9561 Background: The introduction of immunotherapy and targeted therapy has revolutionized the treatment landscape for metastatic melanoma. However, clinical trial data in pts with MBM are scarce; here we summarize the available clinical evidence. Methods: An SLR was conducted by searching EMBASE, MEDLINE, and the Cochrane Central Register of Controlled Trials databases through November 13, 2020. When available, Kaplan-Meier (K-M) curves for overall survival (OS) were digitized and converted to pseudo-individual pt data using the Guyot algorithm. A meta-analysis of the pooled K-M curves was performed for selected interventions, including immunotherapies and targeted therapies. Median OS was calculated either through the meta-analysis of K-M curves or as a weighted average of median OS (table). When interventions were reported in only 1 study, the value reported was used instead of the weighted average and compared qualitatively with the other results. Results for treatment modalities other than systemic agents will also be presented. Results: The SLR included 70 publications that evaluated systemic therapies in pts with MBM for qualitative evidence synthesis: 12 pertaining to 7 randomized controlled trials, 55 pertaining to 46 single-arm studies, and 3 involving nonrandomized comparative studies. The pt population was highly heterogeneous with respect to prior therapies, pt characteristics, and neurological symptoms. For the meta-analysis, a total of 25 K-M curves from 12 studies representing 6 interventions and 1043 pts were digitized. Based on the pooled K-M curves, median OS was numerically longer with nivolumab plus ipilimumab (NIVO + IPI; 28.3 mo; 95% CI, 19.7-31.9) than with the other interventions (range 5.7-11.8 mo; table). Similar OS benefit was also observed with NIVO + IPI when the weighted average of the median was used (in a long-term study, median OS was 29.2 mo) compared with the other interventions. Conclusions: Given the lack of comparative clinical trial data in pts with MBM, there remains an unmet need to determine the best approach to treat these pts. This analysis suggested a clinical advantage with NIVO + IPI compared with other systemic agents analyzed. The heterogeneity of the available data added uncertainty to our treatment assessments. Therefore, these findings warrant further research into the best approach to improve outcomes in pts with MBM.[Table: see text]
Real-World or Controlled Clinical Trial Data in Real-World Practice
