Nonlinearity of dose-response functions for carcinogenicity

Area of Science:

• Toxicology
• Carcinogenesis Research
• Risk Assessment

Background:

• The National Cancer Institute-National Toxicology Program (NCI-NTP) has conducted extensive chemical bioassays.
• Understanding the shape of dose-response curves is crucial for accurate cancer risk assessment.
• Current regulatory approaches often rely on linear dose-response models.

Purpose of the Study:

• To analyze the shape of carcinogenesis dose-response curves for 315 chemicals.
• To investigate the relationship between dose-response curve shape and short-term mutagenicity/genotoxicity assay results.
• To evaluate the appropriateness of linear versus non-linear models in risk assessment.

Main Methods:

• Analysis of carcinogenesis bioassay data from NCI-NTP for 315 chemicals.
• Examination of tumor site data to determine dose-response curve shapes (linear vs. quadratic).
• Correlation analysis with in vivo short-term mutagenicity and genotoxicity assay results.

Main Results:

• Tumor site data were more frequently consistent with quadratic dose-response curves than linear ones.
• Linear models may overestimate cancer risk compared to quadratic models.
• No clear relationship was found between carcinogenesis dose-response curve shape and short-term assay outcomes.

Conclusions:

• Routine use of linear dose-response models may lead to overestimation of chemical risks.
• The findings challenge the regulatory approach of using linear models for short-term assay-positive carcinogens and sublinear models for assay-negative ones.
• A quadratic or other non-linear approach may be more appropriate for modeling carcinogenesis dose-response relationships.