Evaluation of an Anti-thyroid Mode of Action for Thyroid Follicular Cell Adenomas in Female Mice Exposed to Tertiary Butyl Alcohol

Chronic exposure to high concentrations of tertiary butyl alcohol (TBA) in drinking water is associated with, low, but increased (15% vs 3% in controls) thyroid follicular cell (TFC) adenomas in female mice. The U. S. Environmental Protection Agency (EPA) developed linear cancer toxicity criteria on this endpoint. This study evaluates whether TBA, a nongenotoxic alcohol, induces TFC adenomas in female B6C3F1 mice through a nonlinear anti-thyroid mode of action (MOA). Female mice were exposed to TBA either by oral gavage (500 and 1000 mg/kg) for 5 days or via drinking water (5, 10, 20, or 40 mg/mL) for 14 or 28 days. Key events (KEs) within an anti-thyroid MOA framework—including activation of CAR/PXR, induction of phase I and II liver enzymes, changes in serum thyroxine (T4) and thyroid-stimulating hormone (TSH), and histopathological and proliferative alterations in the thyroid—were evaluated. TBA exposure was found to activate CAR/PXR and induce hepatic phase I and II enzymes; however, it did not produce consistent alterations in T4 or changes in TSH levels, nor did it cause histological changes in the thyroid or increased TFC proliferation in these short-term exposure studies. Considered together with the TBA chronic study demonstrating TFC hyperplasia, there is some evidence to support an anti-thyroid MOA. The data, herein demonstrate TBA activates upstream KEs (CAR/PXR signaling, with upregulation of UGT and SULT transcripts, enzymes that can influence T4 serum levels) in the MOA while downstream events (serum TSH changes nor histopathological changes in the thyroid) were not demonstrated in these short-term studies. The inability to unequivocally demonstrate downstream KEs is likely a consequence of the overall weak tumor response following chronic exposure to high-dose levels (~2110 mg/kg-bw/day) of TBA. Notably, the EPA considered the evidence for thyroid adenomas as “suggestive”, a level that would typically not support a dose-response assessment. The MOA analyses herein highlight an important consequence of conducting risk assessments on weak, suggestive tumor data.