Plasma Epstein-Barr viral DNA load after completion of two cycles of induction chemotherapy predicts outcomes for patients with advanced-stage nasopharyngeal carcinoma

Objectives: To evaluate the prognostic value of plasma Epstein-Barr virus DNA level following the completion of two induction chemotherapy cycles (ICT; post(2CICT)-DNA) in locoregionally advanced nasopharyngeal carcinoma (LA-NPC). Materials and methods: This retrospective study included 534 patients with LA-NPC. Recursive partitioning analysis (RPA) was applied to derive a prognostic model for risk stratification. Kaplan-Meier survival analysis was used to determine the survival results, and survival rates were compared using the log-rank test. The Cox proportional hazard model was used for univariate and multivariate analyses. Results: Multivariate analyses revealed that post(2CIC)(T)-DNA and N stage were independent predictors of overall survival (OS; P = 0.001 and P = 0.001, respectively), and post(2CIC)(T)-DNA, pre-treatment DNA, and N stage were independent predictors of progression-free survival (PFS; P = 0.002, P = 0.001, and P = 0.021, respectively). Based on prognostic factors (pre-treatment DNA, post(2CIC)(T)-DNA, and N stage), patients were stratified into three risk subgroups, with 288 patients in the low-, 213 in the intermediate-, and 33 in the high-risk group. The three-year OS rate of the low-, intermediate- and high-risk groups were 99.3% (95% CI 98.3%-100.0%), 90.0% (95% CI 85.5%-94.5%) and 67.0% (95% CI 49.9%-84.1%, P < 0.001 for each of the two groups), respectively. Conclusion: Plasma EBV-DNA level after two ICT cycles is a powerful predictor of prognosis in patients with LA-NPC. RPA analysis revealed that stage N3 patients with detectable post(2CIC)(T)-DNA are at the highest risk of treatment failure, and future clinical trials should focus on early-treatment modification strategies for these patients.