Background: In phase 1-2 trials in patients with resectable, macroscopic stage III melanoma, neoadjuvant immunotherapy was more efficacious than adjuvant immunotherapy.
Methods: In this phase 3 trial, we randomly assigned patients with resectable, macroscopic stage III melanoma to two cycles of neoadjuvant ipilimumab plus nivolumab followed by surgery or surgery followed by 12 cycles of adjuvant nivolumab. Only patients in the neoadjuvant group with a partial response or nonresponse received adjuvant treatment. The primary end point was event-free survival.
Results: A total of 423 patients underwent randomization. At a median follow-up of 9.9 months, the estimated 12-month event-free survival was 83.7% (99.9% confidence interval [CI], 73.8 to 94.8) in the neoadjuvant group and 57.2% (99.9% CI, 45.1 to 72.7) in the adjuvant group. The difference in restricted mean survival time was 8.00 months (99.9% CI, 4.94 to 11.05; P<0.001; hazard ratio for progression, recurrence, or death, 0.32; 99.9% CI, 0.15 to 0.66). In the neoadjuvant group, 59.0% of patients had a major pathological response, 8.0% had a partial response, 26.4% had a nonresponse (>50% residual viable tumor), and 2.4% had progression; in 4.2%, surgery had not yet been performed or was omitted. The estimated 12-month recurrence-free survival was 95.1% in patients in the neoadjuvant group who had a major pathological response, 76.1% among those with a partial response, and 57.0% among those with a nonresponse. Adverse events of grade 3 or higher that were related to systemic treatment occurred in 29.7% of patients in the neoadjuvant group and in 14.7% in the adjuvant group.
Conclusions: Among patients with resectable, macroscopic stage III melanoma, neoadjuvant ipilimumab plus nivolumab followed by surgery and response-driven adjuvant therapy resulted in longer event-free survival than surgery followed by adjuvant nivolumab. (Funded by Bristol Myers Squibb and others; NADINA ClinicalTrials.gov number, NCT04949113.).
Copyright © 2024 Massachusetts Medical Society.