Superfund Research Program
PAHs in Humans at Environmental Levels: Pharmacokinetics, Metabolism and Susceptible Individuals
Project Leader: David E. Williams
Grant Number: P42ES016465
Funding Period: 2013-2020
This is the first pharmacokinetic (PK) study of the polycyclic aromatic hydrocarbon (PAH), benzo[a]pyrene (BaP), a known human carcinogen exhibiting multiple toxicities including cardiovascular, developmental, behavioral, neurocognitive, immunological and reproductive. PAHs, from combustion of carbon (coal, diesel, tobacco, wood-stoves, etc.) comprise three of the top ten chemicals of concern (the Agency for Toxic Substances and Disease Registry (ATSDR)) at priority pollutant sites and are main drivers of remediation at numerous Superfund sites. The EPA cancer risk slope factor is 1 (mg/kg-day)-1. BaP is the EPA reference compound for cancer risk assessment with PAH mixtures (Relative Potency Factor, BaP=1). Risk, as determined by EPA, drives remediation goals. Data for risk estimates are from predominantly mouse models. Utilizing UPLC-accelerator mass spectrometry (Lawrence Livermore), the research group determined absorption, metabolism and urinary excretion of [14C]-BaP after dosing human volunteers with 46 ng (5 nCi) to develop a PK model (Cross-Species and Life Stage Comparisons of PAH Dosimetry). The relative potency factor (RPF) approach was tested by co-administration of a complex PAH mixture from smoked salmon (assistance from Chemistry Core and Community Engagement Core) and the impact of genotype on metabolic profile (Hummel et al., 2018). The major findings were that there is little intra-individual variation (three separate dosings up to a year apart), BaP is rapidly absorbed and metabolized following oral dosing, and individuals exhibited distinct metabolite profiles (Madeen et al., 2018). Finally, there were no detectable BaP-DNA adducts in PBMCs (LOD 0.5 x 10^11 adducts/nucleotide) after 48-72 hours. These results can be incorporated into risk assessment models to set remediation targets that are driven by human data.