Characterizing dose-response relationships in multiple cancer bioassays

Area of Science:

• Toxicology and Carcinogenesis
• Quantitative Risk Assessment
• Chemical Mutagenesis

Background:

• Regulatory agencies traditionally use linear dose-response models for carcinogenic risk assessment.
• Increasing interest in incorporating mechanistic data (e.g., mutagenicity, metabolism) into risk evaluation.
• Limited research on the relationship between dose-response curve shape and chemical mutagenicity.

Purpose of the Study:

• To investigate the association between the shape of dose-response curves and the mutagenicity of chemical compounds.
• To analyze dose-response curve shapes using data from historical rodent bioassays.
• To inform the use of nonlinear models in carcinogenicity risk assessment.

Main Methods:

• Utilized historical data from National Toxicology Program (NTP) cancer bioassays.
• Fitted extended Weibull dose-response models to the bioassay data.
• Employed repeated-measures analysis to assess the estimated shape of the curves.

Main Results:

• Genotoxic chemicals demonstrated dose-response curves that were, on average, closer to linear than those of nongenotoxic chemicals.
• Both genotoxic and nongenotoxic compounds predominantly exhibited convex dose-response curves.
• The influence of genotoxicity on overall curve shape was found to be relatively small.

Conclusions:

• Genotoxicity is associated with a slight shift towards linearity in dose-response relationships.
• The convexity observed in most dose-response curves suggests a threshold effect for many chemicals.
• Findings support the potential utility of nonlinear models, particularly for nongenotoxic compounds, in risk assessment.