Background
Phase III open-label RCT (NADINA), N=423 macroscopic (palpable) resectable stage III cutaneous melanoma (AJCC 8th ed). Stratified by BRAF V600 status, number of positive nodes, region. Enrolled 2020-2022 across 23 centers (Australia, Europe, US). Response-adapted: responders (pCR/near-pCR) received observation; non-responders received adjuvant nivolumab.
Interventions and follow up
Arm A: Neoadjuvant ipilimumab 1 mg/kg + nivolumab 3 mg/kg q3wk × 2 cycles → surgery → response-adapted adjuvant (pCR/near-pCR: observation; non-pCR: adjuvant nivolumab 480 mg q4wk × 1 year)
Arm B: Surgery → adjuvant nivolumab 480 mg q4wk × 1 year
Primary endpoint: Event-free survival (EFS)
mFollow up: 9.9 months
Arm B: Surgery → adjuvant nivolumab 480 mg q4wk × 1 year
Primary endpoint: Event-free survival (EFS)
mFollow up: 9.9 months
Results
12-month EFS: 83.7% vs 57.2%, HR 0.32, 95% CI 0.22-0.49, P<.001
pCR at surgery (Arm A): ~45%; major pathological response (≥90% regression) ~57%; overall response (≥50% regression) ~71%
OS: immature at primary analysis
pCR at surgery (Arm A): ~45%; major pathological response (≥90% regression) ~57%; overall response (≥50% regression) ~71%
OS: immature at primary analysis
Adverse events
Grade ≥3 immune-related AEs (neoadjuvant arm, pre-surgery): 29.7%
Most common Grade ≥3 AEs: colitis 8%, hepatitis 6%, rash 5%
Surgical delay due to AE: 12%
Total IO exposure: comparable between arms given response-adapted design
Most common Grade ≥3 AEs: colitis 8%, hepatitis 6%, rash 5%
Surgical delay due to AE: 12%
Total IO exposure: comparable between arms given response-adapted design
Conclusions
Neoadjuvant ipilimumab + nivolumab followed by response-adapted adjuvant therapy dramatically improved EFS vs adjuvant nivolumab alone (12-month EFS 83.7% vs 57.2%, HR 0.32, P<.001) in resectable stage III melanoma. The response-adapted design allowed 71% of neoadjuvant patients with major responses to avoid adjuvant therapy entirely, reducing treatment burden for responders.
Key Limitations
Short median follow-up (9.9 months); OS immature. The striking HR 0.32 may partly reflect baseline imbalances or limited follow-up. ~29% of neoadjuvant patients had Grade ≥3 immune AEs before surgery and 12% had surgical delays, requiring expert multidisciplinary management. The response-adapted design assumes major pathological responders are cured by neoadjuvant therapy alone, requiring longer follow-up to confirm. Generalizability to community settings (surgical timing, pathological response assessment) requires careful implementation. Optimal sequencing with BRAF/MEK inhibitors in BRAF-mutated patients undefined.
Clinical Context
NADINA (NEJM 2024) is the most practice-changing trial in resectable stage III melanoma surgery since MSLT-I/II. The response-adapted neoadjuvant ipi+nivo approach now represents a new preferred standard for resectable macroscopic stage III melanoma in ESMO guidelines. The design is clinically elegant: pCR patients (45-57%) achieve outstanding outcomes without adjuvant therapy, avoiding its toxicity, while non-responders still receive nivolumab, personalizing the adjuvant decision using pathological response as the biomarker. Together with SWOG S1801, NADINA has established neoadjuvant IO as superior to adjuvant-only IO for resectable melanoma.