Neoadjuvant immune checkpoint inhibitors for patients with resectable stage III/IV melanoma: A nationwide real-life study in France (NEOMEL).

9578 Background: Several studies have shown the efficacy of neoadjuvant treatment (NT) with immune checkpoint inhibitors (ICI) for resectable stage III/IV melanoma. However, real-life data are lacking. Methods: NEOMEL is a national multicenter retrospective study assessing the efficacy and tolerability of NT with ICI (NT-ICI) for resectable metastatic melanoma in a real-life setting. Data were obtained in patients (pts) with resectable stage III/IV melanoma who initiated NT-ICI (at least one infusion). Primary endpoint was pathologic complete response (pCR) (i.e no residual tumor) rate. Secondary endpoints were pathologic response (pR) according to INMC, radiologic response (RECIST), safety profile (CTCAE) and survival outcomes. Results: A total of 166 pts with resectable stage III/IV melanoma were included. Median age was 67 years (range, 24-96). Most pts had primary cutaneous melanoma (n=157, 89%), stage IIIB-D disease (n=156, 94%); a BRAF V600-mutant melanoma was observed in 47% (n=78); 8% (n=14) pts had received a previous systemic treatment. The median number of infusions of NT-ICI was 3 (range, 1-5). ICI regimens were anti-PD-1 monotherapy (n=122, 73%) or nivolumab + ipilimumab (NIVO + IPI) (n=44, 27%). Median follow-up was 6.5 months (range, 1-85). Overall, 143 pts (86%) underwent surgery. A pCR was observed in 44% of pts (n=63/143), including 52% (n=51/99) treated with anti-PD-1 monotherapy, and 27% (n=12/44) treated with NIVO + IPI. Of note, 78 pts (54%) were treated with 3 infusions of pembrolizumab (PBZ) (according to SWOG 1801 protocol) and had a pCR in 45% (n=35). A pR was assessed according to INMC criteria in 58% of pts (n=83/143): 51% of pCR (n=42), 8% of near-pCR (n=7) corresponding to a major pR of 59% (n=49/83), 16% of pathologic partial response (n=13) and 25% of pathologic non-response (n=21). Twenty-three pts (14%) did not undergo surgery because of progressive disease (n=10), clinical and radiologic responses (8 complete responses, 1 partial response, 1 stable disease), toxicity (n=2) or patient’s choice (n=1). Radiologic overall response rate according to RECIST (data available after NT in 49% of pts) was 54% (n = 44/82). All grade immune-related adverse events (irAEs) of NT-ICI occurred in 31% (n=51) of pts, including grade 3-4 irAEs in 10% (n=17) and one grade 5 irAE (pneumonitis with PBZ). Severe irAEs (grade 3 or more) were observed in 4% (n=4/99) of pts treated with anti-PD1 monotherapy and 32% (n=14/44) of pts treated with NIVO + IPI. Adjuvant treatment was started for 120 pts (84%). At last follow-up, 26 pts (18%) have had recurrence after surgery, including 3 pts with a pCR and 23 pts with a non-pCR to NT-ICI. Survival data are not yet available. Conclusions: NT-ICI for advanced melanoma demonstrated its efficacy with high rates of pCR in a real-life setting and an acceptable safety profile, particularly NT anti-PD1 monotherapy.