Authors
Gray,G.M., Cohen, J.T., Cunha, G., Hughes, C., McConnell, E.E., Rhomberg, L., Sipes I.G., Mattison, D.

Title:
Weight of the Evidence Evaluation of Low-Dose Reproductive and Developmental Effects of Bisphenol A

Source:
Human and Ecological Risk Assessment, 10: 875-921, 2004

Summary:
Bisphenol A (BPA) was invented in the 1930's during the search for non-steroidal chemicals with estrogen activity for potential therapeutic use. However, only in the 1950s, when chemists linked BPA together to create polycarbonate material, companies began using BPA in plastics production. Today, BPA is one of the top 50 chemicals in production in the United States. BPA-based polycarbonate plastic and other resins are used extensively in some dental sealants to prevent cavities, as a coating in metal cans to prevent the metal from contact with food, as plastic in food containers, baby bottles and water bottles, returnable containers for juice, milk and water, micro-wave ovenware and eating utensils. Estimated human exposure to BPA has ranged from 0.00048 mg/kg/day to 0.009 mg/kg/day. BPA is readily biodegradable under normal conditions in the environment and does not persist or bioaccumulate.

Since the middle of 1990's, great attention focused on three publications which reported that the male pups of pregnant CF-1 mice, exposed to low environmentally relevant concentrations of BPA, were born with permanently enlarged prostates. According to low-dose hypothesis, BPA acts as a synthetic estrogen through an endocrine-modulating mode of action to cause adverse reproductive and development effects. The hypothesis is based on the concept of a "non-monotonic" dose-response relationship, meaning that health effects may occur at low doses while much higher doses result in no effect. Because billions of pounds of BPA are produced each year and used widely in everyday products, these papers initiated a series of studies about adverse reproductive and development effects of low-dose of BPA exposure. There are now more than 80 publications that demonstrate endocrine disturbances in 15 species at concentrations of BPA below levels of 50 mg/kg (Lowest Observed Adverse Effect Level) and 0.05 mg/kg (Reference Dose) previously determined to be safe using well-established toxicological procedures and principles. Although BPA is clearly estrogenic at high dose levels, the data supporting low-dose effects of BPA are still controversial. To summarize the present state of knowledge about exposure to BPA, and to resolve uncertainties from conflicting data and interpretations, the Harvard Center for Risk Analysis (HCRA) expert scientific panel assessed the evidence regarding the existence of low dose BPA effects in laboratory animals and humans. The work was funded by a grant from the American Plastic Council.

According to the published data from in vitro experiments using isolated receptors from a number of species, including human estrogen receptors (ERs), BPA is an ER alpha agonist (binds ER); however, in vivo may have different activity depending on the dose and the tissue response. Many authors who evaluated the estrogenicity of BPA using the uterotrophic assay also reported an increase in the uterine weight in immature or ovariectomized rats exposed to BPA. The uterotrophic effect of BPA depends on several factors, including the route of administration, the animal species and strain used, and the sexual maturity of the test animals in the experiment.

Nineteen scientific published papers with BPA exposure to rat and mice via oral and non-oral administration at doses below 50 mg/kg (LOAEL) and 0.05 mg/kg/day (RfD) were evaluated for potential development and reproductive toxicity. For investigation of the dose-response relationship and reproducibility of findings and effects across studies, the studies were categorized into four groups by following endpoints: Organ weight- cervix, epididymis, ovaries, preputial gland, prostate, seminal vesicles, testes, uterus, vagina; Perinatal characteristics- anogenital distance; Pubertal characteristics- estrus cycle characteristics, time until first estrus, mammary gland maturation rate, mammary gland ductal migration rate, preputial separation date, testes descent date, vaginal opening date; Other endpoints- avoidance behavior, brain anatomy, daily sperm production, litter gender ratio, hormone level, mammary gland percent ducts, ovary histopathology, prostatic acid phosphatase activity, sperm characteristics, sperm production efficiency, testes histopathology, testosterone levels. The reported evidence of BPA low-dose effects were assessed against three key criteria: consistency, generalizability across species to humans subjected to environmentally relevant exposure, and biological plausibility. The authors noted that evaluation of the evidence for consistent results was complicated by methodological differences, including differences in the type of animal studies; the health effects evaluated the route of exposure, and the age at which the animals were exposed. In the comprehensive review of the existing data for low-dose effects of BPA, the panel found no consistent affirmative evidence of low-dose effects for any endpoints. According to the evaluation, in the case of BPA, the low-dose hypothesis was based on small-scale studies using non-validated protocols, and results have not been independently replicated. The claimed low-dose reproductive effects also have not been found in much larger-scale multi-generation reproductive and developmental studies that specifically examined low doses.

