LAUSR

Abstract Background To analyze patterns of failure for patients with LA-NSCLC receiving definitive chemoradiotherapy and build a nomogram for predicting the failure patterns in these patients. Methods Clinicopathological materials of patients between 2013 and 2016 with LA-NSCLC receiving definitive chemoradiotherapy and following-up in our hospital were collected. The endpoint was the first failure after definitive chemoradiotherapy. Based on logistic regression, the predictive value of each factor was evaluated and nomogram was built. This model was validated by ROC curve, calibration curve and decision curve analysis (DCA). Results With a median follow-up of 28 month, 100 patients were observed failure. Local failure and distant failure were 46 and 54, respectively. Univariate and multivariate analysis indicated that younger (p = 0.016, OR (95%CI): 0.936 (0.887-0.987)), peripheral NSCLC (p = 0.025, OR (95%CI): 2.732 (1.137-6.567)) and epidermal growth factor receptor (EGFR) mutant (p = 0.020, OR (95%CI): 3.747 (1.234-11.381)) were independent predictors of distant failure, which were included in the nomogram. ROC curve showed that area under the ROC curve (AUC) of the nomogram was 0.713, which was better than any factors along. Calibration curve revealed a satisfactory consistency between the predicted distant failure and actual observation. DCA showed most of the threshold probabilities in this model were with good net benefits. Conclusions We concluded that age, tumor location and EGFR mutation status could predict failure patterns in patients with LA-NSCLC receiving definitive chemoradiotherapy. A nomogram was built and validated based on these factors, showing a potential predictive value in clinical practice. Legal entity responsible for the study Shanghai Chest Hospital, Shanghai Jiao Tong University. Funding Has not received any funding. Disclosure All authors have declared no conflicts of interest.