The application of benchmark dose methodology to data from prenatal developmental toxicity studies

Area of Science:

• Toxicology
• Risk Assessment
• Computational Biology

Background:

• Traditional risk assessment relies on No Observed Adverse Effect Levels (NOAELs), which can be highly variable.
• The Benchmark Dose (BMD) modeling approach offers a potentially more consistent and statistically robust alternative.
• Evaluating BMD performance across diverse developmental toxicity datasets is crucial for its adoption in regulatory science.

Purpose of the Study:

• To apply the Benchmark Dose (BMD) concept to a large set of prenatal-developmental toxicity (DT) datasets.
• To compare BMDs derived from various modeling approaches with established NOAELs.
• To assess the consistency and reliability of BMDs for regulatory risk assessment.

Main Methods:

• Utilized 246 prenatal-developmental toxicity datasets.
• Employed five modeling approaches: two generic and three specific to DT.
• Calculated BMDs for both quantal (litter-based) and continuous (e.g., fetal weight) endpoints.
• Compared BMD values (BMD05) with corresponding NOAELs.

Main Results:

• Quantal BMDs (QBMD05) were, on average, 6-fold lower than NOAELs.
• Continuous BMDs (CBMD05) for fetal weight and proportion affected were similar to NOAELs.
• Different continuous modeling approaches yielded similar CBMD05 values.
• Including litter size improved model fit for DT data.

Conclusions:

• BMDs provide a more consistent basis for risk assessment compared to NOAELs.
• Continuous BMD modeling, particularly for fetal weight, shows promise for regulatory application.
• The study supports the broader use of BMDs in evaluating chemical safety and developmental toxicity.