In this phase II clinical trial, patients with high-risk resectable melanoma were treated with neoadjuvant checkpoint blockade. In the 12 patients treated with nivolumab, a 25% ORR (25% pathologic CR) was observed. In the 11 patients treated with nivolumab/ipilimumab, a 73% ORR (45% pCR) was observed and all patients in this arm were surviving at 15 months. Response to nivolumab alone appeared dependent on a pre-existing, antigen-experienced, clonal, and diverse T cell infiltrate. The trial closed early because neoadjuvant nivolumab treatment delayed surgical resection in two patients and toxicity was high in combination-treated patients.
Preclinical studies suggest that treatment with neoadjuvant immune checkpoint blockade is associated with enhanced survival and antigen-specific T cell responses compared with adjuvant treatment(1); however, optimal regimens have not been defined. Here we report results from a randomized phase 2 study of neoadjuvant nivolumab versus combined ipilimumab with nivolumab in 23 patients with high-risk resectable melanoma ( NCT02519322 ). RECIST overall response rates (ORR), pathologic complete response rates (pCR), treatment-related adverse events (trAEs) and immune correlates of response were assessed. Treatment with combined ipilimumab and nivolumab yielded high response rates (RECIST ORR 73%, pCR 45%) but substantial toxicity (73% grade 3 trAEs), whereas treatment with nivolumab monotherapy yielded modest responses (ORR 25%, pCR 25%) and low toxicity (8% grade 3 trAEs). Immune correlates of response were identified, demonstrating higher lymphoid infiltrates in responders to both therapies and a more clonal and diverse T cell infiltrate in responders to nivolumab monotherapy. These results describe the feasibility of neoadjuvant immune checkpoint blockade in melanoma and emphasize the need for additional studies to optimize treatment regimens and to validate putative biomarkers.