharmacological, pharmacokinetic, metabolic, and pharmacodynamic processes for the effect in the test animal and in humans. In summary, the Expert Committee recommends that the NOAEL be used to develop the acceptable human exposure level. The NOAEL should be derived from dose-response data from the most sensitive relevant study, human whenever possible, of adequate design and execution that demonstrates the endpoint(s) that forms the basis for listing. 2. Development of an Accentable Human Exposure Level Proposition 65 mandates the use of a 1,000-fold safety or uncertainty factor to develop human exposure levels for listed substances. The Expert Committee did not consider the impact of the 1,000-fold mandatory "uncertainty factor" for establishing acceptable human exposure levels in its development of listing criteria. The Expert Committee, however, felt that it must comment on the lack of a scientific basis for this fixed approach to establishing acceptable exposure levels for reproductive toxicants. Use of a single, inflexible uncertainty factor as specified in Proposition 65 is not scientifically defensible. No single uncertainty factor is appropriate for all situations. The data available to determine that a substance is "known to cause reproductive toxicity" will vary tremendously in quantity, quality, and type. In addition, the kinds of effects caused, the sensitivity of the species studied, the nature of the dose-response relationship, and the site and mechanism of action should influence the determination of the levels that are acceptable for human exposure. -7-

Endpoints of human reproductive toxicity, which would be adequate to justify listing of a chemical as a reproductive toxicant under Proposition 65, include exposures that reduce the cumulative number of cycles that could result in conception. This could result from exposures that cause a) delay of puberty/disturbance of the pubertal sequence, b) premature menopause, and/or c) increases in cycle length. Exposures causing precocious puberty should certainly be considered an adverse effect due to the psychological, sociological, and general growth compromises that such children suffer. The Committee has not included this endpoint as sufficient to list as a reproductive toxicant because it is not necessarily related to decreased reproductive success. However, the Committee notes that it is an adverse endpoint and should be considered a developmental effect as defined in an earlier ILSI-NF document (16). On the other extreme of the reproductive lifespan, earlier age of menopause within the climacteric interval may not have any impact on lifelong fertility potential; hence the Committee has classified this endpoint as insufficient to list a reproductive toxicant. o Disturbances of Ovulation/Ovarian Cvclicitv Subfertility could result from a chemical exposure if ovulation were adversely affected. Such adverse effects could include anovulation per se, dysfunctional ovarian cycles (with abnormal levels of reproductive hormones) (2, 6, 9, 11, 12, 14, 22), or diminished functional capacity of the released oocyte (to be fertilized, initiate early development, nidate, grow, or differentiate). Strict assessment of ovulation requires detection of conception or observation of an egg outside of the ovary. Since these are not practical screening tests in assessment of ovarian function in humans,

penetration, zona attachment and penetration, and other measures of sperm function. Although it is biologically plausible that alterations in any of these parameters may be associated with decreased fertility in human populations, there is little evidence demonstrating a clear relationship between alterations in these functional parameters and decreased fertility. They are, nevertheless, used frequently as clinical tests. The Expert Committee believes that alterations in these functions are not sufficient to list a substance as a reproductive toxicant at the present time, but that a positive response in any one of them should prompt further investigation. o Sexual Behavior The assessment of sexual behavior among humans (libido or coital frequency) is difficult. Data demonstrating that subtle alterations in libido or coital frequency are associated with alterations in human fertility are not available. Therefore, alterations in these functions are not sufficient to list a substance as a reproductive toxicant. However, they may be considered indicators of the need for further studies of reproductive effects. c. Female-Soecific Endwints Female-specific endpoints of reproductive toxicity which can be reasonably considered to reflect exposure-dependent subfertility include 1) alterations of the reproductive lifespan, 2) disturbances of ovulation/ovarian cyclicity, and 3) compromise of female genital tract function. Female endpoints are summarized in Table 3 and described below. o Alterations of the Reproductive Lifesnan If an exposure shortens the reproductive lifespan of women, then logically the lifetime potential for successful reproduction will be decreased. -18-

