Dose-dependent incidence of hepatic tumors in adult mice following perinatal exposure to bisphenol A

Abstract

Background: Bisphenol A (BPA) is a high production volume chemical with hormone-like properties that has been implicated as a potential carcinogen. Early-life exposure has been linked to increased risk for precancerous lesions in mammary and prostate glands and the uterus, but no prior study has shown a significant association between BPA exposure and cancer development.

Objective: We explored the effects of BPA exposure during gestation and lactation on adult incidence of hepatic tumors in mice.

Methods: Isogenic mice were perinatally exposed to BPA through maternal diets containing one of four environmentally relevant doses of BPA (0, 50 ng, 50 μg, or 50 mg per kilogram of diet), and we followed approximately one male and one female per litter until they were 10 months of age. Animals were tested for known risk factors for hepatocellular carcinoma, including bacterial and viral infections.

Results: We found dose-dependent incidence of hepatic tumors in 10-month-old BPA-exposed mice. Of the offspring examined, 23% presented with hepatic tumors or preneoplastic lesions. We observed a statistically significant dose-response relationship, with an odds ratio for neoplastic and preneoplastic lesions of 7.23 (95% CI: 3.23, 16.17) for mice exposed to 50 mg BPA/kg diet compared with unexposed controls. Observed early disease onset, absence of bacterial or viral infection, and lack of characteristic sexual dimorphism in tumor incidence support a nonclassical etiology.

Conclusions: To our knowledge, this is the first report of a statistically significant association between BPA exposure and frank tumors in any organ. Our results link early-life exposure to BPA with the development of hepatic tumors in rodents, and have potential implications for human health and disease.

Figure 1

Representative photomicrographs of hematoxylin and eosin–stained hepatic lesions in wild-type (a/a) mice exposed to BPA through the maternal diet. (A) Hepatocellular carcinoma in a female mouse exposed to 50 mg BPA/kg maternal diet. (B) Hepatic adenoma in a male mouse exposed to 50 mg BPA/kg maternal diet; arrows indicate the line of demarcation between the neoplasm and compressed adjacent normal parenchyma. (C) Hyperplastic nodule in a female mouse exposed to 50 ng BPA/kg maternal diet; the arrow indicates a bile duct as part of a portal triad within the lesion, indicating preservation of hepatic architecture. (D) Oval cell hyperplasia (arrowheads) and increased Kupffer cells within sinusoids (Kupffer cell hyperplasia) in a male mouse exposed to 50 ng BPA/kg maternal diet. (E) Degenerative changes, including lipofuscin accumulation (arrow), hepatocellular hypertrophy (arrowhead), and steatosis (asterisks) in a female mouse exposed to 50 ng BPA/kg maternal diet. (F) Multinucleated hepatocytes (arrows) in a male mouse exposed to 50 mg BPA/kg maternal diet. Original magnification x400; bar = 50 μm.

Figure 2

Incidence of hepatic tumors (A) or multinucleated hepatocytes (B) in wild-type (a/a) mice exposed perinatally to BPA through the maternal diet [0 (control), 50 ng, 50 μg, or 50 mg/kg of diet]. Mice exposed perinatally to BPA exhibited both linear and nonmonotonic dose responses in a lesion-specific manner. BPA-exposed mice exhibited (A) a statistically significant trend in hepatic tumors (neoplastic and preneoplastic hepatic lesions combined), and (B) a nonmonotonic trend in multinucleated hepatocytes, although this trend was not statistically significant. Numbers of animals per exposure group by lesion type and sex are given in Table 1. **p for trend < 0.05 by Cochrane-Armitage exact test of trend and logistic regression (see Supplemental Material, Tables S1 and S2).

Figure 3

Dose-dependent incidence of hepatic tumors in wild-type (a/a) mice exposed perinatally to BPA through the maternal diet [0 (control), 50 ng, 50 μg, or 50 mg/kg of diet]. BPA-exposed mice exhibited (A) a statistically significant trend in hepatic adenomas (n = 3/78), (B) an increase in hepatocellular carcinomas (n = 13/78), (C) a statistically significant trend in neoplastic hepatic lesions (n = 16/78), and (D) a statistically significant trend in neoplastic and preneoplastic hepatic lesions (n = 18/78). Numbers of animals per exposure group by lesion type and sex are given in Table 1. For full statistical data, see Supplemental Tables S1 and S2. *p for trend < 0.05 by Cochrane-Armitage exact test of trend for total animals. **p for trend < 0.05 on both Cochrane-Armitage exact test of trend and logistic regression for total animals. ^#p for trend < 0.05 on both Cochrane-Armitage exact test of trend and logistic regression for females only. ^##p for trend < 0.1 for females only. ^†p < 0.05 for OR by logistic regression.

Figure 4

Dose-dependent incidence of proliferative lesions in wild-type (a/a) mice exposed perinatally to BPA through the maternal diet [0 (control), 50 ng, 50 μg, or 50 mg per kilogram of diet]. BPA-exposed mice exhibited (A) a statistically significant trend in oval cell hyperplasia, (B) no clear trend in Kupffer cell hyperplasia, (C) a statistically significant trend in hepatocyte hypertrophy, and (D) no clear trend in multinucleated hepatocytes. Numbers of animals per exposure group by lesion type and sex are given in Table 1. For full statistical data, see Supplemental Material, Tables S1 and S2. *p for trend < 0.05 on both Cochrane-Armitage exact test of trend and logistic regression for total animals. ^##p for trend < 0.1 for females only. ^†p < 0.05 for OR by logistic regression.