AUTHORS
Paul G. Richardson, MD, Rahul Banerjee, MD2, Kevin Towle3, Amol Dhamane, BPharm4, Amir Samuel5, Jenny M. H. M.H. Chen6, Sichen Liu, MSc7, Emma Hawe8, Jessica Katz, MD, PhD9, Nisha Joseph, MD10, Joseph Mikhael, MD, MEd11
AFFILIATIONS
1 Dana-Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts, United States
2 Fred Hutchinson Cancer Center, Seattle, Washington, United States
3 Precision AQ, Vancouver, British Columbia, Canada
4 Bristol Myers Squibb, Lawrence Township, New Jersey, United States
5 Bristol Myers Squibb, Lawrence Township, New Jersey, United States
6 Precision AQ, Vancouver, British Columbia, Canada
7 Precision AQ, Vancouver, British Columbia, Canada
8 Precision AQ, Vancouver, British Columbia, Canada
9 Bristol Myers Squibb, Lawrence Township, New Jersey, United States
10 Emory University, Atlanta, Georgia, United States
11 Translational Genomics Research Institute, Seattle, Washington, United States
ABSTRACT
Background: MeziKd significantly improved progression-free survival (PFS) versus Kd in the phase III SUCCESSOR-2 trial among patients with RRMM previously exposed to lenalidomide and an anti-CD38 monoclonal antibody. The Flatiron Health database provides information on contemporary real-world (RW) standard-of-care (SoC) treatment patterns in the United States (US). To contextualize the SUCCESSOR-2 findings against RW treatment patterns, a population-adjusted indirect comparison of MeziKd versus RW SoC was conducted.
Methods: This analysis used individual-level patient data (IPD) from SUCCESSOR-2 (MeziKd; N=288) and the Flatiron Health database. Flatiron cohort was derived by applying key SUCCESSOR-2 eligibility criteria, including age, number of prior lines, prior exposure to lenalidomide and an anti-CD38 monoclonal antibody, prior malignancies, prior stem cell transplantation, laboratory abnormalities, and Eastern Cooperative Oncology Group (ECOG) performance status (PS). The Flatiron cohort was restricted to patients who initiated an eligible line of therapy between January 2022 to October 2025 (data cut off) to align more closely with the enrolment period of SUCCESSOR-2. The base-case analysis included the earliest eligible line of Food and Drug Administration (FDA)-approved or National Comprehensive Cancer Network (NCCN)-recommended treatments in RRMM (broad SoC; N=702 patients). Additional scenario analyses evaluated alternative line-selection and SoC definitions, including broad SoC with a randomly selected eligible treatment line (N=702), a broad SoC basket restricted to triplet regimens (N=465), and a broad SoC basket restricted to the ten most common regimens (N=482).
Baseline imbalances in prespecified, clinician-ranked prognostic factors and/or effect modifiers (N=9 covariates) were adjusted for via inverse probability of treatment weighting (IPTW). Missing covariate data (≤40%) were addressed using multiple imputation. The Flatiron cohort was weighted using average treatment effect on the treated methods for population adjustment. PFS treatment effects were estimated using Cox proportional hazards models, with hazard ratios (HRs) and corresponding 95% confidence intervals (CIs). Differences in restricted mean survival time (dRMST) up to various landmark timepoints (12, 24, and 30.9 months) were also estimated to provide complementary effect measures and to address potential violations in the proportional hazards assumption.
Results: Prespecified covariates for adjustment included triple-class refractory, ECOG PS, high-risk cytogenetics, age, International Staging System (ISS) stage at diagnosis, number of prior lines, platelet count, prior transplantation, and serum albumin. Baseline characteristics were well balanced between SUCCESSOR-2 and Flatiron following adjustment, with mean absolute standardized difference <20% for all covariates. The RW SoC basket included 43 unique regimens; the most common regimens were daratumumab/pomalidomide/dexamethasone (DPd; N=103 [14.7%]), daratumumab/carfilzomib/dexamethasone (DKd; N=66 [9.4%]), and carfilzomib/pomalidomide/dexamethasone (KPd; N=63 [9.0%]). The mean effective sample size was 494.6, representing a 29.5% decrease from the original cohort.
Median PFS was 18.04 months for MeziKd, 9.89 months for RW SoC (unadjusted), and 9.91 months for RW SoC (IPTW-adjusted). MeziKd was associated with statistically significant improvements in PFS relative to RW SoC in both unadjusted (HR 0.58 [95% CI: 0.47, 0.72]) and adjusted (HR 0.58 [95% CI: 0.46, 0.72]) analyses. No violations in the proportional hazards assumption were observed. MeziKd was also associated with statistically significant improvements in PFS as measured by dRMST up to 12 months (+1.53 months [95% CI: 0.46, 2.59]), 24 months (+3.66 months [95% CI: 1.97, 5.35]), and maximum follow-up of 30.9 months (+4.56 months [95% CI: 2.51, 6.61]). Statistically significant PFS benefits favoring MeziKd were observed across all alternative RW SoC definitions, with adjusted HRs ranging from 0.53 to 0.62.
Conclusions: In this population-adjusted external control arm analysis, MeziKd was associated with significantly improved PFS versus RW SoC regimens among patients with RRMM previously exposed to lenalidomide and an anti-CD38 monoclonal antibody. These findings complement the randomized SUCCESSOR-2 results and provide additional context regarding the comparative effectiveness of MeziKd within current US clinical practice.
DISCLOSURES
Paul G. Richardson, MD: Karyopharm - Research Funding; Oncopeptides - Research Funding; Bristol Myers Squibb / Celgene - Consultancy (Includes expert testimony); Sanofi - Consultancy (Includes expert testimony); Oncopeptides - Consultancy (Includes expert testimony); Regeneron - Consultancy (Includes expert testimony); GSK - Consultancy (Includes expert testimony); Karyopharm - Consultancy (Includes expert testimony)