Efficacy and safety of immune checkpoint inhibitors and targeted therapies in resected melanoma: a systematic review and network meta-analysis

Background: Multiple immune checkpoint inhibitors (ICIs) and targeted therapies have been widely used as adjuvant treatments for high-risk resected melanoma, with unclear comparative efficacy and safety. Methods: PubMed, Embase, the Cochrane Library, and ClinicalTrials.gov were searched from database inception until 6 June 2023. We included RCTs that assess adjuvant ICIs or targeted therapies in high-risk resected melanoma. Frequentist random-effect network meta-analyses (NMA) were performed. The primary outcome was recurrence-free survival (RFS). Results: Eleven trials including 10,712 patients and comparing 10 treatments (nivolumab [Nivo], ipilimumab 3 mg/kg [Ipi3], Ipi10, pembrolizumab [Pemb], vemurafenib [Vemu], bevacizumab [Beva], Nivo + Ipi1, Nivo + Ipi3, dabrafenib plus trametinib [Dab + Tram], and placebo/observation [Pla/Obs]) were included. NMA showed that all treatments showed RFS benefit over placebo/observation except Ipi3 (hazard ratio [HR], 0.78; 95% CI, 0.58–1.05). Combination therapy of Nivo + Ipi3 was the most effective treatment, which significantly improved RFS compared with other treatments. NMA also showed that all treatments were associated with an increased risk of grade 3-5 adverse events over placebo/observation except Nivo (HR, 1.25; 95% CI, 0.87–1.80). NMA suggested that Nivo and Pemb were the two safest treatments except for placebo/observation. Although three combination therapies ranked as the top three in terms of RFS, they did not show significant overall survival benefits compared to monotherapies including Pemb, Nivo, Ipi3, and Ipi10. Conclusion: In this NMA, adjuvant Nivo and Pemb are the preferred options in patients with resected melanoma considering the benefits and harms. Combination therapy of Nivo + Ipi3 may be a promising strategy, but more evidence from phase 3 trials is needed. Systematic Review Registration: https://www.crd.york.ac.uk/PROSPERO/display_record.php?RecordID=438667, PROSPERO (CRD42023438667).