Predicting Risk of Colorectal Cancer After Adenoma Removal in a Large Community-Based Setting

Area of Science:

• Gastroenterology and Oncology
• Predictive Analytics in Medicine

Background:

• Current colorectal cancer (CRC) surveillance guidelines rely on polyp characteristics for risk stratification, which is often imprecise.
• Integrating additional patient risk factors may enhance the accuracy of postpolypectomy risk assessment.

Purpose of the Study:

• To compare the predictive performance of a comprehensive risk model against a polyp-only model for postpolypectomy CRC.
• To evaluate the utility of patient demographics and clinical history in refining CRC risk prediction after adenoma removal.

Main Methods:

• A Cox regression model was developed using data from 95,001 patients who underwent baseline colonoscopy with adenoma removal (2004-2016).
• The comprehensive model included age, diabetes, colonoscopy indication, and polyp findings (histology, size).
• A polyp-only model was used for comparison, with performance assessed by AUC and calibration.

Main Results:

• The comprehensive model demonstrated superior predictive performance in both development (AUC 0.71) and validation (AUC 0.70) cohorts compared to the polyp-only model (AUC 0.61 and 0.62, respectively).
• Both models showed adequate calibration, indicating reliable risk estimates.
• The comprehensive model showed a significant improvement in predicting postpolypectomy CRC diagnosis.

Conclusions:

• A comprehensive CRC risk prediction model incorporating patient age, diabetes, and colonoscopy indication, in addition to polyp findings, is more accurate than models based solely on polyp characteristics.
• This enhanced model offers improved risk stratification for patients following colonoscopic polyp removal, potentially optimizing surveillance strategies.