Uncertainty analysis in 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) cancer dose-response for three occupational cohorts

While the U.S. EPA has issued a draft report with a 1% TCDD effective dose (ED₀₁) of 87.9pg/kg/day based on continuous integration of key scientific evidence, a detailed and comprehensive uncertainty analysis has not been well documented. In this study, a new estimate for ED₀₁ was derived based on uncertainty analysis by quantitatively assessing the potential bias arising from the selection of kinetic models, dose-response models and cohorts. The cumulative serum lipid concentration (CSLC) and cumulative body burden (CBB) were reconstructed as dose metrics using a concentration- and age-dependent pharmacokinetic model (CADM), physiologically based pharmacokinetic model (PBPK), and age-dependent half-life model (FV), and the reconstructed dose metrics based on CADM and PBPK were generally higher than those based on the FV model. Three dose-response curves (linear, multiplicative and power) were used to link dose metrics and cancer risk to estimate ED₀₁, and the linear model resulted in the lowest ED₀₁, followed by the power model and multiplicative model, for the same cohort. Meanwhile, ED₀₁ based on the CADM model was the highest, followed by those based on the PBPK model and first-order model. Finally, the ED₀₁ was estimated to be 17.03±7.83pg/kg/day by statistically analyzing the distribution of ED₀₁ values based on various kinetic models, cohorts and dose-response models. The study presented here strengthens the scientific basis for understanding the potential health implications of TCDD exposure.