The use of a physiologically based pharmacokinetic modelling in a “full-chain” exposure assessment framework: A case study on urban and industrial pollution in northern italy

Background and goals: The estimate of the internal dose provided by physiologically based pharmacokinetic (PBPK) modelling is a big step forward in the frame of human health risk assessment (HRA) from contaminating sources. The PBPK model included in the MERLIN-Expo platform was here tested with data collected in a human biomonitoring (HBM) pilot study to check model efficacy in predicting concentrations in human blood and urine of people exposed to a modern solid waste incinerator (SWI). The aim of the study was to investigate if the use of a PBPK model integrated in a computational platform could replace more expensive and invasive pilot studies. Twenty eight subjects living and working within 4 km of the incinerator (exposed) and 21 subjects living and working outside this area (unexposed) were selected among the population recruited in the HBM study. The group of exposed (E) subjects and the group of non-exposed (NE) subjects were comparable for all relevant anthropometric characteristics and exposure parameters except for the exposure to SWI emissions. Three different scenarios were created: an “only diet-scenario” (DS), a “worst case scenario” (WCS) and a “most likely scenario” (MLS). The platform was tested for blood-lead (B-Pb), urinary-lead (U-Pb), urinary-anthracene (U-Ant) and urinary-fluoranthene (U-Flt). Average estimated U-Pb was statistically equal to the measured one (est. 0.411~0.278; meas. 0.398~0.455 µg/L) and estimated vs. measured U-Ant differ by one order of magnitude only (est. 0.018~0.010; meas. 0.537~0.444 ng/L) while for U-Flt and B-Pb, the error was respectively of two and four orders of magnitude. It is likely that the extremely high accuracy in the Pb concentration input values referring to diet led to the very accurate estimate for this chemical in urine, but the higher error in the B-Pb computed value suggests that PBPK model equations cannot entirely capture the dynamics for blood compartments. MERLIN-Expo seems a very promising tool in saving time, energy and money in the screening step of the HRA framework; however, many software validations are still required.