Center for Public Health and Environmental Assessment, Office of Research and Development, US Environmental Protection Agency, Durham, North Carolina 27711.
Department of Environmental Sciences and Engineering, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599.
Toxicol Sci. 2020 Oct 1;177(2):377-391. doi: 10.1093/toxsci/kfaa117.
Naphthalene, a volatile organic compound present in moth repellants and petroleum-based fuels, has been shown to induce toxicity in mice and rats during chronic inhalation exposures. Although simpler default methods exist for extrapolating toxicity points of departure from animals to humans, using a physiologically based pharmacokinetic (PBPK) model to perform such extrapolations is generally preferred. Confidence in PBPK models increases when they have been validated using both animal and human in vivo pharmacokinetic (PK) data. A published inhalation PBPK model for naphthalene was previously shown to predict rodent PK data well, so we sought to evaluate this model using human PK data. The most reliable human data available come from a controlled skin exposure study, but the inhalation PBPK model does not include a skin exposure route; therefore, we extended the model by incorporating compartments representing the stratum corneum and the viable epidermis and parameters that determine absorption and rate of transport through the skin. The human data revealed measurable blood concentrations of naphthalene present in the subjects prior to skin exposure, so we also introduced a continuous dose-rate parameter to account for these baseline blood concentration levels. We calibrated the three new parameters in the modified PBPK model using data from the controlled skin exposure study but did not modify values for any other parameters. Model predictions then fell within a factor of 2 of most (96%) of the human PK observations, demonstrating that this model can accurately predict internal doses of naphthalene and is thus a viable tool for use in human health risk assessment.
萘，一种存在于驱虫剂和石油基燃料中的挥发性有机化合物，已被证明在慢性吸入暴露中会导致小鼠和大鼠产生毒性。虽然存在更简单的默认方法来将动物的毒性起点外推到人类，但通常更倾向于使用基于生理学的药代动力学（PBPK）模型来进行这种外推。当 PBPK 模型使用动物和人体体内药代动力学（PK）数据进行验证时，对其的置信度会增加。先前已经证明，一种已发表的萘吸入 PBPK 模型可以很好地预测啮齿动物的 PK 数据，因此我们试图使用人体 PK 数据来评估该模型。目前可用的最可靠的人体数据来自受控皮肤暴露研究，但吸入 PBPK 模型不包括皮肤暴露途径；因此，我们通过纳入代表角质层和有活力的表皮的隔室以及确定通过皮肤吸收和运输速率的参数来扩展模型。人体数据显示，在进行皮肤暴露之前，研究对象的血液中存在可测量的萘浓度，因此我们还引入了一个连续剂量率参数来解释这些基线血液浓度水平。我们使用受控皮肤暴露研究的数据来校准改良的 PBPK 模型中的三个新参数，但没有修改任何其他参数的值。模型预测结果与大多数（96%）人体 PK 观察结果相差不到 2 倍，这表明该模型可以准确预测萘的体内剂量，因此是用于人类健康风险评估的一种可行工具。