IV. RECOMMENDED CRITERIA FOR IDENTIFYING AND LISTING REPRODUCTIVE TOXICANTS A. Definition of "Known to Cause Reoroductive Toxicitv" The Expert Committee developed the following definition of a reproductive toxicant: (1) A substance is "known to cause reproductive toxicity" within the context of Proposition 65 if there is sufficient evidence in humans of a consistent pattern of biologically plausible adverse effect(s) relevant to reproductive success following exposure, and/or (2) there is sufficient evidence in animals of consistent results from adequate studies of adverse reproductive effect(s) that are relevant to human reproductive success, that are not secondary to other significant systemic toxicity, and are biologically plausible. B. Sufficient Evidence in Humans Sufficient human evidence for listing a substance as being "known to cause reproductive toxicity" should consist of: (1) Convincina Evidence of an Association Between an Exoosure to a Substance and a Consistent Pattern of Abnormal Outcomes That Are Indicative of Imoaired Reproductive Success. Such evidence implies the availability of studies from which the weight-of-evidence supports and establishes the substance as a crucial factor within a chain of causal events. The data from such studies should be valid and sufficiently sensitive and specific, and should provide evidence clearly -26-

other indirect means of assessing ovarian cycles must be used. Rudimentary assessment of human ovarian function can be derived by obtaining documentation of menstrual interval/pattern and basal body temperature charts, but these parameters do not provide much information about the nature of any apparent abnormalities that are detected. Laboratory methods for assessment of ovulation primarily depend upon assessment of luteal function since, short of conception, the corpus luteum is the final product of the ovarian cycle. Endometrial biopsy or serum progesterone levels obtained in the mid-to-late luteal phase are reasonable indicators of ovarian function, but the most persuasive assessments of ovarian function are cycle profiles which include ultrasonic monitoring of follicle growth and collapse and serial measurement of a panel of reproductive hormones. No single parameter is so precise that it can be independently taken to be diagnostic. Combinations of parameters and/or study of several cycles with experienced interpretation of these data are the state of the art for assessing ovarian function. Evidence, based on this type of analysis, of a decreased rate of normal ovulation attributable to exposure to a substance would provide sufficient evidence for listing as a reproductive toxicant. o Comnromised Female Genital Tract Function Abnormalities of the structure or function of the human female genital tract which would be adequate to justify listing of a chemical under Proposition 65 would include abnormalities of the endometrium or myometrium that interfere with nidation or continuation of pregnancy (5, 7, 25, 26). Analysis of possible associations of chemical exposure and other female genital tract factors is confounded by a) disagreement among clinicians about the significance of such factors in clinical infertility and b) lack of biological plausibility for a cause-and-effect relationship of chemical

TABLE 2 Male-Specific Measures of Reproductive Toxicity in Humans Endnoints Sufficient to Identify a Reoroductive Hazard' 1. Significant decrease in the number of motile sperm per ejaculate 2. Significant change in sperm morphology Endooints Insufficient by Themselves to Identify a Renroductive Hazard 1. Alterations in male gamete function as measured by sperm/hamster egg penetration, zona attachment and penetration, and other measures of sperm function 2. Minor change in sperm morphology 3. Libido 4. Coital frequency *Assumes that requirements for adequate test design, conduct, and interpretation described in Section IIIB have been met.

Certainly many of the reproductive processes described in the laboratory animal appear to have correlates in the human. Despite the fact that the basic processes of gamete development and transport, fertilization, and implantation are similar, as is overall neuroendocrine control, the paucity of information for both experimental animals and humans impedes the correlation of reproductive effects between species. Therefore, any expected similarity of toxic responses will be based primarily on general assumptions regarding the similarity of biological processes. When reproductive toxic effects are observed in animals exposed to a substance, the relevance to humans should be assessed by taking into consideration the circumstances under which the effects occur, the frequency and severity of the effects, the dose-effect relationship, the biological plausibility, and relevancy to humans. Other types of data can provide supplementary information that increases or decreases the confidence that a substance is a likely human reproductive toxicant. For example, a structural similarity to known reproductive toxicant(s) may strengthen the association between the substance under consideration and reproductive toxicity. Conversely, evidence of an adverse reproductive effect in a single species or multiple species that are known to metabolize and distribute the substance differently from humans would be weighted less if it is known that such differences are responsible for the observed toxicity. In addition, because genetic heterogeneity is known to be important in individual human susceptibility to toxicants (18, 20) and because considerable interspecies and interstrain variation in susceptibility exists among laboratory animal species and strains, results from animal models having similar sensitivity to humans would be weighted more heavily. The Expert Committee emphasizes that a well-conducted study (animal or human) showing an absence of adverse effect on reproduction can provide valuable information for a weight-of-evidence determination. Because negative studies are reported less frequently in the literature, the Expert Committee recommends that the SAP encourage early submission (as soon as possible after the substance appears on the candidate list) of all pertinent information, including unpublished negative studies having sufficient documentation, so that an informed scientific judgment can be made regarding the results. -25-

