Related Experiment Video
Updated: Jun 26, 2026
07:13
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
An expanded risk prediction model for lung cancer.
Margaret R Spitz1, Carol J Etzel, Qiong Dong
1Department of Epidemiology, Unit 1340, The University of Texas M. D. Anderson Cancer Center, P.O. Box 301439, Houston, TX 77230-1439, USA. mspitz@mdanderson.org
Cancer Prevention Research (Philadelphia, Pa.)
|January 14, 2009
Summary
Adding DNA repair capacity markers to lung cancer risk models modestly improved prediction accuracy for both former and current smokers. These biomarkers enhance the sensitivity of risk assessment for lung cancer.
Area of Science:
- Oncology
- Genetics
- Epidemiology
Background:
- Clinical decision-making relies on risk prediction models.
- A previously developed lung cancer prediction tool showed modest precision.
- Enhancing this tool with DNA repair markers was investigated.
Purpose of the Study:
- To evaluate the improvement in lung cancer risk prediction by incorporating DNA repair capacity markers.
- To assess the impact of host-cell reactivation and mutagen sensitivity assays on model precision.
Main Methods:
- Multivariable models were built using existing variables and adding DNA repair biomarkers (host-cell reactivation, mutagen sensitivity).
- Data from 725 lung cancer cases and 615 controls (smokers) were analyzed.
- Area Under the Receiver Operating Characteristic Curves (AUC) and cross-validation were used for comparisons.
Main Results:
- Expanded models showed statistically significant improvements in AUC for both former (0.67 to 0.70) and current smokers (0.68 to 0.73).
- Individuals with poor DNA repair capacity or heightened mutagen sensitivity had increased lung cancer risks.
- Biomarker addition improved the sensitivity of the expanded risk prediction models.
Conclusions:
- Incorporating DNA repair capacity markers modestly enhances the precision of lung cancer risk prediction models.
- These biomarkers identify individuals with higher absolute risks of developing lung cancer.
- The findings support the use of DNA repair assays to refine lung cancer risk assessment.