A DISTRIBUTED-PARAMETER PHYSIOLOGICALLY-BASED PHARMACOKINETIC MODEL FOR DERMAL AND INHALATION EXPOSURE TO VOLATILE ORGANIC-COMPOUNDS

Estimates of dermal dose from exposures to toxic chemicals are typical ly derived using models that assume instantaneous establishment of ste ady-state dermal mass flux. However, dermal absorption theory indicate s that this assumption is invalid for short-term exposures to volatile organic chemicals (VOCs). A generalized distributed parameter physiol ogically-based pharmacokinetic model (DP-PBPK), which describes unstea dy state dermal mass flux via a partial differential equation (Fickian diffusion), has been developed for inhalation and dermal absorption o f VOCs. In the present study, the DP-PBPK model has been parameterized for chloroform, and compared with two simpler PBPK models of chlorofo rm. The latter are lumped parameter models, employing ordinary differe ntial equations, that do not account for the dermal absorption time la g associated with the accumulation of permeant chemical in tissue repr esented by permeability coefficients. All three models were evaluated by comparing simulated post-exposure exhaled breath concentration prof iles with measured concentrations following environmental chloroform e xposures. The DP-PBPK model predicted a time-lag in the exhaled breath concentration profile, consistent with the experimental data. The DP- PBPK model also predicted significant volatilization of chloroform, fo r a simulated dermal exposure scenario. The end-exposure dermal dose p redicted by the DP-PBPK model is similar to that predicted by the EPA recommended method for short-term exposures, and is significantly grea ter than the end-exposure dose predicted by the lumped parameter model s. However, the net dermal dose predicted by the DP-PBPK model is subs tantially less than that predicted by the EPA method, due to the post- exposure volatilization predicted by the DP-PBPK model. Moreover, the net dermal dose of chloroform predicted by all three models was nearly the same, even though the lumped parameter models did not predict sub stantial volatilization.