c. Female Animals Endpoints which may indicate that a substance causes female reproductive toxicity are listed in Table 6. Changes in endpoints which are sufficient to conclude that a female reproductive hazard exists are inhibition of ovulation, inhibition of implantation due to altered uterine histology, delayed puberty, and early reproductive senescence (10, 15). Changes in other female reproductive endpoints which are suggestive of possible female reproductive hazard are alterations in endocrine and uterine patterns which do not inhibit ovulation or implantation and changes in ovarian and/or uterine weight. While not sufficient to indicate that a reproductive hazard exists, changes in these latter endpoints indicate that further study of substances producing these changes is warranted. D. Weiaht-of-Evidence Considerations In developing the definition of "known to cause reproductive toxicity" and the principles/criteria for listing substances, the Expert Committee relied on a "weight-of-evidence" approach that encourages the evaluation of biological plausibility and human relevance. Making determinations of potential human hazard, such as the determination that a substance is "known to cause reproductive toxicity," requires that all information available on that substance be considered. All available appropriately conducted studies (animal and human) should be evaluated to reach determination based on the "preponderance" or "weight" of evidence. Also, because many chemicals are capable of interfering with reproduction if exposure is sufficiently high, for a substance to be identified as a reproductive hazard, adverse reproductive effects should occur at doses that do not cause other types of toxicity that could interfere with mating ability or frequency, especially other significant systemic toxicity, unless humans are likely to be exposed to systemically toxic doses or if it is determined that the reproductive effects are not secondary effects of the systemic toxicity.

a reproductive hazard are presented in Table 4. Endpoints 1 through 8 pertain primarily to indices and indicators of the ability of animals to mate, conceive, or deliver live offspring and, as such, measure overall effects on male and female fertility. These endpoints should be considered collectively when evaluating the results from such studies. The mating index (endpoint 1), which is a measure of libido, was considered a reliable measure in animals and was therefore judged sufficient to list by the Expert Committee. The same measure was not considered reliable in humans. The survival indices (endpoint 9) measure pup survival from birth through postnatal day 21. Body weights and growth of offspring were considered suggestive of a reproductive hazard but insufficient to define a hazard because of the myriad factors, independent of test substance exposure, which may contribute to alterations in these endpoints. b. Male Animals A variety of endpoints potentially indicative of reproductive toxicity in male animals were considered by the Expert Committee (Table 5). For endpoints that are evaluated by histopathological techniques, only severe inhibition of spermatogenesis was considered to be sufficient evidence to conclude that a substance was a reproductive hazard. In contrast, with other endpoints that are more easily quantified, such as testicular spermatid number or the sperm count in the ejaculate, partial alterations were considered sufficient evidence, provided they were both statistically significant and dose dependent. A significant increase in the proportion of sperm with morphological abnormalities was considered sufficient evidence to conclude that a substance was a male reproductive hazard. Endpoints such as reproductive organ weights and hormone profiles, which either are highly variable in humans and laboratory animals or are inconsistent indicators of changes in reproductive potential, were considered to be only suggestive of potential reproductive hazard.

B. Evaluation of Reoroductive Toxicity Studies 1. Human Studies Data that link exposures of human populations to specific adverse reproductive outcomes may be accumulated from two sources: clinical studies and epidemiological investigations. The advantages and disadvantages of each have been thoroughly discussed in the literature. It is clear, however, that these approaches to human risk assessment are complementary. Clinical investigations, including case reports, are especially valuable in formulating hypotheses and, on occasion, offer unique insights into pathogenic mechanisms and clues to etiology, and hence should not be overlooked. However, in general, case reports alone will not provide sufficient evidence to identify a substance as a reproductive toxicant. Epidemiological investigations of human populations may employ a variety of study designs. Each has unique advantages and disadvantages that must be considered when evaluating the results (8). Particularly important issues to consider are the potential sources of bias and confounding factors, as well as the sensitivity and specificity of a given design. The size of the study population should offer sufficient statistical power to detect the relevant outcomes. Sources, types, and quality (including reliability) of data need to be carefully examined, as does the choice of control (comparison) populations. When there is evidence of differences in susceptibility in human populations, this fact should be considered in the design of epidemiological and clinical studies (3). Detailed discussion of the strengths, weaknesses, and study design considerations for epidemiological and clinical studies have been published elsewhere (13) and are not reviewed in detail here. Many epidemiological investigations are designed as case control studies, assessing odds ratios for a wide variety of toxicants. Demonstration of a statistical association between a toxicant and subfertility or reproductive loss via this design should not be construed as a confirmation of a reproductive hazard, but rather merely

