Molecular analysis of the HCRN GU16-260-Cohort A phase II study of first-line (1L) nivolumab (nivo) and salvage nivo + ipilimumab (ipi) in patients (pts) with advanced clear cell renal cell carcinoma (accRCC).

Abstract

4546 Background: Seven molecular subsets of ccRCC have been proposed: 1: Angiogenic/Stromal; 2: Angiogenic; 3: Complement/Ω-oxidation; 4: T-effector/Proliferative; 5: Proliferative; 6: Stromal/Proliferative; and 7: snoRNA (Motzer et al. Can Cell 2020). In an analysis of IMmotion 151, which compared 1L atezolizumab/bevacizumab to sunitinib, combination immune checkpoint blockade (ICB) and anti-VEGF therapy was associated with improved progression-free survival (PFS) and overall survival (OS) in clusters (CLs) 4 and 5, increased PFS in CL 7, and decreased OS in CL 2 (Motzer Can Cell 2020, Motzer JAMA Onc 2022). CL 4 was also associated with higher PFS and objective response rate (ORR) in >1L nivo compared to everolimus in an analysis of CheckMate-025 (Denize CCR 2022). To date, outcomes have not been compared among accRCC subtypes in pts who received 1L pure ICB. Methods: In HCRN GU16-260-Cohort A, pts with accRCC received 1L nivo monotherapy until progressive disease (PD), unacceptable toxicity, or completion of 96 weeks of therapy. Pre-treatment FFPE tumor samples were obtained. A machine learning model (random forest) was trained on annotated transcriptomic data from bulk RNA-seq and used to classify tumors that passed quality control into one of the 7 molecular subtypes. Results: 70 tumor samples contained adequate RNA quality and were classified as outlined in the table. IMDC rates and ORR (40%) for the RNAseq population closely matched those of the whole study (Atkins JCO 2022). No tumors were classified in CL 7 and only 3 in CL 5. Tumors in each IMDC group contained a mix of subtypes. The highest ORR occurred in CL 4 (6/10, 60.0%); CL 4 median PFS was 15.2 months and 1-year PFS was 60%. ORR was also high in CLs 1 and 3 at 52.4% and 40.0%, respectively and lowest in CL 2 (12.5%). PD as best response (5/10, 50.0%) and short PFS (5.1 months) were seen in CL 6. Conclusions: Consistent with prior analyses with combination ICB + anti-VEGF and >1L ICB, CL 4 was associated with robust 1L ICB monotherapy antitumor efficacy. Surprisingly, ORR was also high in CLs 1 and 3. In this small dataset, transcriptomic signatures enriched for but did not fully predict ORR. Future prospective trials are needed to identify more discriminative biomarkers of efficacy for pure ICB therapy.[Table: see text]