Salivary Extracellular MicroRNAs for Early Detection and Prognostication of Esophageal Cancer: A Clinical Study

Background and aims Early detection of esophageal squamous cell carcinoma (ESCC) will facilitate curative treatment. We aimed to establish a micro-RNA (miRNA) signature derived from salivary extracellular vesicles and particles (EVPs) for early ESCC detection and prognostication. Methods Salivary EVP miRNA expression was profiled in a pilot cohort (n=54) using microarray. Area under the receiver-operator characteristic curve (AUROC) and lasso regression analyses were used to prioritize miRNAs that discriminated ESCC patients from controls. Using quantitative reverse transcription-polymerase chain reaction, the candidates were measured in a discovery cohort (n=72) and cell lines. The prediction models for the biomarkers were derived from a training cohort (n=342) and validated in an internal cohort (n=207) and an external cohort (n=226). Results The microarray analysis identified 7 miRNAs for distinguishing ESCC patients from control subjects. Since one was not always detectable in the discovery cohort and cell lines, the other 6 miRNAs formed a panel. A signature of this panel accurately identified all-stage ESCC patients in the training cohort (AUROC=0.968) and was successfully validated in two independent cohorts. Importantly, this signature could distinguish early-stage (stage Ⅰ/ Ⅱ) ESCC patients from control subjects in the training cohort (AUROC=0.969, sensitivity=92.00%, specificity=89.17%), internal (sensitivity=90.32%, specificity=91.04%) and external (sensitivity=91.07%, specificity=88.06%) validation cohorts. Moreover, a prognostic signature based on the panel was established and efficiently predicted the high-risk cases with poor progression-free survival and overall survival. Conclusion The salivary EVP-based 6-miRNA signature can serve as noninvasive biomarkers for early detection and risk stratification of ESCC. Chinese Clinical Trial Registry, ChiCTR2000031507.