Overall survival in the real-world and clinical trials: a case study validating external controls in advanced melanoma

Clinical trials are the gold standard for measuring the efficacy and safety of new treatments. Comparisons between clinical trials and external controls drawn from real-world data are potentially valuable – especially when randomized trials are not available or feasible – but carry important risks of bias stemming from differences across populations, care settings and measurement of patient characteristics and outcomes. As a case study, we assessed the suitability of a particular real-world database (the Flatiron Health Database) for analyzing overall survival among patients in clinical trials of treatments for metastatic melanoma. Challenges included differences in patient baseline prognostic factors across populations, including high proportions with missing information in real-world data. After accounting for these differences, we observed similar survival between patients receiving ipilimumab monotherapy in clinical trials and in real-world data. We conclude that real-world external controls can be suitable for metastatic melanoma.