MiRNA signature predicts the response of patients with advanced lung adenocarcinoma to platinum-based treatment.

Abstract

Accumulating literature proved that miRNAs can regulate the sensitivity of platinum and act as a promising candidate to predict the response of patients with lung adenocarcinoma to chemotherapy. However, most studies on miRNAs were restricted to in vitro experiments. This study aimed to evaluate whether miRNAs alone or in combination (miRNA signature) can act as predictive biomarkers of platinum-based chemotherapy in patients with lung adenocarcinoma. Eight miRNAs that most probably predict the efficacy of platinum were screened in 111 tumor tissues of lung adenocarcinoma. Univariate and multivariate Cox analyses, Kaplan-Meier survival curve analysis, Chi-square test, and univariate and multivariate logistic regression analyses were employed to determine whether miRNA expression is associated with the response of patients to platinum-based chemotherapy. The maximum significant odds ratio value was acquired by multiple cycles of multivariate logistic regression analysis. The cut-off points of miRNAs were obtained. A miRNA chemo-sensibility index (CI) formula was established, and its prediction performance was confirmed in another independent set (n = 31). Underexpression of three miRNAs (miRNA-21, miRNA-125b, and miRNA-224) was independently associated with the chemotherapy sensitivity of patients with lung adenocarcinoma. The miRNA CI formula containing these three miRNAs was calculated as (1.364 × miR-21) + (1.323 × miR-125b) + (1.131 × miR-224). A high CI was related to platinum-based chemotherapy resistance, and its prediction performance was confirmed in the testing set. The MAPK, PI3K-Akt, Ras, and cGMP-PKG signaling pathways were considered to be most probably correlated with platinum resistance. Our miRNA CI formula can act as an independent predictor to predict the response of patients with lung adenocarcinoma to platinum-based chemotherapy.