Physiologically Based Pharmacokinetic (PBPK) Modeling Reliability in Human Exposure Assessment after a Perfluoroalkyl Substances (PFAS) Contamination Occurred in Northern Italy.

Background and goals: A wide area in the Veneto region (northern Italy) has been faced a perfluoroalkyl substances (PFAS) contamination occurred mainly in groundwater and drinking water. Exposure of the population likely started in the Sixties, had a sure decrease after filters installation in 2013. The Veneto Region is conducting a biomonitoring study to investigate health conditions of the exposed population, collecting hundreds of serum samples. Some Physiologically based Pharmacokinetic (PBPK) models are being tested comparing PFOA and PFOS measured levels with the estimated ones.Methods: Original Loccisano model corrected by EFSA Scientific Panel was tested for this case study. A worst case scenario (WCS) and a best case scenario (BCS) were developed. Both scenarios took into account individual simulations carried out on 25 people (11 men and 14 women), aged 22-39 years, living in an area with both groundwater and drinking water contamination. WCS considered water intake from the private well, company water and bottled water in the same proportion. In the BCS we assumed 1/3 of the water intake from drinking water and 2/3 from bottled water.Results: Estimated average PFOA serum level obtained from the WCS was very near to the measured one for men (E:151.4±68.4;M:108.6±51.0 ng/mL), while was farther but still quite near the measured value for women (E:132.2±108.2;M:30.9±30.1 ng/mL). In the BCS the result is the opposite (men, E:23.9±10.4; women, E:22.1±13.2 ng/mL).Conclusions: Model seems to predict well PFOA serum levels in the exposed population even with high entity of uncertainties. Loccisano model is a multi-compartment PBPK model created specifically for PFAS, but with no modelization of menstruation and pregnancy losses. New simulations are being running to increase statistical significance but we are also looking at other models to take into account gender differences, to explain differences between men and women observed in measured data.