An exact analysis of the multistage model explaining dose-response concavity

Area of Science:

• Toxicology and Carcinogenesis
• Mathematical Modeling
• Biostatistics

Background:

• The multistage (MS) model of carcinogenesis predicts specific dose-response function properties.
• These properties, including convex cumulative hazards, are often not observed empirically for some chemicals.
• The discrepancy arises from approximations in the traditional MS model analysis.

Purpose of the Study:

• To re-evaluate the mechanistic assumptions of the MS model.
• To present an exact probabilistic analysis of a two-stage carcinogenesis model.
• To demonstrate the utility of the exact model in fitting empirical bioassay data.

Main Methods:

• Developed an exact probabilistic analysis for a two-stage MS model.
• Applied the model to scenarios with simultaneous or sequential carcinogen action.
• Utilized least-squares fitting to analyze real bioassay datasets.

Main Results:

• The exact two-stage model explains concave cumulative hazard functions not predicted by the traditional MS model.
• The model successfully fits bioassay data for chemicals like 1,3-butadiene.
• Properties inconsistent with empirical data are resolved using the exact model.

Conclusions:

• The simplifying rare-tumor approximations in the traditional MS model lead to inaccurate predictions.
• An exact two-stage model provides a more accurate representation of carcinogenesis.
• The exact model is recommended for bioassay experiments with tumor rates exceeding 10%.