Screening Level Estimation of Chemical Mixtures Toxicity Using In Silico Models: A JP‑5 Case Study for Environmental Incidents

Environmental incidents can release multiple chemicals
simultaneously,
as seen in the recent JP-5 leak in Hawaii, raising concerns about
human exposure through contaminated drinking water. Assessing the
toxicity of such mixtures is essential, yet experimental data are
often unavailable. This study evaluates the feasibility of applying
publicly available in silico tools, including TTC, QSAR models, and
read-across, to estimate oral points of departure (PODs) for acute,
subchronic, and chronic exposures to JP-5. Applicability domain assessments
indicated broad QSAR model coverage, with OPERA and ToxTree fully
inclusive and VEGA models showing mostly moderate-to-high confidence,
aside from a few additives. Dosimetric adjustments and uncertainty
factors were applied to extrapolate rat PODs to in silico-derived
health guidance values (HGV). Mixture toxicity was estimated by using
the concentration addition model and compared with existing HGV-based
values to derive screening levels. In silico-derived HGVs generally
aligned with available guidance values, and the TTC estimates were
consistently conservative. Potential health effects were also curated
based on available animal and mechanistic studies. Overall, this case
study demonstrates that in silico models can help fill toxicological
data gaps and provide screening-level insights into evaluating chemical
mixture exposure when empirical data are limited.