Mutational landscape of nasopharyngeal carcinoma based on targeted next-generation sequencing: implications for predicting clinical outcomes

BACKGROUND: The aim of this study was to draw a comprehensive mutational landscape of nasopharyngeal carcinoma (NPC) tumors and identify the prognostic factors for distant metastasis-free survival (DMFS). METHODS: A total of forty primary nonkeratinizing NPC patients underwent targeted next-generation sequencing of 450 cancer-relevant genes. Analysis of these sequencing and clinical data was performed comprehensively. Univariate Cox regression analysis and multivariate Lasso-Cox regression analyses were performed to identify factors that predict distant metastasis and construct a risk score model, and seventy percent of patients were randomly selected from among the samples as a validation cohort. A receiver operating characteristic (ROC) curve and Harrell’s concordance index (C-index) were used to investigate whether the risk score was superior to the TNM stage in predicting the survival of patients. The survival of patients was determined by Kaplan–Meier curves and log-rank tests. RESULTS: The twenty most frequently mutated genes were identified, such as KMT2D, CYLD, and TP53 et al. Their mutation frequencies of them were compared with those of the COSMIC database and cBioPortal database. N stage, tumor mutational burden (TMB), PIK3CA, and SF3B1 were identified as predictors to build the risk score model. The risk score model showed a higher AUC and C-index than the TNM stage model, regardless of the training cohort or validation cohort. Moreover, this study found that patients with tumors harboring PI3K/AKT or RAS pathway mutations have worse DMFS than their wild-type counterparts. CONCLUSIONS: In this study, we drew a mutational landscape of NPC tumors and established a novel four predictor-based prognostic model, which had much better predictive capacity than TNM stage. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s10020-022-00479-4.