Summary
A pooled analysis of 1,038 patients found 86.5% five-year overall survival after neoadjuvant immune checkpoint inhibitors for resectable melanoma. Among patients whose tumours had a major pathological response, estimated five-year survival was 98%.
A pooled analysis of 1,038 people with resectable stage IIIB–IV melanoma found an estimated five-year overall survival of 86.5% among those treated with immune checkpoint inhibitors before surgery. For patients whose tumours showed a major pathological response to immunotherapy, estimated five-year survival was 98.0%. The study, published in Nature Medicine on 1 October 2026, brings together data from 26 international centres and extends follow-up of earlier research on treatment given before an operation.
Neoadjuvant treatment is given before definitive surgery. Immune checkpoint inhibitors (ICIs) target proteins that restrain immune responses, helping immune cells attack cancer. After treatment, examination of the removed tumour provides a measure of how much viable cancer remains. In this analysis, a major pathological response (MPR) meant that no more than 10% of the tumour bed contained viable tumour cells.
Five-year outcomes varied by treatment and tumour response
The researchers pooled patient-level data collected between November 2023 and October 2024. The cohort included 735 people treated with ICIs, 119 treated with BRAF and MEK inhibitors, and 184 treated with an ICI plus targeted therapy. Most participants had stage III disease; 35 had resectable oligometastatic stage IV melanoma. About 73% were treated in clinical trials, with the remainder treated outside trials.
Across the ICI group, estimated five-year recurrence-free survival was 70.6% and overall survival was 86.5%. Recurrence-free survival measures the time from surgery to recurrence or death; overall survival measures the time from the first treatment dose to death from any cause.
Within the ICI group, five-year recurrence-free survival was 61.0% with anti-PD-1 treatment alone and 73.9% with anti-PD-1 combined with another immuno-oncology agent. The corresponding five-year overall survival estimates were 83.3% and 87.5%. These pooled results describe outcomes among patients who received each regimen; they are not a randomized comparison of the treatments.
Pathological response sharply separated outcomes. ICI-treated patients with an MPR had estimated five-year overall survival of 98.0% (95% confidence interval 96.4–99.6%). The estimate was 98.8% for patients receiving anti-PD-1 alone and 97.9% for those receiving anti-PD-1 plus another immunotherapy. By contrast, the five-year overall survival estimate for ICI-treated patients with no pathological response was 71.6%.
The study also found that, among ICI-treated patients with an MPR, five-year recurrence-free survival was similar whether or not they received additional ICI treatment after surgery: 90.6% without adjuvant treatment and 90.8% with it. Because postoperative treatment was not assigned randomly, this finding is an observational comparison, not a basis on its own for changing treatment decisions.
How to interpret the pooled results
The findings support pathological response as a useful marker of long-term outlook after neoadjuvant immunotherapy: the amount of viable tumour remaining after treatment was associated with markedly different survival estimates. This may help researchers design studies that focus on people whose tumours respond poorly, while examining whether patients with strong responses need further treatment after surgery.
The analysis was pooled and retrospective, and treatment groups were not interchangeable. For example, the BRAF/MEK inhibitor group had a higher proportion of stage IIIC disease and greater measured tumour burden than the ICI groups. The authors also report a median follow-up of 2.6 years, even though survival estimates are given at five years. The estimates therefore use the available follow-up data rather than representing five years of observation for every participant. Smaller subgroups, including people with acral melanoma, require further study.
The results add long-term context to neoadjuvant immunotherapy for resectable melanoma, but they do not establish that one regimen caused better survival than another. The authors identify prospective research as important, particularly to test postoperative treatment strategies for patients whose tumours show no pathological response.