Real-world comparative effectiveness of triplets containing bortezomib (B), carfilzomib (C), daratumumab (D), or ixazomib (I) in relapsed/refractory multiple myeloma (RRMM) in the US
Faith E. Davies, Robert M. Rifkin, Caitlin Costello, Gareth J. Morgan, Saad Z. Usmani, Rafat Abonour, Antonio Palumbo, Dorothy Romanus, Roman Hájek, Evangelos Terpos, Dasha Cherepanov, Dawn Marie Stull, Hui Huang, Xavier Leleu, Jesús G. Berdeja, Hans C. Lee, Katja Weisel, Michael A. Thompson, Mario Boccadoro, Jeffrey A. Zonder, Gordon Cook, Noemí Puig, Jorge Vela‐Ojeda, Eileen Farrelly, Aditya Raju, Marlo Blazer, Ajai Chari
Abstract
Multiple available combinations of proteasome inhibitors, immunomodulators (IMIDs), and monoclonal antibodies are shifting the relapsed/refractory multiple myeloma (RRMM) treatment landscape. Lack of head-to-head trials of triplet regimens highlights the need for real-world (RW) evidence. We conducted an RW comparative effectiveness analysis of bortezomib (V), carfilzomib (K), ixazomib (I), and daratumumab (D) combined with either lenalidomide or pomalidomide plus dexamethasone (Rd or Pd) in RRMM. A retrospective cohort of patients initiating triplet regimens in line of therapy (LOT) ≥ 2 on/after 1/1/2014 was followed between 1/2007 and 3/2018 in Optum's deidentified US electronic health records database. Time to next treatment (TTNT) was estimated using Kaplan-Meier methods; regimens were compared using covariate-adjusted Cox proportional hazard models. Seven hundred forty-one patients (820 patient LOTs) with an Rd backbone (VRd, n = 349; KRd, n = 218; DRd, n = 99; IRd, n = 154) and 348 patients (392 patient LOTs) with a Pd backbone (VPd, n = 52; KPd, n = 146; DPd, n = 149; IPd, n = 45) in LOTs ≥2 were identified. More patients ≥75 years received IRd (39.6%), IPd (37.8%), and VRd (36.7%) than other triplets. More patients receiving VRd/VPd were in LOT2 vs other triplets. Unadjusted median TTNT in LOT ≥ 2: VRd, 13.9; KRd, 8.7; IRd, 11.4; DRd, not estimable (NE); and VPd, 12.0; KPd, 6.7; IPd, 9.5 months; DPd, NE. In covariate-adjusted analysis, only KRd vs DRd was associated with a significantly higher risk of next LOT initiation/death (HR 1.72; P = 0.0142); no Pd triplet was significantly different vs DPd in LOT ≥ 2. Our data highlight important efficacy/effectiveness gaps between results observed in phase 3 clinical trials and those realized in the RW.