Prediction of Minor Salivary Gland Carcinoma: A Novel Nomogram and Risk Classification System for Overall Survival and Cancer-Specific Survival.

Abstract

Minor salivary gland carcinoma (MiSGC) is rare, and the understanding of this disease is insufficient. This study aimed to identify independent risk factors and develop a nomogram for evaluating the overall survival (OS) and cancer-specific survival (CSS) of patients with MiSGC. Retrospective cohort study. SEER database (Surveillance, Epidemiology, and End Results). We collected data from patients diagnosed with MiSGC between 2004 and 2015 from the SEER database. According to patient registration, all patients were randomly allocated to training sets and validation sets (2:1). Then, Kaplan-Meier product limit curves and Cox proportional hazard regressions were performed to estimate the prognostic effect of variables. Nomograms based on Cox proportional hazard regressions were established to estimate 3- and 5-year OS and CSS. Finally, the nomogram was developed by the training set, and validation was performed with the concordance index, calibration curves, and decision curve analyses. In total, 1787 MiSGC cases were registered in SEER. The concordance index for internal validation of OS and CSS prediction was 0.842 and 0.816; that of external validation was 0.871 and 0.831. The calibration plots showed good consistency between nomogram prediction and actual survival. The decision curve analysis showed substantial net benefits of the new predictive model. We constructed nomograms and a corresponding risk classification system predicting the OS and CSS of patients with MiSGC. These tools can generate simple-to-use clinical risk grouping and determine the relationship between adjuvant therapy and active surveillance.