Induction chemotherapy improved long term outcomes in stage iv locoregional advanced nasopharyngeal carcinoma

Purpose: We aimed to determine whether adding induction chemotherapy (IC) to concurrent chemoradiation (CCRT) improved outcomes in each stage of locally advanced nasopharyngeal carcinoma (LANPC). Methods: From 2007 to 2013, we retrospectively collected 259 histopathologically identified adult LANPC patients from two campuses in south Taiwan. Among the 238 eligibly treated cases, 156 patients received CCRT (CCRT group) upfront and 82 received IC followed by CCRT (IC group). Of these patients, 130 were stage III (92 patients that received CCRT and 38 that received IC adding CCRT) and 108 were stage IV (76 CCRT and 32 IC adding CCRT). Most chemotherapy regimens for IC are composed of cisplatin (P), 5-fluorouracil (F), and ifosfamide (I), while concurrent chemotherapy (CC) was essentially cisplatin-based. For CCRT as the upfront treatment, a P or PF regimen was usually used in CC. Survival outcomes were accessed with a Kaplan-Meier estimate and a p-value by log-rank test to compare the survival distributions of IC added to CCRT or CCRT as the upfront treatment in all LANPC stage III and LANPC IV patients. The failure free survival (FFS), overall survival (OS), local relapse free survival (LRFS), regional recurrence-free survival (RRFS), distant metastasis-free survival (DMFS), first failure site, and other prognostic factors were analyzed. Results: The median follow-up time of all treated LANPC patients was 59 months. For all LANPC patients, there was a significant difference only in the DMFS favoring IC group (91.5% vs 79.4%, p=0.013). In the subgroup study, for the stage III group, there was no significant difference between the groups for overall OS (IC group 71.3% vs CCRT group 78.7%), FFS (71.5% vs 62.4%) and RRFS (91.9% vs 90.9%). However, inferior LRLS (71.7% vs 91.5%; p = 0.03) was noted for the IC group. In contrast, for stage IV, there were significantly longer OS (75.8% vs 52.6%), FFS (66.8% vs 46.8%), and DMFS (86.0% vs 69.6%; p = 0.02, p = 0.04, and p = 0.03, respectively) rates in the IC group. Conclusion: Adding PIF-based IC to CCRT for the LANPC patients resulted in better outcomes for stage IV patients, but not for stage III patients. A future properly designed study should stratify enough LANPC cases under the structure of the AJCC stage grouping system to determine which subgroups truly benefit from adding IC to CCRT.