TABLE 5 Endpoints Indicative of Reproductive Toxicity in Male Laboratory Animals Endnoints Sufficient to Identify a Male Renroductive Hazard* 1. Disruption of seminiferous epithelium resulting in aspermatogenesis evidenced by the absence of mature spermatids in the lumen 2. Alterations in gonadal function causing decreased testicular spermatid number or decreased sperm count in the epididymis, vas deferens, or ejaculate 3. Decrease in percentage of motile spermatozoa 4. Significant change in sperm morphology Endpoints Insufficient to Identify a Male Reproductive Hazard 1. Partial histological disruption of the seminiferous epithelium which does not result in a decrease in sperm number 2. Alterations in reproductive organ weight (e.g., testes, epididymides, seminal vesicles, or prostate) 3. Altered endocrine profiles of testosterone, luteinizing hormone (LH) or follicle-stimulating hormone (FSH) *Assumes that requirements for adequate test design, conduct, and interpretation described in Section IIIB have been met.

TABLE 3 Female-Specific Measures of Reproductive Toxicity in Humans Endpoints Sufficient to Identify a Renroductive Hazard* 1. Alterations of reproductive lifespan o o Delay of puberty Premature menopause 2. Disturbances in ovulation/ovarian cyclicity o o o Anovulation/oligo-ovulation Dysfunctional cycles Diminished functional capacity of oocyte 3. Compromised genital tract function o Structural abnormalities that interfere with nidation or continuation of pregnancy Endooints Insufficient by Themselves to Identify a Renroductive Hazard 1. Perturbations in sexual behavior 2. Earlier age of menopause without change in climacteric interval 3. Intraperitoneal adhesions 4. Tubal occlusion •Assumes that requirements for adequate test design, conduct, and interpretation described in Section IIIB have been met.

14. Jones, G.S., and K. Pourmand. 1962. An Evaluation o.' Etiologic Factors and Therapy in 555 Private Patients with Primary Infertility. Fcrtil. Steril. 13:398-410. 15. Mastroianni, L., Jr., and C.A. Paulsen, eds. 1986. AQinse. Renroduction. and the Climacteric, chap. 1, New York and London: Plenum Press. 16. Mattison, D.R., J.W. Hanson, D.M. Kochhar, and K.S. Rao. 1989. Criteria for Identifying and Listing Substances Known to Cause Developmental Toxicity Under California's Proposition 65. Renrod. Toxicol. 3:3-12. 17. Meistrich, M.L., and C.C. Brown. 1983. Estimation of the Increased Risk of Human Infertility from Alterations in Semen Characteristics. Fert. Steril. 40:220-230. 18. Nebert, D.W., and S.A., Atlas. 1978. The Ah Locus: Aromatic Hydrocarbon Responsiveness of Mice and Men. Hum. Gen. 1(Suppl.): 149-160. 19. Olive, D. L. 1986. Clinical Fertility Trials: A Methodologic Review. Fert. Steril. 45:157-171. 20. Omenn, G.S., and H.V. Gelboin, eds. 1984. Genetic Variability in Resnonse to Chemical Exoosure. Banbury Report 16. Cold Spring Harbor Laboratory, New York. 21. Scientific Advisory Panel, Reproductive Toxicity Subpanel. 1988. Identification of Chemicals as Renroductive Toxicants: Draft(s) for Comment. State of California Health and Welfare Agency. July 26, 1988; September 21, 1988. 22. Soules, M.R. 1987. Luteal Phase Deficiency: an Under diagnosed and Overtreated Reproductive Endocrine Disorder. Obstet. Gynecol. Clin. N. Am. 14:865-886. 23. USEPA. 1988a. Proposed Guidelines for Assessing Female Reproductive Risk. Fed. Rea. 53:24834-24847. 24. USEPA. 1988b. Proposed Guidelines for Assessing Male Reproductive Risk. Fed. Rea. 53:24850-24869. 25. Wild, R.A. 1986. Endometrial Biopsy in the Infertility Investigation. J. Reorod. Med. 31:954. 26. Witten, B.I., and S.A. Martin. 1985. The Endometrial Biopsy as a Guide to the Management of Luteal Phase Defect. Fertil. Steril. 44:460.

