Timing matters: Early vs late recurrence in nasopharyngeal carcinoma and its impact on survival

Vollständiger Abstract

Worum geht es in dieser Arbeit?

BACKGROUND Despite the wide utilization of intensity-modulated radiation therapy and other advanced treatment modalities, approximately 10% of patients with nasopharyngeal carcinoma (NPC) still suffer from local or regional disease recurrence, leading to deteriorated survival outcome. Particularly, early recurrence is often associated with more aggressive biological behavior and greater resistance to initial treatments, subsequently resulting in poorer survival outcomes. AIM To investigate the temporal distribution of site-specific recurrence and evaluate the prognostic impact of early vs late recurrence in NPC. METHODS This study retrospectively included 330 recurrent NPC patients who received initial treatment of radical radiotherapy from January 1990 to December 2020 in our center. Patients were categorized into early and late recurrence groups using the maximally selected rank method. Post-recurrence overall survival (OS) was evaluated utilizing Kaplan-Meier estimates and multivariable Cox regression analysis. RESULTS Among the 330 eligible patients, 167 developed pure local recurrence, 86 had pure regional recurrence, and 77 had locoregional recurrence. With a median follow-up duration of 97.8 months, there was no significant difference in timing distribution across various patterns of failure. The peak incidence of progression was observed in the 2nd year, contributing 29.4% of all progression events, with a cumulative 5-year proportion of 82.8%. The optimal cutoff for differentiating early and late recurrence was identified as 24 months. Advanced T (P = 0.008) and N (P = 0.028) stages were predisposing factors for early recurrence. Survival analysis manifested that late locoregional recurrence was associated with markedly better post-recurrence OS [hazard ratio (HR) = 0.37, P = 0.006], whereas such scenario was not observed in entire cohort or sub-cohorts with pure local or regional failure. The complementary 24-month landmark analysis further unveiled that late recurrence was associated with superior post-landmark OS across the entire cohort (HR = 0.38, P < 0.001) and all failure subgroups. CONCLUSION Recurrence timing is a pivotal prognostic factor in recurrent NPC. Early locoregional recurrence within 24 months is indicative of significantly inferior post-recurrence OS, warranting more intensive surveillance and aggressive salvage strategies for this subset.