Combination therapy with nivolumab and tyrosine kinase inhibitors in patients with metastatic renal cell carcinoma.

e17090 Background: Two immunotherapy (IO) and tyrosine kinase inhibitor (TKI) combinations are FDA-approved for patients (pts) with metastatic renal cell carcinoma (mRCC). Here we present a multicenter off-protocol experience with IO/TKI combinations after progression on monotherapy. Methods: We performed a retrospective analysis of pts with mRCC who received combination non-FDA approved off-protocol IO/TKI combination therapy from 11/2015 – 1/2019 at MD Anderson Cancer Center and Duke Cancer Institute. Results: 48 pts met criteria for study inclusion. At therapy start: median (med) age 65 years; 75% clear cell histology; 68.8% IMDC intermediate risk (20.8% favorable, 10.4% poor); 81.3% prior nephrectomy; med metastatic sites: 2; med prior systemic treatments: 2; most common metastatic sites: lung (58.3%), lymph node (52.1%), and bone (43.8%). Pts received nivolumab (nivo) in combination with the following: cabozantinib (n = 24, 50%), pazopanib (13, 27.1%), axitinib (6, 12.5%), lenvatinib (2, 4.2%), ipilimumab/cabozantinib (3, 6.3%). Med PFS was 13.7 months and med OS was not reached. The two largest cohorts received nivo + cabozantinib (N+C; med dose 40 mg daily) or nivo + pazopanib (N+P; med dose 400 mg daily). The N+C cohort had higher med metastatic sites (3 vs 2) and was more pretreated with agents unique to their IO/TKI combination (med 2 vs 0). In the N+P group, more pts had started on TKI prior to addition of nivo at progression (69.2% vs 45.8%), and fewer had IO monotherapy with TKI addition (30.8% vs 50%). With a med follow up of 14.0 months after combination start, the N+C cohort had a med PFS of 7.3 months (initiated TKI first: 4.8, IO first: 8.2) and med OS of 18.2 months (TKI first: 11.8, IO first: 24.3). The N+P cohort had med follow up of 20.5 months after combination, med PFS of 21.3 months (TKI first: 16.5, IO first: 21.8), and med OS was not reached. In the N+C group, 87.5% experienced any grade adverse event (AE), most common included fatigue (79.2%), diarrhea (42.7%), nausea (29.2%), hypertension (29.2%), and weight loss (25.0%). In the N+P group, 92.3% experienced any grade AE, including fatigue (76.9%), hypertension (38.5%), diarrhea (30.8%), nausea (30.8%), and weight loss (30.8%). There were no grade ≥3 AEs. Conclusions: Slow disease progression on nivo or TKI may be safely controlled with addition of IO/TKI therapy. Med PFS after addition of nivo to TKI appears similar (N+C) and improved (N+P) compared to nivo monotherapy (Checkmate-025). Med PFS after addition of TKI to IO was also similar (N+C) and improved (N+P) compared to historical controls.