As of July 12, 2024, 36 patients had received ≥ 1 dose of epcoritamab. Baseline demographic and disease characteristics have been reported previously [20]. In this challenging-to-treat, highly refractory population, median age was 68.5 years (range 44–89), 6 patients (16.7%) had transformed DLBCL, 28 (77.8%) had Ann Arbor stage III or IV disease, and 21 (58.3%) had an International Prognostic Index of 3–5 (Table 1). Patients had a median of 3 prior lines of therapy (range 2–8); 30.6% of patients had ≥ 4 prior lines of therapy, 55.6% had primary refractory disease, 80.6% had double-refractory disease (i.e., refractory to both anti-CD20 and an alkylating agent, regardless of the treatments being used in the same or different treatment lines), and 58.3% were refractory to ≥ 2 consecutive lines of therapy. A total of 5 patients (13.9%) received prior bendamustine.

Patients initiated a median of 5.5 epcoritamab cycles (range 1–41) and the median duration of treatment was 5 months (range 0–42). At data cutoff, 7 patients (19.4%) were on treatment and 29 (80.6%) had discontinued treatment. The primary reasons for treatment discontinuation were disease progression (63.9%), adverse events (AEs; 13.9%), and patient request (2.8%).

With a median follow-up of 36.7 months (range 1.5+ to 41.7), the ORR was 55.6%, with a CR rate of 47.2% per IRC (Table 2). Median times to response and CR were 1.4 and 2.7 months, respectively (Table 3), aligned with scheduled response assessments (first assessment at 6 weeks, second assessment at 12 weeks). Deepening of responses was observed with 1 patient converting from PR to CR between the primary analysis and current assessment (at the 24-week imaging assessment). ORR/CR rates were consistently high across subgroups, including patients aged ≥ 75 years (75.0%/62.5%), patients with Ann Arbor stage III or IV disease (53.6%/42.9%), patients with 2, 3, or ≥ 4 prior therapies (56.3%/50.0%, 66.7%/55.6%, and 45.5%/36.4%), and patients with prior ASCT (71.4%/71.4%; Supplemental Fig. 1). Patients with primary refractory disease had lower ORR/CR rates (35.0%/20.0%) compared with other high-risk subgroups and patients without primary refractory disease (81.3%/81.3%). Among 5 patients with prior bendamustine, 4 (80.0%) had a response.

The median DOR and DOCR were 15.2 months and not reached, respectively, based on Kaplan–Meier estimates (Fig. 1a, b). At 3 years, an estimated 45.0% of responders remained in response and an estimated 52.9% of complete responders remained in CR (Table 3). As expected, an association between depth of response and long-term outcomes was observed (Fig. 2a, b). Median PFS was 4.1 months and median OS was 14.9 months. In patients who had a CR, median PFS and OS were not reached. At 3 years, an estimated 52.9% of patients with CR remained progression free and 70.6% remained alive (Table 3).

Kaplan–Meier curves for DOR (a) and DOCR (b) per independent review committee assessment among Japanese adults with relapsed or refractory diffuse large B-cell lymphoma. CR complete response, DOCR duration of complete response, DOR duration of response

Kaplan–Meier curves for PFS by response (a), OS by response (b), PFS in complete responders with MRD negativity (c), and OS in complete responders with MRD negativity (d) among Japanese adults with relapsed or refractory diffuse large B-cell lymphoma. PFS and response are per independent review committee assessment. MRD-negativity status was assessed in a cycle 3 day 1 landmark analysis. The landmark analysis included MRD-evaluable patients who, by cycle 3 day 1, had neither a PFS event nor death. Patients were stratified by MRD status and followed from cycle 3 day 1 until PFS event or death. MRD minimal residual disease, OS overall survival, PFS progression-free survival, PR partial response

Among 30 MRD-evaluable patients, 56.7% became MRD negative. Median duration of MRD negativity was 11.6 months (range 0.0+ to 13.0). A cycle 3 day 1 landmark analysis of 15 MRD-evaluable patients, 11 (73.3%) of whom were MRD negative, revealed longer PFS in patients with MRD negativity (median, not reached) versus patients without MRD negativity (median, 7.8 months after cycle 3 day 1); 3-year estimates for PFS were 54.5% and 25.0% for patients with MRD negativity and without MRD negativity, respectively. OS was also improved in patients with MRD negativity by cycle 3 day 1 (median, not reached; 3-year estimate, 72.7%) versus patients without MRD negativity by cycle 3 day 1 (median, 7.0 months after cycle 3 day 1; 3-year estimate, 25.0%). Among complete responders with MRD negativity, 3-year estimates for PFS and OS were 60.0% and 80.0% (Fig. 2c, d).

