A Critical Review of Multigeneration Studies

Abstract

The Chemical Manufacturers Association has an ongoing project to summarize information required to evaluate various approaches to assessment of reproductive hazards in the workplace. The information presented in this manuscript was generated as a part of that project and included (1) a search of the published literature to identify, among other items (a) the procedures used in animal toxicological reproduction testing and (b) whether the database identified could be used for determining the number of generations and litters per generation to be required to provide reproductive risk assessment data for use in a multigeneration evaluation; and (2) a review and tabulation of information on scientific rationale, regulatory history and US and international regulatory guidelines for conduct of animal toxicological reproduction testing procedures. The search and review were inclusive only to their limits, and reports may have been omitted from consideration because they were unobtainable or unidentified. Of 817 publications identified as relevant, it was possible to obtain at least 73 rat or mouse multigeneration studies that met the minimal criteria of at least 4 animals per dosage group and a premating dosage period of at least 10 days (female animals) or 60 days (male animals). Of these 73 studies, 35 were “negative” (no effect reported) and 38 were “positive” (some effect reported). Twenty of the positive studies had effects that were more severe or first detected in the second or subsequent generations than in the first generation. These 20 studies were selected for use in identifying appropriate testing procedures because each (1) reported one or more primary adult or litter reproductive effects, (2) had adequate controls, (3) had more than one dosage group, and (4) used the rat as the test species. The chemicals tested in these 20 studies were classified as insecticides (4 studies), herbicides (3 studies), food/food additives (3 studies), fungicides (2 studies), drugs (2 studies), a metal, an “other agent,” a hydrocarbon, a detergent, a plasticizer, and a dioxin (1 study each). These 20 studies were critically reviewed to identify adult and litter primary reproductive effects (present in the absence of adult toxicity) and adult toxic effects. In general, the quality of the published data was poor. Some measurement of adult toxicity was reported in 18 of the 20 studies; only 13 of these 18 studies reported an effect. Each of the 20 studies reported some evaluation of litter parameters; each had some effect. Only 10 of the 20 studies reported an adult primary reproductive effect (9 in the first generation and 1 in the second generation). The affected second-generation adults came from a population of compromised first-generation pups (primary litter reproductive effects: decreased viability and litter weight). In contrast, all 20 studies had a primary litter reproductive effect present in the first generation. Observations for control litters were erratic from litter to litter, although findings from the first litter from generation to generation were essentially reproducible. Three of the 9 studies with “increasing” primary reproductive effects were of agents with bioaccumulative properties. On the basis of this limited sample size (20 studies), it appears that if the objective of the study is to identify the lowest dosage producing a primary reproductive effect, litter parameters are those that are most sensitive, and one generation appears to be sufficient for evaluation. If there is evidence of bioaccumulation of the agent (accumulation of the agent in fat, bone and/or other tissues, which is most likely to occur with lipid-soluble agents and heavy metals), more than one generation may be necessary to determine when the steady state of the agent is attained. Although evaluation of a second litter may provide data that demonstrate the bioaccumulative characteristic of the test substance, such information generally is available from other sources (chronic studies and/or bioavailability data). In the 20 studies reviewed, secondary reproductive/physiological effects were produced only by dosages that first or simultaneously resulted in toxic and/or primary reproductive effects in adults and/or litters. This finding may indicate that such parameters should be included in studies with identified target organs. As these secondary reproductive/physiological parameters appear to be less sensitive parameters than classic toxic and/or primary reproductive parameters, they probably are not appropriate for routine inclusion in the general reproductive screen. Despite multiple study deficiencies and the small number of studies reviewed, it appears that the multigeneration test provides information that demonstrates primary reproductive effects in adults and litters, effects not apparent in other chronic studies. However, conduct of the study should be improved and better data obtained, particularly in terms of effects in the first-generation adults. Additionally, development of alternate, less expensive procedures is indicated.