OBJECTIVES: Multiple database locks (DBLs) with different levels of data maturity have become available for CheckMate-214 (NCT02231749), a phase 3 trial comparing nivolumab+ipilimumab to sunitinib in first-line advanced/metastatic intermediate- or poor-risk renal cell carcinoma (1L RCC). This study assessed the stability of lifetime overall survival (OS) extrapolations based on different DBLs.

METHODS: Hazard ratios (HRs) for OS of nivolumab+ipilimumab versus sunitinib were compared for CheckMate‑214 DBLs with a minimum of 24, 30, and 37 months follow-up. Parametric functions (traditional distributions and cubic splines) were used to extrapolate OS to a lifetime time horizon based on previously published 18- and 30-month DBLs of CheckMate-214, corrected for British population background mortality. Selection of the best performing model was based on NICE DSU guidance (TSD-14). Functional form of the extrapolations, mean OS, and landmark OS were compared across DBLs to assess stability of lifetime OS estimates.

RESULTS: The OS HR was 0.63 at 18-months follow-up, remaining consistent at 0.66 across the 24-, 30-, and 37-month DBLs; median OS was not reached at the 37-month DBL. Dependent models were used to extrapolate OS as the proportional hazards assumption was upheld in all three DBLs. A log-normal distribution, indicating a long-term survival plateau, provided the best fit to observed OS in all DBLs. Mean lifetime OS for nivolumab+ipilimumab was 7.15 and 7.36 years, in the 18- and 30-month DBLs, respectively. Extrapolated 18-month DBL landmark survival at 5, 10, and 20 years was 39.9%, 23.0%, and 11.1%, which remained stable at the 30-month DBL as 40.7%, 23.9%, and 11.8%, all respectively.

CONCLUSIONS: Nivolumab+ipilimumab is the first immuno-oncology combination that has shown significant and stable long-term OS benefit in 1L RCC compared with standard of care (sunitinib). The new DBLs confirm the robustness of the 18-month DBL results, including the choice of a log-normal distribution for extrapolation.