Cabozantinib (C) in combination with nivolumab (N) and ipilimumab (I) in previously untreated advanced renal cell carcinoma (aRCC): Final results of COSMIC-313.
Laurence Albigès, Robert J. Motzer, Sergio E. Trevino, Ravindran Kanesvaran, Piotr Centkowski, Melissa A. Reimers, Juan Pablo Sade, Damien Pouessel, Elisa Biscaldi, Emilio Esteban, Jóse Ángel Arranz Arija, Scott S. Tykodi, Haijun Ma, Lei Zhou, Maximiliano A.G. Van Kooten Losio, Andrew D. Simmons, Fatima Rangwala, David A. Braun, Toni K. Choueiri, Thomas Powles
Abstract
438 Background: In the randomized, double-blind, phase 3 COSMIC-313 study (NCT03937219), C+N+I significantly improved PFS compared with N+I in first-line intermediate or poor risk aRCC, meeting the primary endpoint (Choueiri et al. N Engl J Med 2023). Here, we present the secondary endpoint of OS, updated efficacy and safety results, and biomarker analyses. Methods: Patients (pts) with previously untreated, IMDC intermediate or poor risk aRCC were randomized to receive C 40 mg QD or placebo (P). Both groups received N (3 mg/kg IV Q3W) + I (1 mg/kg IV Q3W) for 4 cycles, followed by N (480 mg IV Q4W) for up to 2 y. The primary endpoint was PFS by blinded independent central review per RECIST 1.1 in the first 550 randomized pts (previously reported). The secondary endpoint was OS in all randomized pts. A random forest model was used to identify immune subsets (deconvoluted from RNA-Seq data) associated with improved OS with C+N+I vs P+N+I. Results: A total of 855 pts were randomized to C+N+I (n=428) or P+N+I (n=427); IMDC risk was intermediate for 75% and poor for 25%. At a median follow-up of 45.0 months, an improvement in PFS with C+N+I was maintained (Table). OS was not significantly different between C+N+I and P+N+I in the ITT population or by IMDC risk group. ORR was higher with C+N+I, with a lower incidence of PD as best response. Grade 3/4 treatment-emergent AEs (TEAEs) occurred in 81% (C+N+I) vs 62% (P+N+I); most common grade 3/4 TEAEs were increased ALT (27% vs 6%) and increased AST (20% vs 5%). Grade 5 treatment-related AEs (TRAEs) occurred in 1% of pts in each group. No significant differences in OS outcomes were observed based on baseline c-Met or PD-L1 levels. Exploratory biomarker analyses showed that higher M2 macrophage abundance was associated with improved OS with C+N+I (HR, 0.51; 95% CI, 0.31–0.86), poor risk (per IMDC), high baseline sum of target lesions, and the presence of visceral metastasis. Biomarker analysis of angiogenic and immune signatures is ongoing. Conclusions: First-line treatment with C+N+I continued to demonstrate PFS and ORR benefit over P+N+I for pts with intermediate or poor risk aRCC. OS was comparable between the two arms, and no new safety signals emerged.Pts whose tumors have high M2 macrophage abundance had improved OS with C+N+I treatment. Clinical trial information: NCT03937219 . C+N+I (n=428) P+N+I (n=427) Median OS (95% CI), mo 41.9 (34.8–47.9) 42.0 (34.9–53.1) HR (95% CI); P- value 1.02 (0.85–1.23); P= 0.84 Median PFS (95% CI), mo 16.6 (14.0–22.6) 11.2 (9.3–14.0) HR (95% CI) 0.82 (0.69–0.98) ORR (95% CI), % 46 (41–51) 37 (32–41) Complete response, % 4 3 Partial response, % 42 33 Stable disease, % 40 36 Progressive disease, % 8 20