Although no patients had prior CAR T-cell therapy, 4 patients went on to receive subsequent CAR T-cell therapy after epcoritamab (median of 154 days from last epcoritamab dose [range 56–327]). Three of these 4 patients remained alive at data cutoff, although 2 patients progressed after subsequent CAR T-cell therapy. No patients received subsequent ASCT; 1 patient received subsequent allogeneic stem cell transplant.

The most common treatment-emergent AEs (TEAEs) were CRS (83.3%); injection-site reaction, including injection-site reaction and erythema (69.4%); neutropenia, including decreased neutrophil count and neutropenia (38.9%); hypokalemia (27.8%); lymphopenia, including decreased lymphocyte count and lymphopenia (27.8%); decreased appetite (25.0%); thrombocytopenia, including decreased platelet count and thrombocytopenia (25.0%); leukopenia, including decreased white blood cell count and leukopenia (22.2%); and rash (22.2%; TEAE overview shown in Table 4). Overall, 50.0% of patients had grade 1 CRS events, 25.0% had grade 2, and 8.3% had grade 3; median time to onset was 15.5 days, median time to resolution was 4.5 days, and no events led to epcoritamab discontinuation.

Infections and infestations by System Organ Class occurred in 58.3% of patients. Six patients (16.7%) experienced COVID-19 and 3 patients (8.3%) experienced cytomegalovirus infection. There were no fungal infections. Immunoglobulin levels decreased from baseline after initiation of treatment and remained generally stable through cycle 36 (Supplemental Fig. 2). A total of 15 patients (41.7%) were given concomitant immunoglobulins.

Grade 3 or 4 infections occurred in 33.3% of patients at any timepoint, the most frequent of which were COVID-19, COVID-19 pneumonia, pneumonia, and urinary tract infection, in 8.3% each. Infections occurred in 33.3% of patients in the first 12 weeks (grade ≥ 3 infections, 11.1%) and in 16.7–42.9% of patients in subsequent 12-week intervals between weeks 109 and 144 (range for grade ≥ 3 infections, 14.3–16.7%). Beyond week 108, the most common type of infection was COVID-19 (14.3–28.6% between weeks 109 and 144). Overall, 11 patients (30.6%) experienced serious infections, 4 of whom had 2 events and 1 of whom had ≥ 4 events; 8.3% of patients had a serious infection in the first 12 weeks versus 14.3–16.7% of patients in subsequent 12-week intervals between weeks 109 and 144.

Cytopenias occurred in 25 patients (69.4%), including 3 patients who had 2 events, 5 who had 3 events, and 15 who had ≥ 4 events. A total of 52.8% of patients experienced cytopenias in the first 12 weeks versus 14.3–33.3% of patients in subsequent 12-week intervals between weeks 109 and 144. Cytopenias with ≥ 15% incidence are shown in Table 4. Febrile neutropenia occurred in 1 patient (2.8%; grade 3) and resolved in 5 days. Eleven of 14 patients with neutropenia (78.6%) required granulocyte colony-stimulating factor treatment for neutropenia. Nine patients (25.0%) had thrombocytopenia, 5 (55.6%) of whom required treatment.

TEAEs led to treatment discontinuation in 5 patients (13.9%): COVID-19, chronic myelomonocytic leukemia, muscular weakness, pancreatic carcinoma, and prolonged electrocardiogram QT in 1 patient each. Of those who discontinued treatment due to TEAEs, the median duration of epcoritamab treatment was 11.1 months; median DOR and DOCR were 13.4 months and 11.1 months, respectively. The median DOCR after treatment discontinuation due to TEAEs was 9.8 months.

There were 5 complete responders who had an epcoritamab pause for > 6 weeks. Reasons for dose delay were AEs in all cases. Among these patients, the median duration of treatment pause was 7.6 weeks (range 6–44) prior to restarting treatment. All patients maintained their CR after resuming epcoritamab.

No fatal TEAEs were reported. Rates and severity of CRS, immune effector cell–associated neurotoxicity syndrome (ICANS), and clinical tumor lysis syndrome (CTLS) were unchanged since the primary analysis [20]. CRS mostly followed the first full dose, with 71.4% of patients experiencing a CRS event during that dosing period.