3684 - Survival Outcomes and Recurrence Patterns of Nasopharyngeal Carcinoma Patients Achieving Complete Response after Induction Chemotherapy: A Retrospective Cohort Study

Purpose/Objective(s):

Nasopharyngeal carcinoma (NPC) has a good prognosis, but treatment-induced toxicities remain a concern. This has driven interest in radiotherapy dose de-escalation, which is currently guided mainly by complete response (CR) after induction chemotherapy (IC) or during radiotherapy. However, whether this approach increases recurrence risk remains unclear. This study aimed to explore whether the post-IC CR supports the radiotherapy dose de-escalation.

Materials/Methods:

This retrospective cohort study reviewed 600 newly diagnosed locally advanced NPC patients who received IC and CCRT/RT at a single-center hospital between January 1, 2012, and December 31, 2023. Pre- and post-IC MRIs were available for all patients. Post-IC responses were classified using RECIST version 1.1. Recurrence patterns were categorized as in-field, marginal, or out-field based on dosimetric analysis. The chi-square test or Fisher's exact test was used to compare categorical variables, with significant variables entered into multivariate logistic regression. Kaplan-Meier analysis and log-rank test were used to compare survival outcomes between the CR and non-CR groups. All tests were two-sided, P < 0.05 was considered statistically significant.

Results: Of the 600 patients, 163 (27.2%) achieved CR, while the remaining 437 (72.8%) had non-CR. There were no significant differences in clinical characteristics between the two groups. With a median follow-up of 62.4 months, the 3-year OS, DMFS, DFS, LRRFS and LRFS rates for CR vs. non-CR were 98.7% vs. 89.6% (P < 0.001), 91.9% vs. 83.4% (P < 0.05), 84.6% vs. 72.1% (P < 0.001), 93.2% vs. 90.7% (P = 0.370), and 94.4% vs. 94.3% (P = 0.945), respectively. Achieving post-IC CR did not reduce recurrence risk (LRFS: HR 0.979, 95% CI 0.551-1.743, P = 0.945; LRRFS: HR 0.804, 95% CI 0.499-1.296, P = 0.370). Among the entire cohort, 83 patients (13.8%) experienced recurrence and had significantly worse OS than those without recurrence (3-year OS: 86.7% vs. 93.0%, P < 0.001). Among patients with recurrence, OS did not differ between post-IC CR recurrent and non-CR recurrent subgroups (3-year OS: 90.0% vs. 84.0% P = 0.143). The rate of in-field failure was similar between the subgroups (70.0% vs. 76.2%, P = 0.132), while marginal failure rates were 30.0% and 14.3%, respectively.

Conclusion:

In this study, patients with post-IC CR had better OS and DMFS than those with non-CR, while the LRFS and LRRFS were not affected. Patients with recurrence had worse OS, but OS was similar in the CR and non-CR recurrence subgroups. These findings suggest that post-IC CR can predict better OS and DMFS, but it is not a reliable predictor for locoregional control. The practice of using post-IC CR to guide radiotherapy dose de-escalation in LANPC requires caution. The high rate of in-field failure pattern in both subgroups indicated the intrinsic treatment-resistance might not be monitored by MRI alone. Thus, further investigation into imaging biomarkers is needed.