A Bayesian network meta-analysis of the primary definitive therapies for locoregionally advanced nasopharyngeal carcinoma: IC+CCRT, CCRT+AC, and CCRT alone.

Abstract

The major aim of this Bayesian network analysis was to determine the optimal treatment strategy for locoregionally advanced nasopharyngeal carcinoma (LANPC). We systematically searched databases and extracted data from randomized clinical trials involving LANPC patients randomly assigned to receive induction chemotherapy followed by concurrent chemoradiotherapy (IC+CCRT), CCRT followed by adjuvant chemotherapy (CCRT+AC), or CCRT. In the network analysis, IC+CCRT was significantly better than CCRT alone for 5-year FFS (odds ratio [OR]: 1.63, 95% credible interval [CrI] 1.16-2.29), DMFS (OR: 1.56, 95% CrI 1.08-2.22), and LFRS (OR: 1.62, 95% CrI 1.02-2.59), but not OS (OR: 1.35, 95% CrI 0.92-2.00). Rank probabilities showed that IC+CCRT was ranked the best followed by CCRT+AC and CCRT for all 5-year outcomes. Although compared to IC+CCRT and CCRT, CCRT+AC did not significantly improve survival but had the highest 5-year survival rates. IC+CCRT could be recommended as a front-preferred primary definitive therapy for patients with LANPC.