An Individualized EMT-Related Gene Signature to Predict Recurrence-Free Survival in Stage II/III Colorectal Cancer Patients.

BACKGROUND: Approximately 30% of stage II and 50-60% of stage III colorectal cancer (CRC) patients who have undergone surgery will develop recurrence within 5 years. Thus, more reliable prognostic biomarkers are urgently needed to identify the high-risk subset of patients who will benefit from postoperative adjuvant therapy. METHODS: We retrospectively analyzed 911 stage II/III CRC patients in multiple cohorts. Using a series of bioinformatic and statistical approaches, an individualized prognostic signature was established in the training cohort and validated in four other independent cohorts. An integrated decision tree was generated to improve risk stratification, and a nomogram was built to quantify risk assessment for individual patients. RESULTS: Epithelial-mesenchymal transition (EMT) was identified as a dominant risk factor for recurrence-free survival (RFS) in stage II/III CRC patients. The EMT-related gene signature could discriminate high-risk subsets in a training cohort and four independent validation cohorts (with 473, 89, 130, 74 and 145 patients, respectively). Survival analyses demonstrated that the EMT-related gene signature served as an independent risk factor for RFS in different subgroups. The decision tree could optimize the risk stratification, and the nomogram could predict the 5-year RFS probability accurately. CONCLUSION: The proposed EMT-related prognostic signature is a useful biomarker to predict RFS and identify the high-risk subset in stage II/III CRC patients.