Exposure Conditions and Pharmacokinetic Principles: Interpreting Bisphenol A Absorption in the Canine Oral Cavity

Gayrard et al. (2013) reported significant (~ 80%) absorption of bisphenol A (BPA) from solutions applied to the oral cavity of dogs, leading to higher serum BPA (aglycone) concentrations than occurred when BPA was absorbed through the gastro intestinal tract. This finding is consistent with first prin ciples and experience with orally absorbed drugs. The implications for human exposure and health will be clear when experi mental evidence is available regarding the extent and frequency of sublingual absorption in orally exposed humans. Arguments made by Gayrard et al. that “nanogramspermilliliter” serum concen trations of BPA resulting from sublingual absorption are plausible in humans ignore key pharmaco kinetic and exposure data and con flate issues of serum BPA concentrations with serum BPA/BPAG (BPA glucuronide) ratios. For a given dose, bolus dosing yields peak concentrations of a parent compound that are higher than those with non bolus dosing. Peak concentrations in mixed systemic serum following sublingual exposure cannot exceed those following intra veneous (iv) dosing. Therefore, in humans, peak systemic serum concentrations following sublingual expo sure cannot be higher than those resulting from bolus iv administration of a given BPA dose. Scaling by dose from pharmaco kinetic studies of non human primates (Patterson et al. 2013), the human serum concentra tion of aglycone BPA immediately after an iv bolus adminis tra tion of a dose equiva lent to a 95th percentile of aggregate daily U.S. human exposure (0.22 μg/kg body weight; Lakind and Naiman 2010) would be < 0.1 ng/mL, which is below current quanti fication limits (LOQ) using stateoftheart technology. Therefore, peak systemic serum concentrations of aglycone BPA following sublingual exposure/absorption of this dose would also be < LOQ. This result is consis tent with the data of Gayrard et al. (2013) in which dog serum BPA concentrations would be ~ 0.07 ng/mL (i.e., < LOQ) when sampled from a site reflecting systemic expo sure (e.g., leg vein). Therefore, in 95% of the U.S. popu lation, serum aglycone BPA con centrations of nanograms per milliliter are not possible, even with complete sublingual absorption. Several studies in the literature have reported BPA/BPAG ratios in human serum samples higher than those (< 1%) consis tently observed in human and animal oral pharmaco kinetic studies (Vandenberg et al. 2007). Gayrard et al. (2013) introduced a new hypothesis (i.e., sublingual exposure/ absorption) to justify this discrepancy. Previously proffered hypotheses for high BPA and BPA/BPAG ratios in human serum samples included exposures 10,000 times higher (Vandenberg et al. 2007) than those measured in humans (Lakind and Naiman 2010), extensive dermal exposure/absorption (Vandenberg et al. 2013), and sample contami nation (Teeguarden et al. 2011). Only the latter hypothesis is supported by ample evidence that contamination of blood samples is not only common, but difficult to avoid (Ye et al. 2013). Finally, extrapolation of BPA levels from BPAG levels using the measured ratio of < 1% following ingestion of BPA is appro priate under conditions where BPA reaches the systemic blood via gastro intestinal tract absorption [e.g., food (see Teeguarden et al. 2011)]. Similarly, values of approxi mately 10–20% would be valid for serum levels 0–4 hr after iv dosing or complete sub lingual absorption, when sampled from a site representing systemic exposure. J.G.T. has received grants from the U.S. Environmental Protection Agency, the National Institute for Occupational Safety and Health, and the American Chemistry Council. J.W.F. and D.R.D. declare they have no actual or potential competing financial interests.