A Risk Prediction Model for Sporadic CRC Based on Routine Lab Results

Area of Science:

• Oncology
• Preventive Medicine
• Biostatistics

Background:

• Current colorectal cancer (CRC) risk scores rely on demographic and behavioral factors, demonstrating limited predictive accuracy.
• There is a need for improved risk assessment tools for sporadic CRC.

Purpose of the Study:

• To develop and validate a novel risk prediction model for sporadic CRC.
• The model aims to utilize clinical and laboratory data from electronic medical records.

Main Methods:

• A nested case-control study was conducted using a UK primary care database.
• Predictors were identified through univariate and multivariate logistic regression, with discrimination assessed via receiver operating characteristic curves.
• Internal validation was performed to confirm model robustness.

Main Results:

• A demographic-based model showed limited predictive value (AUC 0.58).
• A laboratory-based model incorporating hematocrit, MCV, lymphocytes, and neutrophil-lymphocyte ratio (NLR) achieved an AUC of 0.76.
• A combined model including laboratory markers and clinical factors demonstrated superior performance (AUC 0.80) with improved calibration and reclassification.

Conclusions:

• Laboratory-based risk models demonstrate substantial predictive power for sporadic CRC.
• Integrating clinical and laboratory data offers a significant advancement in CRC risk prediction.