You are here:
Interindividual Variability Assessments through Benchmark Dose-Response Modeling of Primary Human Bronchial Epithelial-Fibroblast Co-Culture Responses to Acrolein
Citation:
Hickman, E., A. Simmons, M. Wheeler, S. Auerbach, S. McCullough, AND J. Rager. Interindividual Variability Assessments through Benchmark Dose-Response Modeling of Primary Human Bronchial Epithelial-Fibroblast Co-Culture Responses to Acrolein. Society of Toxicology Annual Meeting, Nashville, TN, March 19 - 23, 2023.
Impact/Purpose:
The use of in vitro systems that utilize cells isolated directly from a range of individual human donors (i.e., “primary” cells) provide a novel opportunity to simultaneously increase the physiological and public health relevance of in vitro new approach methodologies. Unfortunately, the utility of incorporating interindividual variability into in vitro testing has not yet been integrated into toxicity testing strategies due to a lack of computational methods and infrastructure. To address this gap, we hypothesized that benchmark dose-response (BMD) modeling could be leveraged to evaluate human interindividual sensitivities to chemical exposures.
Description:
Human health risks are known to significantly vary according to individuals across populations; however, this interindividual variability cannot be captured by traditional in vivo or in vitro studies that utilize inbred animal strains and isogenic cell lines, respectively. The use of in vitro systems that utilize cells isolated directly from a range of individual human donors (i.e., “primary” cells) provide a novel opportunity to simultaneously increase the physiological and public health relevance of in vitro new approach methodologies. Unfortunately, the utility of incorporating interindividual variability into in vitro testing has not yet been integrated into toxicity testing strategies due to a lack of computational methods and infrastructure. To address this gap, we hypothesized that benchmark dose-response (BMD) modeling could be leveraged to evaluate human interindividual sensitivities to chemical exposures. To test this hypothesis, we performed BMD modeling on a dataset containing a broad range of endpoints reflective of in vivo tissue physiology derived from a study that applied an inhalation assay battery, including 6 phenotypic and 11 secreted cytokine/growth factor endpoints, to evaluate the effect of an acute exposure to the ubiquitous reactive volatile organic gas acrolein (0-4 ppm) on primary human bronchial epithelial-fibroblast co-cultures (n=14). We then implemented a novel approach leveraging clustering and data reduction across BMD model curve fit parameters to assess whether model parameters were associated with cell donor demographic variables. This study identified the following: First, we found that benchmark doses varied greatly on a per-donor basis, with BMDs spreading up to 10 orders of magnitude after model averaging approaches (e.g., BMDs for vascular endothelial growth factor A ranged 0.84-7.32 ppm [at benchmark responses of 2 SD]). Second, BMDs were generally lower when analyzing response trends on a per-donor basis, as opposed to aggregating averaged responses across donors; this demonstrated the need to capture and quantify sensitive populations by evaluating individual-specific responses rather than aggregating responses at the population-level. Third, we found that BMD modeling of secreted cytokines and functional culture endpoints yielded similar BMDs, with potential sex-based differences captured through model parameter clustering. This study provides critical information towards the improvement and implementation of in vitro-based methods to capture human response variability in chemical risk assessments. This abstract does not reflect EPA policy.