Epcoritamab Yields Superior PFS vs Chemoimmunotherapy in Pretreated LBCL

At the European Hematology Association (EHA) 2026 Congress, Christopher P. Fox, MBChB, PhD, FRCP, FRCPath, of Nottingham University Hospitals NHS Trust, Nottingham, UK, presented primary results from the phase 3 EPCORE FL-1 trial (NCT05409066) evaluating subcutaneous epcoritamab-bysp (Epkinly) monotherapy vs investigator's choice of chemoimmunotherapy (CIT) in patients with relapsed/refractory large B-cell lymphoma (LBCL) who were ineligible for, or had relapsed after, high-dose therapy with autologous stem cell transplant (HDT-ASCT).¹

Epcoritamab met the co-primary progression-free survival (PFS) endpoint. The median PFS was 3.5 months (95% CI, 2.9–4.6) with epcoritamab vs 3.0 months (95% CI, 2.9–4.1) with CIT (stratified HR, 0.74; 95% CI, 0.60–0.92; P = .0059), reflecting a 26% reduction in the risk of progression or death. Kaplan-Meier curves showed progressive separation over time: at 12 months, 35% of epcoritamab-treated arm remained progression-free vs 18% in the CIT arm; at 24 months it was 30% vs 13%; and at 36 months it was 27% vs 8%.

The complete response (CR) rate was 38% with epcoritamab vs 26% with CIT (nominal P = .0032). Duration of complete response (DOCR) was not reached (95% CI, 32.9–NR) with epcoritamab vs 10.8 months (95% CI, 5.6–20.6) with CIT, with 59% of patients achieving a complete response with epcoritamab still in complete response at 3 years vs 24% with CIT.

Time to next treatment (TTNT) also significantly favored epcoritamab, with a median of 6.6 months (95% CI, 4.5–13.3) vs 4.3 months (95% CI, 3.7–5.2) for CIT (stratified HR, 0.62; 95% CI, 0.50–0.78; P < .0001). Reflecting higher disease progression in the CIT arm, patients there were considerably more likely to go on to receive CAR T-cell therapy (11% vs 3%), CD3×CD20 bispecific antibody therapy (14% vs 1%), or stem cell transplant (SCT; 6% vs 2%), accounting for 31% of CIT-arm patients vs 6% in the epcoritamab arm receiving these subsequent therapies.

In the subgroup of patients enrolled in the second-line setting, epcoritamab showed numerically favorable PFS, with a median of 6.0 months (95% CI, 3.0-38.7) vs 5.6 months (95% CI, 3.0-7.9) with CIT (unstratified HR, 0.67; 95% CI, 0.44–1.04). Additionally, for overall survival (OS), the median was 40.2 months (95% CI, 13.8-NR) vs 17.8 months (95% CI, 8.4-30.9), respectively (unstratified HR, 0.80; 95% CI, 0.51–1.25), with a CR rate of 45% vs 38% in the CIT arm. Epcoritamab has also demonstrated activity in relapsed/refractory follicular lymphoma in combination regimens, extending the clinical evidence base for this bispecific antibody across B-cell malignancies.²

The co-primary OS endpoint was not met. The median OS was 11.3 months (95% CI, 7.8–15.7) with epcoritamab vs 10.3 months (95% CI, 9.0–12.5) with CIT (stratified HR, 0.96; 95% CI, 0.77–1.20; P = .7285). Fox attributed this to 2 compounding factors: COVID-19 mortality, which accounted for 22 deaths (9%) in the epcoritamab arm vs 6 deaths (3%) in the CIT arm, a consequence of enrollment during the Omicron phase; and the substantially higher use of effective subsequent therapies in the CIT arm, which contributed to 25% of OS in CIT-arm patients vs 5% in the epcoritamab arm.

In a post hoc analysis adjusting for both COVID-19 mortality and receipt of subsequent therapies, the median OS was 16.7 months (95% CI, 10.9–26.3) with epcoritamab vs 9.5 months (95% CI, 8.7–11.7) with CIT (stratified HR, 0.76; 95% CI, 0.59–0.99).

Investigators in the EPCORE DLBCL-1 trial enrolled 483 patients across 166 sites in 24 countries with a data cutoff of October 13, 2025, and median follow-up of 43.2 months in the epcoritamab arm vs 42.3 months in the CIT arm. Eligible patients had CD20-positive R/R LBCL—including de novo diffuse large B-cell lymphoma (DLBCL), transformed disease, double-hit/triple-hit DLBCL, follicular lymphoma grade 3B, and T-cell/histiocyte-rich LBCL—with at least 1 prior systemic line of therapy and ECOG performance status of 0 to 2. Enrollment of patients with 1 prior line of therapy was capped at 25% to ensure robust representation of the third-line-plus population.

