A major uncertainty of many site-specific human health and mammalian risk assessments is the digestive bioavailability of Polycyclic Aromatic Hydrocarbons (PAHs) consumed via the incidental ingestion of contaminated soils.
Relative to the bioavailability of PAHs used in spiked foods used in laboratory toxicity tests, soil-associated PAHs at contaminated sites are generally much less available. For example, laboratory tests that have measured in vivo exposures have noted that PAHs in soil contaminated with solid materials (e.g., soot) or weathered petroleum are many times less bioavailable compared to PAHs that have been freshly spiked with solvent or unweathered petroleum. This reduced bioavailability can be quantitatively accounted for in risk assessments using a Relative Bioavailability (RBA) value. However, laboratory feeding trials with live animals are expensive ($50,000 to $100,000 range or more) and require many weeks of experimental work. Fortunately, less expensive and more rapid chemical availability measures can be used to provide RBA information. This presentation provides an evaluation of two such tools: 1) Physiologically Based Extraction Tests (PBET) that extract PAHs from soil using a simulated digestive matrix; and 2) determination of freely-dissolved PAHs in soil via a commercially-available polyethylene passive sampling device. PBET and passive sampling results were evaluated for soils contaminated with solids-associated PAHs and compared to dose and RBA estimates for benzo(a)pyrene values generated from a laboratory experiment with rodents fed small amounts of the soils. Freely-dissolved PAHs data from the passive sampling measurements indicated that PAHs were approximately 10 to 100 times less available than PAHs spiked with unweathered petroleum and solvent, respectively, confirming the low availability of the solids-associated PAHs in the soils tested. Both chemical measures were positively correlated with benzo(a)pyrene bioavailability measurements obtained from the rodent feeding trial, indicating that empirical models may be able to be developed to relate chemical PAH availability measures to RBA estimates such that PBET and passive sampling data may be used to predict RBA values for use in risk assessments. PBET and passive sampling chemical measurements appear to be cost- and time-effective tools for predicting mammalian digestive PAH availability and improving the accuracy of site-specific PAH risk assessments.