they do affect reproduction, the Expert Committee suggests that substances causing reversible effects be listed under Proposition 65, but recommends that the listing include the information that a reversible hazard exists and describe the pattern of resumption of reproductive function after cessation of exposure. Alterations in reproductive capacity measured in animals may be sufficient to classify an agent as a hazard to reproduction in humans. Alternatively, less conclusive results may indicate the potential of a chemical to interfere with reproductive processes and suggest the need for further investigation. Based on current understanding of the association of individual endpoints with impaired reproductive success, the Expert Committee has classified endpoints measured in animal experiments as those being sufficient evidence to list or insufficient evidence to list. The Expert Committee has developed its listing of endpoints based on a compilation of separate, but related, endpoints measured in breeding studies (Table 4), in male animals (Table 5), and in female animals (Table 6). Decrements in endpoints that can change without having a direct effect on reproductive success were not considered to be an adequate indication of reproductive hazard for a listing for Proposition 65. A case in point is a decrease in testicular weight, which is often, but not always, indicative of compromised testicular function. Unless a decrease in testicular weight is accompanied by depletion of the seminiferous epithelium as measured by histopathological examination, or a reduction in spermatid or sperm numbers, it was not considered an adequate criterion for listing. Such changes are indicative that the chemical should be tested further, preferably in studies which provide more conclusive evidence. The chemical should not bP lsd fPl~ l~yta such data are a..,. voaa..nal+ls .,.. .`. o.., ~. a. Animal Breedin¢ Studies Endpoints from laboratory animal breeding studies which were judged by the Expert Committee to be sufficient or insufficient (only "suggestive") to indicate

TABLE 4 Laboratory Animal Breeding Endpoints Endpoints Sufficient to Identify a Renroductive Hazard* 1. Male (Female) = m r of m 1 f whi h m i w nfirm x 100 Mating Index Number of males (females) used for mating 2. Male Fertility Index Number of males oroducina a nreenant female x 100 Number of males for which mating was confirmed 3. Female Fertility Index e Number of females confirmed nregnant x 100 Number of females for which mating was confirmed 4. Gestation Index Number of females delivering at least one live offspring x 100 Number of females confirmed pregnant 5. Number of Implantations per Pregnant Female 6. Pre- and Post Implantation Losses 7. Litter Size at Birth 8. Live Birth Index e Mean number of live offsorinst per litter x 100 Mean number of offspring per litter 9. Survival Indices = Number live offsprinQ on day 4 x 100 Number live offspring born © Number live offsnrina on day 7 x 100 Number live offspring on day 4 Number live offspring on day 14 x 100 Number live offspring on day 7 = Number live offsorinQ on day 21 x 100 Number live offspring on day 14 10. Reproductive Capacity of Offspring resulting from the mating of animals exposed to a potential reproductive toxicant (as assessed in 1-9 above) Endpoints SuaQestive of a Reproductive Hazard 1. Decreased offspring weights and/or retarded growth *Assumes that requirements for adequate test design, conduct, and interpretation described in Section IIIB have been met.

TABLE 8 Summary of Significant and Supplemental Endpoints in Animals Indicating Reproductive Toxicity Significant Sufficient to List Breedinsr Studies 1. Mating index 1. 2. Fertility index 3. Gestation index 4. Number of implantations 5. Pre- and Post implantation loss 6. Litter size at birth 7. Live birth index 8. Survival indices 9. Reproductive capacity of offspring Male 1. Aspermatogenesis 1. 2. Decreased sperm count 3. Decreased sperm motility 2. 4. Severe changes in sperm morphology 3. Female 1. Estrous cycle disruption 2. Reduction in number of ovarian follicles &/or oocyte 3. Sufficient alterations in uterine histology to prevent implantation 4. Delayed puberty 5. Premature reproductive senescence 1. 2. 3. 4. Supplemental Insufficient to List Breeding_Studies Offspring weights and growth Male Partial disruption of seminiferous epithelium Alteration in reproductive organ weights Altered endocrine profiles Female Alterations in ovarian histopathology Alterations in uterine histopathology Altered endocrine patterns which do not inhibit ovulation Alterations in ovarian or uterine weight

