Abstract
Introduction
Odronextamab (Odro), a CD20×CD3 bispecific antibody, has shown compelling efficacy as monotherapy in patients (pts) with relapsed/refractory (R/R) DLBCL in the Phase 2 ELM-2 study (NCT03888105). OLYMPIA-3 (NCT06091865), a Phase 3, randomized, open-label, multicenter study, is the first to investigate whether Odro + chemotherapy (CHOP) is superior to standard-of-care rituximab + CHOP in pts with previously untreated DLBCL (Part 2). The study consists of Part 1A (dose escalation) and Part 1B (dose optimization). Here we report the first results from Part 1A.
Methods
Part 1A included pts aged ≥18 years with untreated CD20+ DLBCL not otherwise specified or high-grade B-cell lymphoma with MYC, BCL-2, and/or BCL-6 rearrangements, an International Prognostic Index (IPI) score of ≥2. Odro-CHOP was administered in 6 × 21-day cycles. Odro was administered intravenously with weekly step-up dosing (0.7/4/20 mg) starting on Cycle (C)1 Day (D)8, followed by full doses on C2 (D8 and 15), C3–4 (D1, 8, and 15), and C5D1, and 2 × full dose on C5D8 and C6 (D1 and 15). Odro full doses of 80 mg (dose level 1 [DL1]) and 160 mg (DL2) were tested. Part 1A primary endpoints were the incidence of dose-limiting toxicities (DLTs) during the DLT observation period, and the incidence and severity of treatment-emergent adverse events (TEAEs). Secondary endpoints included objective response rate (ORR), complete response (CR) rate, and duration of response (DOR) assessed by local investigator per Lugano response criteria. Biomarker analyses were exploratory endpoints.
Results
At the data cutoff (May 5, 2025), 22 pts with DLBCL were enrolled in Part 1A and were evaluable for efficacy and safety (DL1, n=9; DL2, n=13). Median age was 66 years (range
24–81), 11 pts were male, 20 pts had an IPI score of 2–5, and one pt had an IPI score of 1. The median duration of follow-up was 9.0 months (95% CI 6.5–not evaluable [NE]) for DL1 and 4.5 months (2.8–6.0) for DL2. In the DL1 and DL2 groups, 77.8% and 92.3% of pts completed 6 cycles of induction, with median durations of treatment exposure of 18.1 weeks (range 1.6–21.9) and 18.9 weeks (16.3–22.7), respectively. Fifteen pts (68.2%) were considered to have completed treatment in Part 1A following initiation of C6. DLTs were evaluable in 18 pts (DL1, n=6; DL2, n=12). No pts experienced a DLT during the DLT observation period. All 22 pts experienced ≥1 TEAE, which led to treatment interruption/delay in 17 pts (DL1, n=6; DL2, n=11) and to treatment discontinuation in 2 pts (myocardial infarction [DL1], and septic shock [DL2]). The most common TEAEs were neutropenia (DL1, 77.8%; DL2, 84.6%; all Grade [Gr] ≥3), cytokine release syndrome (CRS; 33.3%; 76.9%), and anemia (33.3%; 53.8%). Gr ≥3 TEAEs occurred in all pts, with neutropenia being the most common. Infections occurred in 66.7% of pts at DL1 and 92.3% of pts at DL2; Gr ≥3 infection rates were 22.2% and 61.5%, respectively. CRS events were all Gr 1 (DL1, 11.1%; DL2, 69.2%) or Gr 2 (22.2%; 7.7%). No tumor lysis syndrome or immune effector cell-associated neurotoxicity syndrome events were reported.
For DL1, ORR was 77.8% (95% CI 40.0–97.2), with a CR rate of 66.7% (95% CI 29.9–92.5). The ORR and CR rate in the DL2 group were both 100% (95% CI 75.3–100). Among 20 evaluable pts, median DOR was not reached for both DL1 and DL2 (95% CIs, 2.5–NE and
2.9–NE, respectively), with 6-month event-free probabilities of 85.7% (95% CI 33.4–97.9) and 90.0% (95% CI 47.3–98.5).
Serum cytokines IL-6, IL-8, IFNγ, and TNFα were elevated with the first Odro-CHOP doses only, and T-cell margination was similar to that previously reported with Odro monotherapy. After the C2D8 first full dose, cytokine and T-cell counts were comparable between DL1 and DL2. B-cell counts decreased during Week 1 of CHOP treatment and cleared completely after the first dose of Odro-CHOP. No B-cell recovery was observed ≤90 days after end of treatment. Baseline CD20 status and on-treatment ctDNA clearance will be presented.
Conclusions
In Part 1A of the Phase 3 OLYMPIA-3 study, the safety profile of Odro-CHOP induction was generally manageable, with no new safety signals. Preliminary efficacy of this simplified regimen was encouraging, with a CR rate of 100% for DL2, suggesting that rituximab may not be required to achieve depth of response when combining Odro with CHOP in previously untreated pts with DLBCL. Updated Part 1A data will be presented.
