You are here:
Examining Models for the Pharmacokinetics of Perfluorooctanoic Acid
Citation:
WAMBAUGH, J. F., R. W. SETZER, AND H. A. BARTON. Examining Models for the Pharmacokinetics of Perfluorooctanoic Acid. Presented at Society of Toxicology Annual Meeting, Baltimore, MD, March 15 - 19, 2009.
Impact/Purpose:
The maximum likelihood estimates for the saturable resorption model parameters for the Lou et al. (2008) data predict that after 600 hours, 30% remains after a dose of 1 mg/kg while only 5% remains following a 60 mg/kg dose. For the saturable deep tissue transport model the amount predicted for the plasma is similar, but more PFOA is accumulated in deep tissue at high doses: for a dose of 1 mg/kg, only 4% of the dose remains in the deep tissue compartment after 600 hours, while for a 60 mg/kg dose the deep tissue compartment holds 76% of the total dose. Determining the body-burden of PFOA will require biological models for tissues other than plasma. Careful statistical analysis provides a needed crucible for data-driven development of these models.
Description:
Perfluorooctanoate (PFOA) is a man-made surfactant used in a variety of industrial and consumer applications. Because of its wide-spread environmental distribution and stability, PFOA is found in human blood from the general population (Calafat et al., 2007). PFOA displays complicated pharmacokinetics (PK) in that plasma serum concentration indicates a long half life but also rapidly achieves steady-state. Past attempts to address this have included using larger volumes of distributions for larger doses as well as the saturable renal resorption model of Andersen et al. (2006). In the saturable resorption model, PFOA is cleared from the plasma into a kidney filtrate compartment from which it is either excreted or resorbed into the plasma by a process with a Michaelis-Menten form. Thus, the clearance rate for PFOA is only slow at low concentrations and at higher concentrations PFOA is rapidly excreted from the body. Lou et al. (2008) recently examined the saturable renal resorption model for CD1 mice and though the model described the plasma data, the parameter uncertainty was large and the estimated flow through the filtrate compartment was too large to be consistent with a biological interpretation. We have investigated alternative model structures for PFOA PK, including a modified empirical two-compartment model in which transport from the deep tissue back to the plasma is saturable. To discriminate between models we have applied maximum likelihood and Bayesian statistical approaches that, for example, balance the likelihood of model predictions with the flexibility gained by additional parameters.