Patients were randomly assigned 1:1 to subcutaneous epcoritamab at 48 mg on a 28-day cycle schedule, weekly in cycles 1–3, biweekly in cycles 4–9, then every 4 weeks from cycle 10 onward; (n = 241) or investigator's choice of R-GemOx (n = 174) or bendamustine plus rituximab (BR; n = 68). Dual co-primary endpoints were PFS and OS per independent review committee (IRC) assessment using Lugano criteria, with only 1 co-primary endpoint required to be met in the EU regulatory context. Notably, two-thirds of patients were enrolled during the Omicron wave of the COVID-19 pandemic.

The enrolled population was older and heavily pretreated: median age was 72.0 years (range, 31–89) in the epcoritamab arm and 71.0 years (range, 28–88) in the CIT arm, with 77% of patients aged 65 years or older in the epcoritamab arm. The majority of patients had de novo DLBCL (81% vs 79%) and Ann Arbor stage III/IV disease (77% vs 78%). Regarding treatment history, approximately 72% of patients in each arm had received 2 or more prior lines of therapy; 52% and 55% were primary refractory; 15% and 14% had prior ASCT; and 11% in each arm had prior CAR T-cell therapy. Median time from initial diagnosis to randomization was 16.3 months in the epcoritamab arm and 17.1 months in the CIT arm.

Mean treatment duration was 11.2 months in the epcoritamab arm vs 2.1 months in the CIT arm; 73 patients (31%) in the epcoritamab arm completed or initiated 12 or more cycles, and 28 patients (12%) remained on treatment at the data cutoff. When adjusted for the substantially longer exposure duration with epcoritamab, the safety profile was favorable. Grade 3 to 4 treatment-emergent adverse events (TEAEs) occurred in 76% of the epcoritamab arm vs 79% of CIT arm; the exposure-adjusted event rate (EAER) was 20 vs 85.7 per 100 patient-months, respectively. Serious TEAEs were reported in 69% of the epcoritamab arm and 32% of the CIT arm.

Cytokine release syndrome (CRS) occurred in 53% of epcoritamab-treated patients, with grade 1–2 occurring in 94% of cases vs grade 3 accounting for 6%. Fatal TEAEs were reported in 17% of the epcoritamab arm vs 6% of the CIT arm, with COVID-19 accounting for 22 of 40 fatal events in the epcoritamab arm. Grade 3–4 infections were more frequent with epcoritamab, at 30% vs 12%; however, febrile neutropenia was less common, at 2% vs 5%. Immune effector cell-associated neurotoxicity syndrome (ICANS) was rare and predominantly low-grade, occurring in 9 in the epcoritamab arm, 2 of which were grade 3–4. Common hematologic toxicities included neutropenia (36% vs 46%), anemia (26% vs 35%), and thrombocytopenia (18% vs 53%).

Fox concluded that EPCORE DLBCL-1 is the first randomized phase 3 trial to demonstrate a statistically significant PFS improvement with CD3×CD20 bispecific antibody monotherapy in R/R LBCL, with clinically meaningful gains in CRR, duration of response, and DOCR — including 59% of complete responders remaining in remission at 3 years. OS was not significantly different, a finding Fox attributed to COVID-19 pandemic confounding and high crossover to effective salvage therapies in the CIT arm. The safety profile was consistent with epcoritamab's established label. These data support a potential role for epcoritamab monotherapy in R/R LBCL patients ineligible for HDT-ASCT.

References

1. Fox C, Inchiappa L, Ferhanoğlu B, et al. Results from EPCORE DLBCL-1: randomized phase 3 study of epcoritamab (EPCOR) vs investigator's choice chemoimmunotherapy (CIT) in patients with relapsed/refractory large B-cell lymphoma (R/R LBCL). Presented at: European Hematology Association 2026 Congress; June 11–14, 2026; Stockholm, Sweden. Abstract S235.
2. Nijland M, Falchi L, Linton K, et al. Clinically relevant subgroup analysis from the randomized phase 3 EPCORE FL-1 trial: treatment (tx) effect of epcoritamab with lenalidomide and rituximab (R²) IN R/R follicular lymphoma (FL). Presented at: European Hematology Association 2026 Congress; June 11-14, 2026; Stockholm, Sweden. Abstract S229.