Impact of number of treatment lines following first-line (1L) immuno-oncology (IO) combination on overall survival (OS) in patients with metastatic renal cell carcinoma (mRCC).

Abstract

673 Background: Across the world, treatment of patients with mRCC is heterogeneous with different access to treatment sequences and number of lines of therapy (LOTs) employed. For instance, patients receiving first line (1L) nivolumab+ipilimumab (NIVO+IPI) may be offered second-line (2L) and third line (3L) vascular endothelial growth factor receptor targeted kinase inhibitor (VEGFR-TKI), whereas patients receiving 1L pembrolizumab or avelumab in combination with axitinib (IO+AXI) may only receive one subsequent VEGFR-TKI in 2L. We aimed to examine whether these different treatment strategies impact overall survival (OS). Methods: Adult mRCC patients who received at least three LOTs starting with 1L NIVO+IPI or at least two LOTs starting with 1L IO+AXI from the International Metastatic RCC Database Consortium (IMDC) centers were included. Kaplan-Meier analyses were used to estimate median OS (time from 1L to death). Results were stratified by 1L IMDC prognostic risk. Results: Among 128 patients who received at least three LOTs starting with 1L NIVO+IPI (median age 61 years, 77% White, 77% male, 37% from the US), 14% had favorable, 61% had intermediate, and 26% had poor IMDC risk. The most common 2L treatments following 1L NIVO+IPI were sunitinib (38%), cabozantinib (27%), and pazopanib (20%). Among 104 patients who received at least two LOTs starting with 1L IO+AXI (median age 62 years, 75% White, 67% male, 38% from the US), 28% had favorable, 48% had intermediate, and 25% had poor IMDC risk. The most common 2L treatments following 1L IO+AXI were cabozantinib (57%) and sunitinib (10%). Median OS are presented in the table, which suggested no difference in survival for patients who received at least two LOTs starting with 1L IO+AXI compared to patients who received at least three LOTs starting with 1L NIVO+IPI. Conclusions: Treatment for patients with mRCC varies depending on the 1L regimen chosen and by country. Our results demonstrate that, even with potential guaranteed time bias and IMDC imbalances, there is no statistically significant difference in OS for patients who received at least three LOTs starting with 1L NIVO+IPI and patients who received at least two LOTs starting with 1L IO+AXI, suggesting that selecting effective treatments in 1L resulting in fewer LOTs may have similar clinical outcomes as multiple LOTs. [Table: see text]