Document resource
Background Randomized trials have demonstrated superior outcomes with neoadjuvant immunotherapy (IO) compared to adjuvant IO alone in macroscopic Stage III or resectable Stage IV melanoma. However, the optimal peri-operative strategy, including IO regimen and surgical approach, remains undetermined. Real-world data, including the feasibility of limited resection and utility of ctDNA in predicting melanoma recurrence, can help inform optimal practice.Methods In our retrospective, single-institution study, we identified 76 patients with resectable melanoma treated with neoadjuvant IO from 2020 to 2025. Data pertaining to treatment course, anti-tumor outcomes and safety was collected. Signatera ctDNA results were available and reviewed for 24 patients.Results In our neoadjuvant cohort (n=76); the majority had a cutaneous primary (81%) and macroscopic stage III disease (93%). 55% (42/76) were treated with Ipilimumab plus Nivolumab (I/N), 41% (31/76) anti-PD1 monotherapy (PD1) and 4% (3/76) Nivolumab plus Relatlimab (Nivo/Rela). Compared to the PD1 cohort, patients treated with I/N were younger, more likely to have a BRAF mutation and higher burden of disease ( table 1). Sixty-four patients (84%) underwent planned surgical excision; 26 (41%) had an upfront CLND; and 31 (48%) had an index node excision (INE) with reflex CLND ultimately performed in 7 patients. The remaining patients had resection of non-nodal disease. Recurrence rates were comparable between INE (10%; 3/31) and CLND (12%; 3/26) cohorts. The MPR of the total cohort was 55% with a MPR rate of 49% with I/N and 65% with PD1. Thirty-one patients received adjuvant systemic therapy, 25 IO, and 6 BRAF/MEK inhibitors. Most patients with MPR did not receive adjuvant treatment (71%). Seven of 29 patients (24%) without MPR did not receive adjuvant therapy, 3 due to toxicity, 4 physician’s discretion. Four of these 7 patients recurred.Nineteen patients had baseline ctDNA results (figure 1a). At the time of surgery, ctDNA was undetectable in the majority of MPR patients (9/11), while the majority of non-MPR patients remained detectable (7/11). Pre-surgery ctDNA assessment prognosticated RFS (p=0.07, figure 1d). After surgery, ctDNA was detectable for only 1 patient, who ultimately recurred. To date, 3 patients have recurred; ctDNA positivity pre-dated clinical and/or radiographic recurrence.Conclusions Patients who underwent INE have comparable outcomes to patients who underwent upfront CLND. Subsequent adjuvant therapy in patients with MPR did not impact recurrence free survival (RFS). However, in patients without MPR, subsequent adjuvant therapy improved RFS (p= 0.046). Longitudinal ctDNA results correlate with clinical outcomes, and pre-surgery ctDNA detectability may predict recurrence.Ethics Approval This study was approved by Dana Farber Cancer Institute’s Ethics Board; approval numbers 11-181 and 23-692.Abstract 35 Table 1Table summarizing baseline characteristics and surgical and adjuvant approaches for each cohortAbstract 35 Figure 1Longitudinal ctDNA results