OBJECTIVE: This study was intended to identify prognostic biomarkers for lymph node (LN)-positive locoregional esophageal squamous cell carcinoma (ESCC) patients.
SUMMARY OF BACKGROUND DATA: Surgery is a major treatment for LN-positive locoregional ESCC patients in China. However, patient outcomes are poor and heterogeneous.
METHODS: ESCC-associated miRNAs were identified by microarray and validated by quantitative real-time polymerase chain reaction analyses in ESCC and normal esophageal epithelial samples. A multi-miRNA based classifier was established using a least absolute shrinkage and selection operator model in a training set of 145 LN-positive locoregional ESCCs, and further assessed in internal testing and independent validation sets of 145 and 243 patients, respectively.
RESULTS: Twenty ESCC-associated miRNAs were identified and validated. A 4-miRNA based classifier (miR-135b-5p, miR-139-5p, miR-29c-5p, and miR-338-3p) was generated to classify LN-positive locoregional ESCC patients into high and low-risk groups. Patients with high-risk scores in the training set had a lower 5-year overall survival rate [8.7%, 95% confidence interval (CI): 0-20.3] than those with low-risk scores (50.3%, 95% CI: 40.0-60.7; P < 0.0001). The prognostic accuracy of the classifier was validated in the internal testing (P < 0.0001) and independent validation sets (P = 0.00073). Multivariate survival analyses showed that the 4-miRNA based classifier was an independent prognostic factor, and the combination of the 4-miRNA based classifier and clinicopathological prognostic factors significantly improved the prognostic accuracy of clinicopathological prognostic factors alone.
CONCLUSION: Our 4-miRNA based classifier is a reliable prognostic prediction tool for overall survival in LN-positive locoregional ESCC patients and might offer a novel probability of ESCC treatment individualization.
