Test of four colon cancer risk-scores in formalin fixed paraffin embedded microarray gene expression data

Area of Science:

• Oncology
• Genomics
• Biostatistics

Background:

• Prognosis prediction for resected primary colon cancer relies on the TNM staging system.
• Investigated the utility of four gene expression risk scores for enhancing patient stratification.

Purpose of the Study:

• To evaluate if gene expression risk scores improve patient stratification beyond the TNM system.
• To assess the prognostic value of risk scores for relapse-free survival (RFS), survival after relapse (SAR), and overall survival (OS).

Main Methods:

• Applied microarray-based risk scores to 688 stage II/III colon tumors from the PETACC-3 trial.
• Assessed prognostic value using regression analysis and receiver operating characteristic (ROC) curves.
• Evaluated improvement over existing models including T-stage, N-stage, and microsatellite instability (MSI) status.

Main Results:

• All four risk scores (RSs) showed significant univariate association with OS or RFS.
• Individual RSs marginally improved RFS or OS models (AUC gains < 0.025).
• A combined score demonstrated higher prognostic value for OS (AUC increase from 0.6918 to 0.7321) and RFS (AUC increase from 0.6723 to 0.6945).

Conclusions:

• Tested gene expression-based risk scores offer prognostic information but have limited impact on established models.
• A combination of risk scores may provide more robust prognostic information.
• Distinct predictors for RFS and SAR may be necessary.