Related Experiment Videos
Statistical aspects of extrapolation of dichotomous dose-response data
Journal of the National Cancer Institute
|January 1, 1978
Summary
This study proposes a multistage carcinogenesis model to predict cancer risk at low doses from high-dose bioassay data. A graphical method assesses extrapolation uncertainty for diethylstilbestrol (DES) and dimethylnitrosamine (DMN) exposures.
Area of Science:
- Quantitative toxicology
- Cancer research
- Statistical modeling
Background:
- Carcinogenesis is a complex, multistage biological process.
- Extrapolating high-dose animal bioassay data to low-dose human risk is statistically challenging.
- Existing models for cancer risk extrapolation have limitations.
Purpose of the Study:
- To develop a general mathematical model for multistage carcinogenesis.
- To address the statistical problem of extrapolating cancer risk from high to low doses.
- To quantify the uncertainty associated with such extrapolations.
Main Methods:
- Proposed a general multistage carcinogenesis model.
- Utilized statistical likelihood procedures for uncertainty assessment.
- Applied a graphical technique for visualizing extrapolation uncertainty.
Main Results:
- The general multistage model demonstrated similar extrapolation characteristics to commonly used specific models.
- The graphical technique effectively provided a measure of uncertainty in risk extrapolation.
- The model was exemplified using diethylstilbestrol (DES)-induced mammary carcinoma and dimethylnitrosamine (DMN)-induced liver carcinoma data.
Conclusions:
- The proposed multistage carcinogenesis model offers a flexible framework for low-dose risk extrapolation.
- Statistical likelihood and graphical methods are valuable tools for assessing extrapolation uncertainty.
- This approach aids in a more robust estimation of carcinogenic risk at environmentally relevant exposure levels.