The most controversial evidence of a low-dose BPA effect related to studies that investigated the impact of BPA exposure on prostate weight in mice. The authors concluded that the existing scientific data does not support the presence of any association between BPA exposure and prostate weight in mice. Among possible factors contributing to the absence of an association between BPA exposure and prostate weight, Gray et al. discussed the difficulty of measuring prostate weights, the inability of independent laboratories to replicate the positive prostate weight findings, inadequate statistical power used by investigators, and the potential differences in sensitivity to estrogenic effects among the strain types used in different studies. The authors hypothesized that failure to control for potential confounders and for litter effects could be a reason for the reported positive association in some studies. Gray et al. also noted that reported low-dose effects in laboratory animals, due to their inconsistency, cannot be generalized to humans. The panel of scientists concluded that there is no credible evidence that BPA exhibits carcinogenic activity at high doses in animals. The fact that diethylstilbestrol (DES) and BPA do not have the same high dose carcinogenic effects does not support estrogenicity as a biologically plausible mechanism for low-doses of BPA. Finally, the available indirect evidence (the higher estrogen levels in humans during pregnancy compare to rodents) indicates that humans may be less sensitive to possible estrogenic effects from BPA exposure due to pharmacodynamic factors. In summary, the authors advised the replication of existing studies under carefully controlled conditions with BPA pharmacokinetic and pharmacodynamic studies.

A recent study (Hunt et al., 2003) has demonstrated a credible link between low-dose BPA exposure and oocyte aneuploidy in the mouse. It was found that oral BPA exposure to mice at daily exposure levels in the range of 14-72 ng/g body weight with contaminated drinking water (100 and 360 ng/ml of BPA) produced increased rates of aneuploidy in treated mice. The BPA treated animals exhibited a dose-dependent increase in aneuploid oocytes (5.8-10.9%), suggesting that low-dose BPA exposure is correlated with meiotic abnormalities in the mouse oocyte. The molecular mechanism by which BPA exerts its adverse effect on aneuploidy in unknown, however some studies of somatic cells exposed in vitro have indicated BPA-induced disturbances on microtubule organization thereby causing aneuploidy. These kinds of chromosomal abnormalities are leading causes of miscarriage, congenital birth defects and mental retardation in humans. Although no direct conclusions can be drawn on human health effects without further study, these findings provide the link between mammalian aneuploidy and an environmental exposure. The study also suggests that the mouse oocyte may provide a sensitive system for the study of reproductive toxins. The authors conclude that additional studies are required to investigate the possible association between BPA exposure and meiotic nondisjunction in the mouse oocyte, and the mechanism of action of BPA on oocyte aneuploidy, as well as the potential relevance of BPA exposure to human health.

The potential toxicity of BPA has been examined extensively with numerous toxicological studies conducted over more than 40 years. In recent years, a hypothesis has been advanced claiming that exposure to extremely low doses of BPA could cause adverse health effects in humans, including disruption of normal hormonal functions. According to this "low-dose hypothesis", health effects occur at doses far below levels previously determined to be safe using well-established toxicological procedures and principles. The present review of published animal studies was conducted to evaluate the weight of evidence for potential developmental and reproductive toxicity of BPA at doses below the LOAEL of 50 mg/kg day and RfD of 0.05 mg/kg day. The review was prepared in response to concerns about the estrogenicity as a low-dose mechanism of action for BPA.

The scientific panel of experts concluded that at present time there is credible evidence that low doses of BPA can cause effects on specific endpoints in animal experiments. However, studies in rodent models have yielded conflicting results regarding the effects of exposure to BPA that are influenced by the species and by the specific strains examined as well as by the dose, the route of administration, and the time of exposure. Due to the lack of clear dose-response relationship, selection bias, uncontrolled confounding factors which affect the consistency of these results, as well as the inability of other well-established animal studies in several different laboratories to observe low dose effects of BPA and/or replicate the findings in similar experiments, the experts considered that there is no significant scientific evidence to support the low dose hypothesis in the mechanism of BPA as a general or reproducible finding. In addition, the mechanism(s) of action is still unclear (i.e., hormone related or otherwise). As a result, it is not clear how these findings can be related to specific toxicological endpoints that are relevant to human health outcomes. The lack of adequate and well-designed epidemiological studies limits the ability for generalization of the animal effects to health outcomes in humans.

Gray et al., suggest that additional studies will need to understand potential similarities and differences in the response of different species and strains to low-dose exposure, to identify an alternative mechanism of action or substantial differences in pharmacokinetics in humans and to determine the extent to which experimental findings in test animals are applicable to humans. The further experimental work with EDCs should also focus on oral administration study designs as a common route of exposure among humans. Future work should be also aimed to improve the understanding of factors influencing outcomes and variability in reproductive and developmental toxicity studies (housing, diet, species, strains and sub-strains, study design and issues of specific extrapolation). The combination of animal- mechanistic studies with human epidemiologic studies should provide the data needed to protect the public health from endocrine disrupting compounds.

References:
Hunt PA, Koehler KE, Susiarjo M, Hodges CA, Ilagan A, Voigt RC, Thomas S, Thomas BF, Hassold TJ. Bisphenol A exposure causes meiotic aneuploidy in the female mouse. Current Biology. 13(7):546-53; 2003.

An on-line supplement of additional material is available at: http://www.hcra.harvard.edu/pdf/bisphenolA.pdf.