A Nomogram Based on Serum Biomarkers and Clinical Characteristics to Predict Survival in Patients With Non-Metastatic Nasopharyngeal Carcinoma

Objective: This study focused on developing an effective nomogram for improving prognostication for patients with primary nasopharyngeal carcinoma (NPC) restaged according to the eighth edition of the AJCC/UICC TNM staging system. Methods: Based on data of 5903 patients with non-metastatic NPC (primary cohort), we used Cox regression analysis to identify survival risk factors and created a nomogram. We used the nomogram to predict overall survival (OS), distant metastasis-free survival (DMFS) and disease-free survival (DFS) in the primary and independent validation (3437 patients) cohorts. Moreover, we compared the prognostic accuracy between the 8th TNM system and the nomogram. Results: The nomogram included gender, age, T stage, N stage, Epstein-Barr virus DNA, hemoglobin, C-reactive protein, lactate dehydrogenase, and radiotherapy with/without induction or concurrent chemotherapy. In the prediction of OS, DMFS and DFS, the nomogram had significantly higher concordance index (C-index) and area under ROC curve (AUC) than the TNM system alone. Calibration curves demonstrated satisfactory agreements between nomogram-predicted and observed survival. The stratification in different groups permitted remarkable differentiation among Kaplan-Meier curves for OS, DMFS and DFS. Conclusion: The nomogram led to a more precise prognostic prediction for NPC patients in comparison with the 8th TNM system. Therefore, it could facilitate individualized and personalized patients' counseling and care.