Assessment of Minimal Residual Disease in a Phase 1b Study of Once-Weekly Carfilzomib Combined with Lenalidomide and Dexamethasone in Patients with Multiple Myeloma

Introduction: While minimal residual disease (MRD) negativity is not yet an established regulatory surrogate for a clinical endpoint in multiple myeloma (MM), it does have value as a prognostic biomarker and in assessing disease status. Previously we reported results from a phase 1b trial that described the safety and efficacy of the triplet regimen carfilzomib (given weekly), lenalidomide, and dexamethasone (KRd) in patients with relapsed and/or refractory MM (RRMM) or newly diagnosed MM (NDMM) (Biran et al, Am J Hematol 2019;94:794-802; Alsina et al, Clin Lymphoma Myeloma Leuk 2019;19:Suppl E52). Here, we evaluate MRD status by flow cytometry after treatment with weekly KRd in this trial.

Methods: A total of 56 patients with RRMM and 51 with NDMM were enrolled and treated with weekly KRd. Treatment was given in 28-day cycles. Carfilzomib was given on days 1, 8, and 15; lenalidomide 25 mg on days 1-21; and dexamethasone 40 mg on days 1, 8, and 15 (also day 22 for cycle 1-8). In the NDMM cohort, patients were enrolled regardless of transplant eligibility, and treatment interruption for mobilization and collection with or without autologous stem cell transplant was allowed after cycle 4. Per the protocol, aspirate was collected for flow cytometry evaluation at two separate endpoints: at cycle 8 day 1 (C8D1); and at the time when laboratory data supported a response of complete response (CR) or better (on the basis of negative immunofixation on serum and urine). Sample quality was assessed by several measures (including the presence of mast cells, erythroid precursors, and immature B cells), and viability was determined using a flow assay. Adequate samples were tested for MRD using 8-color flow cytometry in two tubes, with a sensitivity of 10^-5. MRD was reported as positive when a minimum of 2 x 10⁶ CD138+ cells per tube were evaluated, and ≥ 20 abnormal events detected.

Results: Overall, 58 of 79 (73.4%) study-specified samples were obtained and evaluated by flow cytometry. Of the NDMM patients, 15 had bone marrow aspirate evaluable for MRD at C8D1 (in total 22 patients were treated to C8D1), and 12 had bone marrow aspirate evaluable for MRD at the time of CR (14 patients achieved CR). At C8D1, 53% (8/15) of NDMM patients with flow data achieved MRD negativity, and at time of CR, 83% (10/12) of NDMM patients were MRD negative (the remaining patients with flow data were determined to be MRD positive). Among the 12 NDMM patients evaluated for MRD at time of CR, 17% had high-risk cytogenetics, 58% had standard-risk cytogenetics, and 25% had unknown cytogenetic risk status. Of the RRMM patients, 26 had bone marrow aspirate evaluable for MRD at C8D1 (30 patients were treated to C8D1), and five had bone marrow aspirate evaluable for MRD at the time of suspected CR or better (in total 13 patients achieved CR). Of RRMM patients with available flow data at time of CR, 40% (2/5) achieved MRD negativity, and of those with flow data at C8D1, 50% (13/26) achieved MRD negativity. Of all patient samples evaluated for MRD, 67% of NDMM and 48% of RRMM samples were MRD negative. The results reported here are directionally comparable to the rates of MRD-negative CRs observed in previous studies of twice-weekly KRd in NDMM (Jasielec J et al, Blood 2014;124:2127; Kazandjian D et al, JAMA Oncol 2018;4:1781-1783; Zimmerman T et al, Blood 2016;128:675). Our findings are limited by incomplete acquisition of samples to support a full MRD analysis per protocol (approximately 73% of the intended time point samples were acquired). Additionally, we reported MRD status without censoring for missing MRD data, as would be required to analyze MRD for a randomized controlled trial to eliminate acquisition bias. A robust comparison of MRD-negative CR rates between regimens would require an RCT with full MRD sampling, and sensitivity analyses that treat missing MRD data as MRD-positive (Chari A et al, Blood 2017;130:974-981; Voorhees PM et al, Blood 2020).

Conclusions: We have previously shown that once-weekly KRd is active and has acceptable toxicity in both the RRMM and NDMM settings. We found that among MRD-evaluable patients who had a CR or better, MRD negativity rates were impressive in both the NDMM setting (83%) and in the RRMM setting (40%), suggesting that the weekly KRd regimen can induce MRD-negative CRs in both settings.