e16549 Background: Numerous immunotherapy (IO) and vascular endothelial growth factor receptor inhibitors (VEGFI) options are approved for first-line (1L) treatment of mRCC. We performed a systematic review and meta-analysis to compare the oncologic outcomes and adverse events of these therapies. Methods: We searched databases for randomized controlled trials (RCT) evaluating agents that are currently FDA-approved or on the NCCN compendium for 1L mRCC, with sunitinib (S) as a comparator. Primary outcomes were overall survival (OS), progression-free survival (PFS), overall response rate (ORR), and grade 3 and 4 (G3/G4) adverse events (AEs). We attempted network meta-analysis (NMA) to estimate relative effects of the comparative treatments. Survival outcomes were summarized using hazard ratios (HR). In the case that the HR was not reported but a Kaplan-Meier curve was available, the curve was digitized using Engauge Digitizer software. A frequentist fixed-effects NMA model was applied to each outcome. Results: 798 articles were identified, of which 47 were eligible for full text screening, and 8 were included for evidence synthesis. Eight RCTs recruiting a total of 5,565 pts with advanced or metastatic treatment-naïve RCC were included. The summary of all included studies is reported in Table 1. Ordered from the most to least effective, treatments with greatest mOS benefit include Nivolumab plus Cabozantinib (NC), Pembrolizumab plus Axitinib (PA), and Nivolumab plus Ipilimumab (NI); treatments with greatest mPFS benefit include Cabozantinib (C), NC, and Avelumab plus Axitinib (AA); treatments with the highest ORR include C, NC, and AA. Ordered from least to most toxic, treatments with G3/G4 AEs less frequent than sunitinib include Pazopanib (P), NI, and AA. Conclusions: OS generally favored IO-IO or IO-VEGFI combinations, while PFS and ORR favored single agent C or IO-VEGFI. The lowest risks of G3/G4 AE were seen with P or IO-IO. IO-VEGFI combinations had similar risks for G3/G4 AE. Longer real-world follow-up is needed. We recognize the limitations of cross-trial comparisons, but sought a summative view of the current data.[Table: see text]
Read full abstract