Multistage risk models and the age pattern in familial polyposis coli

Area of Science:

• Oncology
• Cancer Biology
• Mathematical Modeling

Background:

• Multistage risk models are widely used to explain adult-onset cancer patterns.
• These models suggest cancer develops through a sequence of cellular transformation events.
• Previous studies using approximate models indicated fewer stages for Familial Polyposis coli (FPC) related colon cancer.

Purpose of the Study:

• To evaluate the accuracy of multistage cancer models using exact formulations.
• To investigate discrepancies in estimated carcinogenic stages between the general population and FPC gene carriers.
• To assess the utility of population-based multistage models for inferring cellular mechanisms.

Main Methods:

• Fitting approximate and exact multistage risk models to cancer incidence data.
• Comparing estimated numbers of carcinogenic stages across different populations and model types.
• Analyzing age-onset patterns for colon cancer in the general population and FPC gene carriers.

Main Results:

• Approximate models yield inconsistent estimates of carcinogenic stages for FPC vs. general populations.
• Exact multistage models produce stage estimates incompatible with experimental carcinogenesis data for the general population.
• Exact models estimate more stages for FPC cases than for the general population, contradicting prior findings.

Conclusions:

• Commonly formulated multistage models, even when exact, are insufficient for inferring cellular mechanisms from population-level cancer onset data.
• The "inherited-hit" hypothesis for FPC may be valid at the cellular level, but population models fail to capture it accurately.
• Rethinking the application and formulation of multistage models is necessary for understanding cancer development and genetic predispositions.