Parameterizing dose-response models to estimate relative potency functions directly.

Area of Science:

• Toxicology
• Pharmacometrics
• Biostatistics

Background:

• Comparative toxicity assessments often assume constant relative potency between chemicals, which can lead to inaccurate conclusions.
• Existing methods estimate relative potency functions secondarily after fitting dose-response models, which can be complex.

Purpose of the Study:

• To investigate an alternative approach for characterizing chemical potency by directly estimating relative potency functions.
• To evaluate a priori relative potency models, specifically a power function, for direct estimation using combined dose-response data.

Main Methods:

• Developed and applied a direct estimation method for relative potency models using nonlinear regression.
• Utilized a power function as the relative potency model, ensuring dose-response functions remain within common toxicological model families.
• Illustrated the approach with Hill models and data from U.S. National Toxicology Program (NTP) bioassays.

Main Results:

• The direct estimation approach is straightforward when response limits do not differ or are due to experimental factors.
• This method simplifies the calculation of simultaneous confidence bands for relative potency.
• The power function model effectively integrates dose-response data from test and reference chemicals.

Conclusions:

• Directly estimating relative potency functions offers a valuable alternative to secondary estimation methods in toxicology.
• This approach enhances the accuracy and interpretability of comparative chemical potency assessments.
• The method is broadly applicable across various dose-response models commonly used in toxicological research.