Cancer ResearchResearch PaperPaywall

Neoadjuvant Immunotherapy Delivers Near-Perfect 5-Year Survival in Melanoma Responders

A 1,038-patient international analysis finds 5-year overall survival exceeding 97% in melanoma patients who achieve a major pathological response to neoadjuvant checkpoint inhibitors.

Saturday, October 3, 2026 1 view
Published in Nat Med
A surgeon reviewing pre- and post-treatment MRI scans of a melanoma patient on a lightbox in a clinical oncology setting

Summary

Giving immunotherapy before surgery is now the preferred treatment for melanoma that has spread to lymph nodes but can still be removed. This pooled analysis of 1,038 patients at 26 centers gives the longest follow-up yet. About 58% of immunotherapy patients had a major pathological response, meaning almost no living tumor remained at surgery. Those patients had a five-year survival of about 98%, and continuing immunotherapy after surgery did not appear to help them. Combining two immunotherapy drugs worked better than one drug alone. Targeted BRAF/MEK drugs given before surgery performed poorly, with five-year recurrence-free survival of only 37%. Patients who did not respond remain at high risk, with five-year recurrence-free survival of 35%, and need new strategies. Results were weaker in acral melanoma, and subgroup sizes were small.

Detailed Summary

Neoadjuvant immunotherapy, giving checkpoint inhibitors before surgery, has replaced adjuvant treatment as the preferred approach for resectable stage IIIB–D melanoma. Earlier trials showed event-free survival hazard ratios of 0.52 and 0.32 versus adjuvant anti-PD-1. Those trials followed patients for only one to three years, though. Clinicians still lacked answers on which regimen is best, whether BRAF/MEK inhibitors have a role, whether to continue drugs after surgery, and how well therapy works in subgroups that trials underrepresented.

The International Neoadjuvant Melanoma Consortium pooled participant-level data from 1,038 patients treated at 26 centers between 2011 and 2024, 755 in trials and 283 outside them. Patients received checkpoint inhibitors (n = 735; 242 anti-PD-1 alone and 489 anti-PD-1 plus another immuno-oncology agent), BRAF/MEK inhibitors (n = 119) or checkpoint inhibitors plus targeted therapy (n = 184). Median age was 60, 61.8% were men, 49.1% carried BRAF V600 mutations and 96.6% had stage III disease. Median follow-up was 2.6 years, and pathological response followed 2018 INMC criteria.

Major pathological response (MPR) occurred in 57.8% of checkpoint-inhibitor patients, 51.4% with BRAF/MEK inhibitors and 44.6% with combined therapy. Combination immunotherapy outperformed anti-PD-1 alone (61.4% vs 49.5%). Five-year recurrence-free survival was 61.0% with anti-PD-1 alone, 73.9% with anti-PD-1 plus another agent, 37.4% with BRAF/MEK inhibitors and 76.4% with combined therapy. Five-year overall survival was 83.3%, 87.5%, 69.8% and 83.9%. The 70.6% five-year recurrence-free survival with immunotherapy compares favorably with the roughly 50–55% reported for adjuvant regimens.

Pathological response strongly stratified outcomes. In immunotherapy-treated patients, five-year overall survival was 98.0% with MPR and 71.6% with no response. Five-year recurrence-free survival was 91.2% with MPR versus 35.2% with no response. Among 681 patients with response data, those with MPR who received no adjuvant drug therapy had five-year recurrence-free survival of 90.6%, versus 90.8% with adjuvant checkpoint inhibitors, and similar overall survival. In non-responders, adjuvant therapy trended toward better recurrence-free survival (38.3% vs 25.6%), but assignment was not randomized.

Subgroup results varied. BRAF V600-mutant tumors had lower MPR rates (46.3% vs 55.6% in wild-type) and poorer recurrence-free survival, but overall survival did not differ. Acral melanoma had a 38.1% MPR rate and a five-year recurrence-free survival of 28.5%, and all six deaths occurred within two years. In-transit-only disease reached a 66.7% MPR rate. Among 17 resectable stage IV patients, 14 were alive at a median of 2.1 years. Nontrial patients had similar response rates but poorer event-free and recurrence-free survival. Outcomes did not differ by sex.

The authors conclude that anti-PD-1-based neoadjuvant therapy is preferred, that neoadjuvant BRAF/MEK inhibitors can largely be abandoned, that MPR patients may forgo adjuvant therapy, and that non-responders should be triaged to trials. Caveats include retrospective pooling, small subgroups, no central pathology or imaging review, nonrandom adjuvant assignment, wide confidence intervals and shorter follow-up for anti-PD-1 alone, and the risk that expert centers overstate real-world effectiveness. Whether lymph node dissection can safely be omitted remains under study.

Key Findings

  • Major pathological response occurred in 57.8% of checkpoint-inhibitor patients, 51.4% with BRAF/MEK inhibitors and 44.6% with inhibitor-plus-targeted therapy (n = 959 assessed)
  • Anti-PD-1 plus another immuno-oncology agent beat anti-PD-1 alone on MPR (61.4% vs 49.5%) and on 5-year RFS (73.9% vs 61.0%)
  • Patients with MPR after immunotherapy had 5-year overall survival of 98.0% and 5-year RFS of 91.2%, versus 71.6% and 35.2% with pathological nonresponse
  • Among MPR patients, 5-year RFS was 90.6% with no adjuvant therapy and 90.8% with adjuvant checkpoint inhibitors, suggesting no added benefit
  • Neoadjuvant BRAF/MEK inhibitors had a 5-year RFS of only 37.4% and 5-year OS of 69.8%, versus 70.6% RFS and 86.5% OS with checkpoint inhibitors
  • BRAF V600-mutant tumors had a lower MPR rate (46.3% vs 55.6% wild-type) and poorer RFS, but no overall survival difference
  • Acral melanoma had a 38.1% MPR rate and a 5-year RFS of 28.5%, while in-transit-only disease had a 66.7% MPR rate

Methodology

This was a pooled, participant-level analysis from the International Neoadjuvant Melanoma Consortium covering 1,038 patients with clinically detected, resectable stage IIIB–IV cutaneous melanoma from 26 centers, 72.7% in trials and 27.3% outside them. Outcomes were MPR by 2018 INMC criteria, EFS, RFS (measured from surgery) and OS, with a median follow-up of 2.6 years. Regimens were compared using Kaplan–Meier estimates, log-rank tests, multivariable models for response and interaction tests. The interaction between treatment and MPR was significant for RFS (P = 0.0113) but not for OS (P = 0.38). Adjuvant therapy was not adjusted for in the primary survival analysis because it was not randomly assigned.

Study Limitations

The analysis is retrospective and pooled, with small subgroups (for example 26 acral and 17 stage IV patients), no central review of pathology or imaging, and missing FDG-PET data. Follow-up is shorter for anti-PD-1 alone and for nontrial patients, confidence intervals are wide, and adjuvant therapy was non-randomly assigned, which introduces bias by indication. The targeted-therapy combination cohort was heterogeneous. The supplied text does not include the conflict-of-interest disclosures.

Enjoyed this summary?

Get the latest longevity research delivered to your inbox every week.

Enter your email to subscribe: