Document 065vmzRppa2pKJ5VVw1b0L6Qm

< AR226-3379 Development Of A Biologically Based Model To Describe Perfluorooctanoic Acid (PFOA) Kinetics P.M. Hinderliter, and G.W. Jepson The DuPont Company, Haskell Laboratory for Toxicology and Industrial Medicine, Newark, Delaware, USA IIntroduction Perfluorooctanoic acid (PFOA) is a fluoroorganic surfactant used in fluoropolymer processes. . PFOA is reported to be metabolicaily inert. PFOA is abeileved to be relatively biopersistent In humans. The recent removal ot perfluorooctanyi sulfonate (PFOS) from the marketplace has brought attention to other fluoroorganic compounds Uver effects including peroxisome proliferation have been noted in rats. Sexdifferences have been shown in the rat. PFOA is cleared more rapidly in the female rat than in the male ra t Species differences have also been observed with the rat reportedly clearing PFOA more rapidly than the human. Protein binding has been reported in the ra t Target organs and prolans are not yet clearly identified. Preliminary modeling efiorfs are based on available literature data. Perfluorinated Compounds: Steric hindrance by large fluorine C-F bonds forms a spiral structure in the carbon chain and causes a more rigid molecule than in hydrocarbons. Fluorine Is highly electronegative and the carbon backbone is completely covered by the "electron cloud* of fluorine atoms presenting an even charge distribution. C-F bonds are very strong and short La.C i .C l - ITable 1 - Input Parameters MchseflsConslant KM&imoKJ 0.1 IMAFigure 4 - Two-Compartment Results liM Table 2 - Data Gaps Tim(hours) IFigure 5 -P B P K Results IHf Protein Binding IMaterials and Methods All models were constructed using Advanced Continuous Simulation Language (ACSL). The preliminary models were tested against available laboratory and worker exposure data for acceptability. The initial model was a non-physiologica!, 2-compartment model (Figure 2). The lung and dermal compartments are included to accommodate potential exposure via these routes, but the primary focus ot the work is on the oral route of exposure. While the liver is not a site of PFOA metabolism, it Is described as a separate model compartment since it is a reported toxicity target in some animal species and may represent a source of significant protein binding. Physiological and chemical specific data input into the model are shown in Table 1. Blood data compared against model simulations in Figures 4 and 5 were obtained from the literature. WBS Figure 1 - Modeling Pathway IFigure 2 - Two-Compartment Model Body k2 Blood Body lumped into angle storage compartment fFigure 3 - PBPK Model IDiscussion The two-compartment model simulates the general trends in the rat data, it is not satisfactory, however, tor predictive uses. The preliminary PBPK model performs batter than the twocompartment model, but includes many estimated or "lit" parameter values. One function of this investigation was to identify the data gaps in the existing literature. Table 2 lists the relevant areas and their relative importance to the modeling effort. The human data in the literature at this lime are not complete enough for modeling purposes. Most of the human data are taken from monitoring of production workers and thereby lacks any exposure information. In addition, there is no associated 'event' on which to base a lime scale. The reported human blood concentrations of PFOA are generally below 35 ppm. While protein binding has been noted in the literature, there is no intormation available to quantify the binding parameters. A binding term is included in the preliminary PBPK mode! but the values used have no physical significance at this point. Qualitative Binding Determination Determine key tissues from mass balance data and literature. Identify target proteins In key tissues. Determine branched chain effects. Check tor the presence of non-protein (e. g. lipid) Interactions. - Quantitation of Model Parameters Calculate stability, stoichiometry and kinetic parameters of PFOA interaction with applicable proteins and biomoiecules. Determine differences in binding of proteins with linear and branched PFOA g lU Conclusions and Future Direction Classical compartmental modeling was not sufficient to describe PFOA kinetics in mammalian systems. A preliminary rat PBPK model was developed. ' Data gaps were identified and a pathway for further model development presented. Future Directions: Conduct Mass Balance, Adsorption. Distribution and Bimination studies in mate and temale rats. Characterize the rate, extent and type of PFOAbinding in rats and human systems. Develop model descriptions for the human. Include quantitative protein binding terms into rat and human PBPK models. Use model to simulate the plasma kinetics, uptake, distribution and elimination of PFOA in humans. .