TABLE 1 Couple-Specific Measures of Adverse Reproductive Effects in Humans Endooints Sufficient to Identify a Renroductive Hazard' 1. Infertility o o Absence of fertility throughout entire reproductive lifespan Absolute loss of a known, essential component of the reproductive process 2. Subfertility o Increased time to pregnancy o Alterations in the cumulative fertility curve o Decreased Standardized Fertility Ratio 3. Reproductive Loss o Decrease in rate of continuing pregnancies Endooints Insufficient by Themselves to Identify a Renroductive Hazard 1. Libido 2. Coital frequency 3. Postcoital test 4. Increase in antisperm antibodies 'Assumes that requirements for adequate test design, conduct, and interpretation described in Section IIIB have been met.

TABLE 7 Summary of Significant and Supplemental Endpoints in Humans Indicating Reproductive Toxicity Significant Sufficient to List Supplemental Insufficient to List Counle Counle 1. Infertility 1. Libido 2. Subfertility 2. Coital frequency 3. Reproductive loss 3. Postcoital test 4. Increase in anti-sperm antibodies ~I~ Male 1. Decrease in number of motile sperm 1. Altered gamete function 2. Significant change in sperm 2. Minor change in sperm morphology morphology 3. Libido 4. Coital frequency Female Female 1. Alterations in 1. Perturbations in 2. reproductive lifespan Annovation/oligo-ovulation/ 2. sexual behavior Earlier age of menopause without 3. dysfunctional ovarian cycles Compromised genital tract 3. change in climacteric interval Intraperitoneal adhesions function 4. Tubal occlusion -38-

TABLE 6 Endpoints Indicative of Reproductive Toxicity in Female Laboratory Animals EndQoints Sufficient to Identify a Female Renroductive Hazard* 1. Estrous cycle disruption resulting in anovulation 2. Significant reduction in the number of ovarian follicles or oocytes 3. Altered uterine histology which is sufficiently out of sequence with the estrous cycle to prevent implantation 4. Delayed puberty 5. Premature reproductive senescence Endooints Insufficient by Themselves to Identify a Female Reoroductive Hazard 1. Altered ovarian histology characterized by reduced corpora lutea or increased number of ovarian cysts 2. Altered uterine histology without disruption of implantation 3. Altered endocrine patterns of estrogen, progesterone, luteinizing hormone (LH) follicle-stimulating hormone (FSH), or prolactin which do not inhibit ovulation 4. Alterations in ovarian or uterine weight *Assumes that requirements for adequate test design, conduct, and interpretation described in Section IIIB have been met.

associations and (ii) evidence indicating biological plausibility of mechanism of action. These factors must be consistent with human biologic principles. When data on human outcome are completely lacking, which will be true for many substances, the decision whether or not to list the substance as a reproductive toxicant must be based on data from animal studies. A preponderance of positive or negative data from multiple animal studies is to be appropriately weighed by application of principles presented throughout this document.

species that are predictive of human responses with exposure route(s), level, duration, and frequency that are relevant to human exposure. Data from other routes should be taken into consideration if supported by pharmacokinetic information. Data from replicate studies and multiple independent study types should be consistent and reinforcing. When data are discordant, sufficient additional evidence should be available to reconcile the differences. The resolving power and statistical treatment of the studies should be appropriate. Confidence may be increased by the demonstration of a dose-response relationship (as measured by increased incidence or severity of effect with increasing dose). In all cases, endpoints evaluated in such studies should be predictive of reproductive outcomes. Reproductive endpoints identified as sufficient to list were presented in Tables 4-6 and are summarized in Table 8 along with those endpoints judged to be suggestive but insufficient to list. (2) Evidence of Selective Renroductive Toxicity. The Expert Committee recommends that the dose-response distinction between general adult toxicity and reproductive toxicity be carefully assessed. A substance should be considered a reproductive (male or female) toxicant if it produces its effects at dosage levels lower than those which produce general adult toxicity that is severe enough to interfere with mating ability or frequency. For substances that produce secondary, nonspecific reproductive effects only at or above the dosage levels which produce general toxicity to adult male or female animals, the weight-of-evidence should be substantially reduced in the absence of evidence demonstrating comparable levels of human exposure or additional evidence indicative of reproductive toxicity. (3) Evidence of Bioloaical(y Plausible Effects. Consistency with a hypothesis (even a general one) concerning the possible mechanism of action for a substance is sufficient to satisfy this criterion. Evidence of relevancy to humans is based on (i) consistency among animal studies of patterns of exposure, abnormal outcome, and causal

