AUTHORS
Monica Wielgos-Bonvallet, PhD, Kubra Karagoz, PhD, MSc2, Daniele Bizzarri3, Joshua D. Brody, MD4, Chan Y Cheah5, Catherine Thieblemont, MD, PhD6, Anoek Zomer7, John Karavitis8, Edith Szafer Glusman, PhD9, Malene Risum, MD, PhD10, David Soong11, Han Si12, Andrew J Steele, PhD13
AFFILIATIONS
1 Genmab, Plainsboro, New Jersey, United States
2 Genmab, Plainsboro, New Jersey, United States
3 Genmab, Utrecht, Netherlands
4 Icahn School of Medicine at Mount Sinai, New York, New York, United States
5 Sir Charles Gairdner Hospital and the University of Western Australia, Perth, Western Australia, Australia
6 Hemato-oncology, Hôpital Saint-Louis – Assistance Publique - Hôpitaux de Paris, University of Paris Cité, Paris, France
7 Genmab, Utrecht, Netherlands
8 Genmab, Plainsboro, New Jersey, United States
9 AbbVie Inc., South San Francisco, California, United States
10 Genmab, Copenhagen, Denmark
11 Genmab, Plainsboro, New Jersey, United States
12 Genmab, Plainsboro, New Jersey, United States
13 Genmab, Plainsboro, New Jersey, United States
ABSTRACT
Introduction: Epcoritamab (epcor) is a subcutaneous CD3×CD20 bsAb with clinical activity in R/R DLBCL. PET/CT-based response assessment remains the gold standard for evaluating clinical benefit. However, circulating tumor DNA (ctDNA)-based assessment of MRD offers a molecular approach to monitor disease in blood that may provide insights into clinical responses beyond imaging alone. We investigated whether differences in response durability observed through serial MRD profiling after epcor monotherapy were associated with immune system variability in patients (pts) with R/R DLBCL.
Methods: In the EPCORE NHL-1 (NCT03625037) expansion cohort, pts with R/R DLBCL received epcor with a 2-dose step-up followed by the full 48-mg dose on cycle (C)1 day (D)15, continuing until PD or unacceptable toxicity. PET/CT-defined best overall response was assessed by IRC using Lugano criteria. MRD was assessed by plasma ctDNA dynamics using the AVENIO Oncology Assay for NHL (Roche). The biomarker-evaluable population included pts with available MRD data who were MRD-positive (MRDpos) at C1D1 and had ≥1 on-treatment MRD assessment. MRD negative (MRDneg) was defined as <1 mutant molecule per mL and MRDneg results were categorized as sustained (≥3 consecutive MRDneg results) or transient (1–2 consecutive MRDneg results). Baseline absolute lymphocyte counts were analyzed locally, and peripheral blood immunophenotyping was assessed centrally by flow cytometry at baseline and throughout treatment. P values were based on Wilcoxon rank-sum testing.
Results: In the biomarker-evaluable population (N=109 of 139 pts enrolled in the expansion cohort) as of May 2025, median PFS was 18.5 mo in responders (CR; n=49: 39.1 mo; PR; n=27: 4.1 mo) vs 1.2 mo in nonresponders/non-evaluable (nonR) pts (n=33; P<0.0001). The duration of MRDneg further stratified outcomes: median PFS was not reached in sustained MRDneg responders (35 CR, 2 PR), was 8.3 mo in transient MRDneg responders (10 CR, 8 PR), 3.7 mo in MRDpos responders (4 CR, 17 PR), and 1.2 mo in MRDpos nonR pts (2 SD, 24 PD, 7 NE; P<0.0001). Pts with sustained MRDneg had longer median DOR than those with transient MRDneg (37.7 mo [95% CI, 26.5–not available] vs 5.6 mo [95% CI, 2.8–26.5]) and became MRDneg earlier (median, 57 vs 113 d, respectively). Despite these differences, baseline clinical risk factors were similar between sustained and transient MRDneg groups.
Baseline peripheral blood immune profiling showed higher absolute lymphocyte counts in sustained and transient MRDneg pts versus MRDpos pts (responders + nonR; median, 0.91 and 0.98 vs 0.53×10⁹/L; P=0.0016 and P=0.0219, respectively), with a similar pattern for CD4+ and CD8+ T cells. Baseline PD-1+ CD8+ effector memory T cell (TEM) frequencies were lower in sustained and transient MRDneg pts versus MRDpos pts (median, 7.7% and 16.1% vs 25.1%; P<0.001 and P=0.0226). Similar patterns were observed in clinical responders versus nonR. On treatment, a rapid, significant increase from baseline in CD4+ T cells and CD8+ terminally differentiated TEM (TEMRA) cells was observed in sustained versus transient MRDneg pts following T-cell margination during C1. PD1+ CD8+ TEMRA cell counts expanded up to C4D1 then later declined in sustained, but not transient, MRDneg pts. Early increases in PD1+ CD8+ naive, central memory, and TEM cell counts were similar between groups. However, these populations decreased over time in sustained MRDneg pts, whereas transient MRDneg pts maintained higher expression throughout treatment.
Conclusions: Sustained MRD negativity identified pts with R/R DLBCL treated with epcor who had durable clinical benefit, suggesting that serial ctDNA-based MRD assessment provides insight into response depth and durability beyond PET/CT-based response and single-timepoint MRD negativity. MRDneg pts showed greater baseline immune fitness than MRDpos pts. Among MRDneg pts, sustained MRDneg was associated with more robust on-treatment T-cell recovery and early T-cell activation that resolved over time, whereas persistent T-cell activation in transient MRDneg pts may reflect continued target engagement and trimer formation. Together, these data highlight differences in the immune repertoire between sustained and transient MRDneg pts that may explain durable efficacy following epcor treatment.
DISCLOSURES
Monica Wielgos-Bonvallet, PhD: Genmab - Current equity holder in publicly-traded company; Genmab - Current Employment