OBJECTIVES: To gain a better understanding of effectiveness, and patient outcomes of the first-line treatment options, cabozantinib+nivolumab and ipilimumab+nivolumab combination therapies, for advanced renal cell carcinoma (aRCC) by the generation of high-quality and robust real-world evidence (RWE).

METHODS: Linked physician and patient cross-sectional surveys, with retrospective data collection, of consulting aRCC patients receiving active second-line therapy for 12+ weeks were completed in 2022-23 in France, Germany, Italy, Spain and the UK. Patients were analysed according to first-line and second-line (post-ipilimumab-nivolumab only) treatment groups.

RESULTS: 216 physicians provided data on 1,473 patients; 497 (33.7%) completed a patient survey. Between 28-36% of patients were female, with a mean age at initiation of first-line therapy ranging from 65-67 years old. Most patients had an intermediate Memorial Sloan-Kettering Cancer Centre (MSKCC) score at first-line therapy initiation and 39-46% of patients had a nephrectomy. Physician determined real-world objective response rate of first-line therapy ranged from 71.6% to 78.0%.

CONCLUSIONS: Representativeness of this patient sample to the wider aRCC population was likely impacted by the requirement for patients to be actively consulting and treated with second-line therapy for inclusion. Consequently, our effectiveness findings contrasted with lower estimates observed in existing literature. Initiating a multi-country study to generate RWE promptly following a treatment launch introduces bias within the design, making real-world effectiveness difficult to assess. However, this study enabled generation of RWE on patient-reported outcomes and physician-reported medical data, highlighting discrepancies between patients and treating physicians on symptom burden. The relevance of the design, data type, sampling criteria and/or analysis methods to each research question and outcome should be evaluated, with separate alternative designs considered to maximize the quality of RWE and mitigate potential biases. There is a trade-off to consider between the fit for purpose of real-world data and expediting broad RWE generation.