Diffuse large B-cell lymphoma (DLBCL) is typically managed with rituximab-based chemoimmunotherapy; however, patients with high-risk disease continue to have suboptimal outcomes. Bispecific antibodies targeting CD20×CD3 are being explored in earlier lines of therapy to enhance response depth and improve disease control in this setting. The phase 3 OLYMPIA-3 study (NCT06091865; part 1B) evaluated the safety and preliminary efficacy of odronextamab in combination with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) as first-line treatment for patients with previously untreated CD20-positive DLBCL or high-grade B-cell lymphoma with at least 1 high-risk feature.

In this open-label study, patients aged ≥18 years were randomized 1:1 to receive odronextamab 160 mg with 2 different dosing schedules in combination with CHOP for 6 cycles:160 mg weekly followed by 320 mg every 2 weeks (regimen [R] 1) or 160 mg weekly followed by 160 mg every 3 weeks (R2). The primary endpoint was safety, and secondary endpoints included investigator-assessed objective response rate (ORR) per Lugano criteria, complete response (CR) rate, and duration of response. Additionally, minimal residual disease by circulating tumor DNA (ctDNA) was an exploratory endpoint.

At data cutoff on April 6, 2026, 40 patients were enrolled (20 per arm), with a median age of 67.5 years. Most patients (75.0%) had high-risk disease, defined by an International Prognostic Index score of 3 to 5. Treatment exposure was comparable across groups, with a median duration of approximately 18 weeks in R1 and 16.6 weeks in R2, and 67.5% of patients completed all 6 planned cycles.

The safety profile was generally manageable and similar across dosing regimens. The most common treatment-emergent adverse events (AEs) included neutropenia (R1: 65% vs R2: 40%), cytokine release syndrome (CRS; R1: 50.0% vs R2: 60%), and anemia (R1: 50% vs R2: 40.0%). CRS was largely grade 1 in severity (R1: 35% vs R2: 50%), and immune effector cell–associated neurotoxicity syndrome occurred in 3 patients, all grade 1 and fully resolved. Treatment-related AEs led to dose interruptions or delays in approximately 65% and 50% of patients and CHOP dose reductions in 15% and 5% of patients in R1 and R2, respectively. One treatment discontinuation due to grade 4 neutropenic sepsis was reported.

Despite relatively short follow-up (median, 7.9 months in R1; median, 8.0 months in R2), high response rates were observed across both dosing schedules. The ORR was 95.0% in R1 and 90.0% in R2, with CR rates of 85.0% in both groups, indicating robust early antitumor activity in this high-risk, frontline population. Median duration of complete response was not reached. Both regimens achieved rapid decreases in ctDNA, with no meaningful differences in ctDNA undetectability between regimens at the end of treatment and a 90-day follow-up.

These findings indicate that odronextamab can be safely combined with CHOP and produces high early response rates in untreated DLBCL. Longer follow-up and ongoing randomized evaluation will be important to assess the durability of responses and confirm clinical benefit in this setting.

Source: Matasar MJ, et al. First-line odronextamab plus CHOP in DLBCL: results from the OLYMPIA-3 part 1B study. Presented at: ASCO Annual Meeting. May 29-June 2, 2026; Chicago, IL. Abstract 7009.