A molecular signature for the prediction of recurrence in colorectal cancer

BACKGROUND: Several clinical and pathological factors have an impact on the prognosis of colorectal cancer (CRC), but they are not yet adequate for risk assessment. We aimed to identify a molecular signature that can reliably identify CRC patients at high risk for recurrence. RESULTS: Two hundred eighty-one CRC samples (stage II/III) were included in this study. A two-step gene expression profiling study was conducted. First, gene expression measurements from 81 fresh frozen CRC samples were obtained using Affymetrix Human Genome U133 Plus 2.0 Arrays. Second, a focused gene expression assay, including prognostic genes and genes of interest from literature reviews, was performed using 200 fresh frozen samples and a Taqman low-density array (TLDA) analysis. An optimal 31-gene expression classifier for the prediction of recurrence among patients with stage II/III CRC was developed using logistic regression analysis. This gene expression signature classified 58.5% of patients as low-risk and 41.5% as high-risk (P < 0.001). The signature was the strongest independent prognostic factor in the multivariate analysis. The five-year relapse-free survival (RFS) rates for the low-risk patients and the high-risk patients were 88.5% and 41.3% (P < 0.001), respectively. CONCLUSION: We identified a 31-gene expression signature that is closely associated with the clinical outcome of stage II/III CRC patients.