Neoadjuvant immunotherapy sets a new standard in stage III melanoma

More than 40% of patients with resectable stage III melanoma have disease recurrence within the first few years after initial treatment and the lack of overall survival (OS) benefit with adjuvant immune checkpoint inhibitors (ICIs) (Cancer Discov. 2023;13:2319–2338) prompted researchers to investigate different approaches. As far back as 10 years ago, it was suggested that neoadjuvant ICIs may have greater efficacy than adjuvant therapy due to the presence of the entire tumour and therefore induction of the complete neoantigen repertoire. Other potential advantages of neoadjuvant therapy include a reduction in tumour burden to enhance resectability and the possibility of gaining an early readout of pathological response as a surrogate of long-term OS.

Initial evidence for feasibility came from the phase I OpACIN trial in which neoadjuvant ipilimumab plus nivolumab expanded more tumour-resident T-cell clones than the same treatments given after surgery (Nat Med. 2018;24:1655–1661). Favourable efficacy results were seen in the phase II OpACIN-neo and PRADO trials, which were undertaken to optimise the neoadjuvant regimen (Lancet Oncol. 2019;20:948–960; Nat Med. 2022;28:1178–1188). In addition, the phase II SWOG S1801 trial demonstrated that neoadjuvant pembrolizumab improved event-free survival (EFS) compared with adjuvant-only pembrolizumab (N Engl J Med. 2023;388:813–823).

Now, with the recent publication of the phase III NADINA trial, the value of the neoadjuvant approach is conclusively confirmed (N Engl J Med. 2024;391:1696–1708). Patients with resectable, macroscopic stage III melanoma received either 2 cycles of neoadjuvant ipilimumab plus nivolumab or 12 cycles of adjuvant nivolumab. Estimated 12-month EFS was 84% in the neoadjuvant group compared with 57% in the adjuvant group, with an impressive hazard ratio for progression, recurrence or death of 0.32 (99.9% confidence interval 0.15–0.66). The extent of the benefit in NADINA is consistent with that seen in the OpACIN-neo and PRADO trials. Furthermore, based on updated data from the OpACIN-neo trial, outcomes after neoadjuvant therapy are expected to be durable: 82% of patients were recurrence free at 3 years and 3-year OS was 92% (Ann Oncol. 2023;34:420–430).

NADINA incorporated a personalised response-driven adaptive strategy such that only patients in the neoadjuvant group without a major pathological response – around 40% – received adjuvant treatment. A de-escalated treatment regimen for most patients, consisting of 2 cycles of neoadjuvant ICIs, could positively impact quality of life and the health economics may be substantial. One could also envision response-directed surgery as used in the PRADO trial, where patients with a major pathological response may not need therapeutic lymph node dissection (Nat Med 2022;28:1178–1188).

The OpACIN, OpACIN-neo and NADINA trials suggest that the addition of neoadjuvant CTLA-4 inhibition to PD-1 blockade improves efficacy over the regimen of anti-PD-1 antibody alone used in SWOG S1801; however, the advantages of combination therapy come with higher toxicity. An important next step is to identify subgroups of patients who may benefit from single or combination ICIs. Ideally, we will see biomarker-based personalisation in the future to ensure all patients receive the most appropriate forms of treatment. Various candidates for response and toxicity prediction are under investigation including tumour genomic features (e.g. tumour mutational burden and driver mutations); immune cell characteristics; host-related factors, such as the microbiome; and liquid biopsy markers, including circulating tumour DNA, extracellular vesicles and cytokines (Cancer Discov. 2023;13:2319–2338). Having seen such success in stage III melanoma, the scene is also now set for studies of neoadjuvant therapies earlier, in stage II melanoma.