Heterogeneity in Multistage Carcinogenesis and Mixture Modeling
Author Information
Author(s): Sandro Gsteiger, Stephan Morgenthaler
Primary Institution: Institute of Mathematics, Swiss Federal Institute of Technology, Lausanne, Switzerland
Hypothesis
Can mixture modeling effectively describe population heterogeneity in multistage carcinogenesis?
Conclusion
The study successfully extends the multistage carcinogenesis model using mixture modeling to account for population heterogeneity, achieving good fits with human lung cancer incidence data.
Supporting Evidence
- The mixture model allows for variability among individuals, which is biologically meaningful.
- Good fits were achieved for models combining a small high-risk group with a large quasi-immune group.
- The study highlights the importance of considering population heterogeneity in cancer modeling.
Takeaway
This study looks at how cancer develops in different people and shows that using a mixture model helps explain why some people get cancer while others don't.
Methodology
The authors applied finite mixture models to human lung cancer data and used analytic graduation for parameter estimation.
Potential Biases
Potential model mis-specification and reliance on biological assumptions may introduce bias.
Limitations
The model's assumptions may oversimplify the complexity of carcinogenesis, and the heavy censoring in the data affects the estimation.
Participant Demographics
The study focuses on human lung cancer incidence data from several birth cohorts of European Americans.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website