ColoFinder: a prognostic 9-genes signature improve prognosis for 871 stage II and III colorectal cancer patients

Abstract

Colorectal cancer (CRC) is the heterogeneous disease with a high mortality rate, and still lacks an effective treatment. Our goal is to develop a robust prognosis model for predicting the prognosis in CRC patients. In this study, 871 stage II and III CRC samples were collected from six gene expression profilings. ColoFinder was developed using 9-genes signature based Random Survival Forest (RSF) prognosis model. The 9-genes signature recurrence score was derived with 5-fold cross validation to test the association with relapse-free survival and the value of AUC was gained with 0.87 inGSE39582(95% CI [0.83 − − 0.91]). The low-risk group had significantly better relapse-free survival (HR, 14.8; 95% CI [8.17–26.8];P< 0.001) than the high-risk group. We also found the 9-genes signature recurrence score contributed more information about recurrence than standard clinical and pathological variables in univariate and multivariate Cox analyses when applied toGSE17536(p= 0.03 andp= 0.01 respectively). Furthermore, ColoFinder improved the predictive ability and better stratified the risk subgroups when applied to CRC gene expression datasetsGSE14333,GSE17537,GSE12945andGSE24551. In summary, ColoFinder significantly improves the risk assessment in stage II and III CRC patients. The 9-genes prognostic classifier informs patient prognosis and treatment response.