Management of resected stage III/IV melanoma with adjuvant immunotherapy.

Background: Adjuvant anti-PD1 therapy reduces the risk of recurrence in resected stage III/IV melanoma and is now standard care. Limited data exist beyond registration trials. We sought to explore the use of adjuvant immunotherapy in routine clinical practice. Methods: Patients (pts) from 11 Australian centres who received adjuvant nivolumab (nivo) for resected stage III/IV melanoma were included in this study. Efficacy, toxicity, surveillance, recurrence characteristics, management and further treatment outcomes were examined. Results: 471 pts received adjuvant nivo between 8/2018 to 3/2020. 318 (68%) were male, median age 64y (range 17-94), 28 (6%) were AJCC v8 IIIA, 194 (41%) IIIB, 175 (37%) IIIC, 11 (2%) IIID, and 63 (13%) IV. 65 (14%) pts had in-transit only disease, 152 (37%) pts were sentinel lymph node biopsy (SLNB+) and only 9 (6%) of these had CLND. 128 (27%) had BRAF mutant (BRAFmt) melanoma. Median time from resection to start of adjuvant nivo was 1.8 months (mo) (range 0.2-4.0). Median FU was 17.5 mo. 256 (54%) pts completed 12 months of nivo, 86 (18%) ceased early for toxicity, 76 (16%) for disease recurrence, 25 (5%) other reasons (COVID-19 8, co-morbidities 7, pt choice 10); 28 (6%) pts were still receiving nivo at data cut. Median duration of treatment was 10.4 mo (range 0-16.8). 117 (25%) pts recurred; 76 (65%) while ON nivo and 41 (35%) OFF nivo ( > 1 month after last dose, including 20 pts who stopped early for toxicity). 24 mo RFS was 69%. Median time to recurrence was 6.0 mo (95% CI 5.1, 7.5). 56 (48%) had first recurrence with locoregional (LR) disease only and 61 (52%) had distant +/- LR recurrence. Of those who recurred with LR disease only, 46/56 (82%) underwent surgery, 15/46 (33%) then had adjuvant radiotherapy, and 15/46 (33%) had ‘second adjuvant’ therapy with BRAF/MEK inhibitors (15/21, 71% BRAFmt pts). 10/56 (37%) pts who recurred with LR disease subsequently recurred distantly. 58/80 (73%) pts received systemic therapy at either 1st or subsequent unresectable recurrence. For recurrences ON nivo, 18 pts received combination ipilimumab (ipi) and nivo (ORR 44%), 4 pts had ipi monotherapy (ORR 0%), 7 pts had anti-PD1 + investigational agent (ORR 57%), 11 pts had BRAF/MEK inhibitors (ORR 82%). 1 pt had PD with ongoing PD1 monotherapy. For recurrences OFF nivo, no patients responded to PD1 alone (n = 1) or with an investigational agent (n = 1), ipi+nivo (n = 3), ipi monotherapy (n = 4) or chemotherapy (n = 2); 6 pts received BRAF/MEK inhibitors (ORR 50%). 2-year OS was 92%. Conclusions: Despite higher rates of discontinuation due to toxicity compared with clinical trial cohorts, the efficacy data appear similar. Most early recurrences are distant, and many with LR recurrence soon recur distantly thereafter. Second line adjuvant BRAF/MEK inhibitors are frequently used for resected LR recurrence. Both ipi+nivo and BRAF/MEK inhibitors appear to have activity after distant recurrence