KRd Exhibits PFS Advantage vs VRd in Newly Diagnosed Multiple Myeloma

KRd reduced the risk of progression or death by 43% vs VRd in the intent-to-treat population at a median follow-up of 53 months; the median PFS was not reached in the KRd group vs 48.8 months in the VRd arm.

The addition of Carfilzomib (Kyprolis) to lenalidomide (Revlimid) and dexamethasone (KRd) significantly improved progression-free survival (PFS) vs bortezomib (Velcade) plus lenalidomide and dexamethasone (VRd) among patients with newly diagnosed multiple myeloma, according to interim results from the phase 3 COBRA trial (NCT03729804) presented at the 2025 American Society of Hematology (ASH) Annual Meeting and Exposition.1

Efficacy findings demonstrated that KRd (n = 126) reduced the risk of progression or death by 43% vs VRd (n = 124) in the intent-to-treat population (HR, 0.57; 95% CI, 0.37-0.88; P = .0095) at a median follow-up of 53 months. The median PFS was not reached (NR) in the KRd group vs 48.8 months in the VRd arm.

The data demonstrated consistent directional benefits across cytogenetic subgroups. Among patients with standard-risk disease, KRd (n = 97) achieved a statistically significant improvement in PFS vs VRd (n = 96; HR, 0.59, 95% CI, 0.36-0.98; P = .04); the median PFS was NR in the KRd group compared with 48.8 months for VRd. In this subgroup, 27% of patients treated with KRd experienced progression or death vs 40% with VRd. In the high-risk cohort, KRd (n = 29) again demonstrated favorable outcomes vs VRd (n = 28), with 31% vs 48% of patients experiencing progression or death. The median PFS was NR vs 34.9 months, respectively (HR, 0.52; 95% CI, 0.22-1.22; P = .12).

“COBRA showed superior efficacy of KRd vs VRd in newly diagnosed multiple myeloma, achieving both co-primary endpoints of MRD-negative CR at 12 months andPFS,” noted presenting author Dominik Dytfeld, MD, PhD. “The PFS benefit of KRd was observed regardless of cytogenetic risk, and KRd produced deeper responses, with higher rates of complete response [CR] and minimal residual disease [MRD]–negativity. KRd [also] demonstrated anticipated toxicity profiles with higher rates of neutropenia and cardiac [adverse effects (AEs)], but less neuropathy. COBRA results support further evaluation of KRd-based induction regimens in newly diagnosed multiple myeloma.”

Dytfeld is an associate professor of medicine at the Poznan University of Medical Sciences in Poland and founder/chief executive officer of the Polish Myeloma Consortium

What was the design of the COBRA trial?

The COBRA trial was a multicenter, randomized, open-label phase 3 study designed to compare KRd with VRd in patients with newly diagnosed multiple myeloma who had an International Myeloma Working Group (IMWG) Frailty Score of less than 2.1,2 Randomization was stratified by cytogenetic risk category (standard vs high risk defined by t(4;14) or del(17p)) and by history of venous thromboembolism (yes vs no). A total of 250 patients were randomly assigned in a 1:1 ratio to the KRd or VRd treatment arms.1

Patients in the KRd arm received up to 24, 28-day cycles of therapy consisting of carfilzomib at 56 mg/m² administered on days 1, 8, and 15 (with twice-weekly dosing during cycles 1 and 2), lenalidomide at 25 mg on days 1 through 21, and dexamethasone at 40 mg weekly (or 20 mg for patients 75 years of age or older). Stem-cell collection occurred after cycle 4 in this transplant-deferred design. Following completion of induction, patients continued on lenalidomide 5 mg daily as maintenance until disease progression.

Patients in the VRd arm received eight 28-day cycles of bortezomib at 1.3 mg/m² on days 1, 4, 8, and 11; lenalidomide at 25 mg on days 1 through 14; and dexamethasone at 20 mg on days 1, 2, 4, 5, 8, 9, 11, and 12. After completion of the VRd induction phase, patients transitioned to consolidation with lenalidomide and dexamethasone for eighteen 28-day cycles, followed by lenalidomide 5 mg maintenance until progression.

The trial’s co-primary end points were the rate of MRD-negative CR rate at a sensitivity of 10⁻⁵ at 12 months and PFS. Secondary end points include MRD rate, sustained MRD negativity, overall response rate (ORR), overall survival (OS), and safety.

How did KRd and VRd compare in transplant-eligible vs transplant-ineligible patients?

The subgroup analysis by transplant eligibility showed distinct patterns in PFS favoring KRd among patients considered eligible for autologous stem cell transplant. In the transplant-eligible cohort, KRd reduced the risk of progression or death by 60% relative to VRd (HR, 0.40; 95% CI, 0.21-0.75; P = .003). Only 23% of patients treated with KRd experienced progression or death compared with 45% in the VRd arm. Median PFS was NR with KRd, whereas VRd demonstrated a median PFS of 40.1 months.

In contrast, outcomes were comparable between the two regimens in the transplant-ineligible population. In this subgroup, both KRd and VRd produced identical progression or death rates of 30%. Median PFS remained NR with KRd vs 52.9 months for VRd, resulting in no statistically significant difference between the arms (HR, 1.06; 95% CI, 0.21-0.75; P = .87).

What were the safety outcomes of the COBRA trial?

The safety profile of KRd and VRd in the COBRA trial demonstrated that both regimens were associated with high rates of AEs, although specific toxicity patterns differed between the treatment arms. Grade 3 or higher AEs occurred in 73% of patients receiving KRd and 62% of those receiving VRd, while any-grade AEs were nearly universal across both cohorts (96% with KRd and 94% with VRd). Treatment discontinuation due to AEs occurred in 11% of KRd-treated patients and 8% of VRd-treated patients.

Grade 5 AEs were infrequent but observed in both arms, occurring in 5 patients (4%) in the KRd arm and 7 patients (6%) in the VRd arm. Reported fatal AEs in the KRd cohort included COVID-19, stroke, pneumonia, sepsis, and acute kidney failure, whereas the VRd arm included deaths from COVID-19, pneumonia, respiratory failure, and two AEs of unknown cause. Neutropenia of any grade occurred in 29% of KRd-treated patients and 17% of VRd-treated patients, with grade 3 or higher neutropenia reported in 21% and 11%, respectively.

Neuropathy, a known bortezomib-associated toxicity, was more common with VRd: 56% of patients experienced neuropathy of any grade compared with 17% in the KRd arm, although grade 3 or higher neuropathy remained low in both groups (2% in each arm). Cardiac AEs occurred more frequently with KRd (18% any grade; 6% grade ≥3) than with VRd (10% any grade; 2% grade ≥3), consistent with the known cardiovascular risk profile of carfilzomib. Infection rates were high in both groups but were more common in the KRd arm, where 75% experienced infections of any grade and 25% developed grade 3 or higher infections, compared with 60% and 23% in the VRd arm, respectively.

References

1. Dytfeld D, Kubicki T, Tyczynska A, et al. Carfilzomib, lenalidomide, and dexamethasone (KRd) versus bortezomib, lenalidomide, and dexamethasone (VRd) in patients with newly diagnosed multiple myeloma: results of the randomized phase III COBRA trial. Blood. 2025;146(suppl 1):99. doi:10.1182/blood-2025-99
2. Carfilzomib, lenalidomide, and dexamethasone versus bortezomib, lenalidomide and dexamethasone (KRd vs. VRd) in patients with newly diagnosed multiple myeloma (COBRA). ClinicalTrials.gov. Updated October 3, 2025. Accessed December 6, 2025. https://tinyurl.com/3adcfx9j