V. REFERENCES 1. Amann, R.P. 1986. Detection of Alterations in Testicular and Epididymal Function in Laboratory Animals. Environ. Health Persoect. 70:149-158. 2. Apter, D., I. Raisanen, P. Ylostalo, and R. Vihko. 1987. Follicular growth in relation to serum hormonal patterns in adolescent compared with adult menstrual cycles. r il ril. 47:82-88. 3. Bloom, A.D., ed. 1981. Guidelines for Studies of Human Ponulations Exposed to Mutagenic and Renroductive Hazards. March of Dimes/Birth Defects Foundation, White Plains, N.Y. 4. Christian, M. 1986. A Critical Review of Multigeneration Studies. J. Am. Coll. Toxicol. 5:161-180. 5. Cumming, D.C. 1985. The Late Luteal Phase in Infertile Women: Comparison of Simultaneous Endometrial Biopsy and Progesterone Levels. Fertil. Steril. 43:715. 6. Daly, D.C. 1989. Treatment Validation of Ultrasound Defined Abnormal Follicular Dynamics as a Cause of Infertility. Fertil. Steril. 51:51-57. 7. Daly, D.C. 1983. Endometrial Biopsy During Treatment of Luteal Phase Defects is Predictive of Therapeutic Outcome. Fertil. Steril. 40:305. 8. Feinstein, A. 1985. Chemical Eoidemiologv_: The Architecture of Clinical Research. Philadelphia: Saunders. 9. Foster, D.L., F.J. Karsch, D.H. Olster, K.D. Ryan, and S.M. Yellon. 1986. Determinants of Puberty in a Seasonal Breeder. In: Greep, R.O., ed. Recent Progress in Hormone Research, 42:331-384. Orlando, FL: Academic Press. 10. Greep, R.O., ed. 1986. Recent ProAress in Hormone Research. Proceedings of the 1985 Laurentian Hormone Conference, vol. 42, Chaps. 8-9, Orlando, FL: Academic Press. 11. Hartman, C.G. 1932. Studies in the Reproduction of the Monkey Macacus (pithecus) rhesus, with Special Reference to Menstruation and Pregnancy. Contrib. Embrvol. 134: 1-161. 12. Hartman, C.G. 1931. On the Relative Sterility of the Adolescent Organism. Science. 14:226-227. 13. Hennekens, C.H., and J.E. Buring. 1987. E i i l in M i i . Boston: Little, Brown & Company. -30-

linking exposure to individuals with known reproductive impairment. Evidence of a dose-response relationship would further support a causal relationship. The most persuasive evidence for a cause-effect relationship arises when a number of studies, conducted by different investigators at different times using different methodologies in different geographic or cultural settings with different populations, all show similar results. In some cases, data from several studies can be pooled for analysis to test a hypothesis. In all cases, the effects measured must have a demonstrated association with impaired reproductive function. The Expert Committee identified human endpoints that are sufficient to identify a reproductive toxicant, as well as those measures that are not sufficient in and of themselves. These endpoints were presented in Tables 1-3 and are summarized in Table 7. (2) Evidence That a Causal Relationship is Biologically Plausible. As the chain of causality is forged, arguments of biological relevance and plausibility must be constantly kept in mind. Unexplained inconsistencies require caution in interpretation of even well-designed and powerful studies that identify statistically significant associations. Statistical significance does not confer biological validity. This criterion is met by demonstrating that the evidence for the reproductive toxicity of a substance is consistent with a hypothesis (even a general one) for causation. Failure of the data to meet either of the above two criteria signals a need for further investigation before a causal relationship is firmly asserted. C. Sufficient Evidence in Animals Sufficient evidence in animals for listing a substance as being "known to cause reproductive toxicity" should consist of: (1) Consistent Evidence of an Adverse Renroductive Effect That is Relevant to Human Reproduction. Data should be derived from studies of acceptable quality in mammalian -27-