CD3xCD20 bispecific antibodies in transformed Waldenström macroglobulinemia/lymphoplasmacytic lymphoma

Histological transformation (HT) of Waldenström macroglobulinemia (WM)/lymphoplasmacytic lymphoma (LPL) represents a rare complication, occurring in less than 5% of cases. Patients usually present with high‐risk features, including extra‐nodal involvement, elevated lactate dehydrogenase levels, high IPI scores, and frequent central nervous system (CNS) involvement.1, 2, 3, 4, 5 These patients carry a poor prognosis after standard treatments, with shorter survival compared to non‐transformed WM and de novo large B‐cell lymphoma (LBCL),2, 6 especially those who are not eligible or relapsing after high‐dose chemotherapy with autologous stem cell transplant (SCT) and/or chimeric antigen receptor (CAR) T‐cell therapy.7, 8 CD3xCD20 bispecific monoclonal antibodies (BsAb) (epcoritamab and glofitamab) provide overall and complete responses of around 60% and 40%, respectively, in relapsed/refractory (R/R) aggressive B‐cell lymphomas, including de novo LBCL and transformed indolent lymphomas.9, 10, 11 However, no data exist for BsAb in patients with transformed WM. In the pivotal studies, only patients with transformed follicular lymphoma (FL) were included in the glofitamab study, and no information was provided on the antecedent histology for the 40 patients with transformed indolent lymphomas in the epcoritamab study. The aim of this study was to evaluate the efficacy and safety of CD3xCD20 BsAb in patients with R/R transformed WM.

Here, we retrospectively identified patients with HT‐WM treated with CD3xCD20 BsAb in FILO/LYSA centers (n = 16 patients), two other European centers (n = 3 patients), and one US center (n = 1 patient). Adult patients with a confirmed diagnosis of WM/LPL and R/R HT‐WM who were treated with epcoritamab or glofitamab for biopsy‐proven LBCL were included. The diagnosis of LBCL was made in accordance with the World Health Organization classification of malignant lymphoma and confirmed for FILO/LYSA cases by an expert hematopathologist from the Lymphopath network. The primary endpoint was the best overall response rate (ORR), according to the Lugano 2014 classification. ¹² The secondary endpoints included the best complete response (CR) rate, progression‐free survival (PFS), overall survival (OS), and safety, including specific toxicities such as cytokine‐release syndromes (CRS), immune effector cell‐associated neurotoxicity syndrome (ICANS), infections, and hematological toxicity. CRS and ICANS were graded according to the American Society for Transplantation and Cellular Therapy (ASTCT) grading system. Other adverse effects were graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE version 5.0).

This study enrolled 20 patients treated between February 2023 and May 2025. The median age at diagnosis of WM was 63 years (range, 38–83). MYD88 ^L265P and CXCR4 mutations were detected in 91% (10/11) and 17% (1/6) of patients with available data, respectively. The median time from WM to HT diagnosis was 3.3 years (interquartile range [IQR], 0.2–12.4), including 5 cases (25%) with concurrent diagnosis of WM and LBCL. The patients received a median of 3 prior lines of treatment (range, 2–9, median of 1 line for WM and 2 lines for HT before BsAb), and 8 patients (40%) had received prior CAR T‐cell therapy (Supporting Information S1: Table S1). Nine patients (45%) received a BTK inhibitor (ibrutinib, n = 6; zanubrutinib, n = 3) for WM (n = 7) or HT (n = 2); all were refractory or eventually progressed on a BTK inhibitor. CNS involvement due to LBCL was present in three patients at HT and in one patient at the time of BsAb. Additionally, four patients presented CNS involvement by WM, that is, Bing–Neel syndrome (BNS) at the time of BsAb therapy. The median time from HT diagnosis to BsAb administration was 16 months (IQR, 5–33); the median age was 71 years (range, 46–86), with seven patients (37%) older than 80 years. Twenty percent of patients had an ECOG performance status of 2 or higher, nearly two‐third of patients (68%) had elevated LDH, and 74% had Ann Arbor stage III or IV disease. Sixty‐five percent of patients were refractory to the last therapy before BsAb and 55% were refractory to at least two consecutive lines. Patient characteristics according to prior exposure to CAR T‐cell therapy are depicted in Table 1. While the CAR T‐cell‐exposed cohort was more heavily pre‐treated for HT (median of 3 lines, prior autologous SCT for 38% of patients), the CAR‐T‐naïve cohort consisted of older patients (median of 81 years at BsAb), with a higher proportion of refractory disease.

Table 1.

Patient characteristics.

Variable	All patients n = 20	CAR‐T‐naïve n = 12	CAR‐T‐exposed n = 8	P value
Sex male/female (ratio)	12/8 (1.5)	7/5 (1.4)	5/3 (1.7)	1
WM characteristics
Median age at WM (years, range)	63 [38–83]	67 [38–83]	60 [46–70]	0.35
MYD88 L265P mutation	10/11 (91%)	6/6 (100%)	4/5 (80%)	0.45
CXCR4 mutation	1/6 (17%)	1/4 (25%)	0/2 (0%)	1
Median number of lines for WM (range)	1 [0–9]	1 [0–9]	0 [0–2]	0.04
HT characteristics
Median age at HT (years, range)	70 [45–85]	79 [45–85]	60 [52–70]	0.005
Histology
LBCL, not otherwise specified	20 (100%)	12 (100%)	8 (100%)	1
Hans algorithm				1
Non‐GC	18 (95%)	11 (92%)	7 (100%)
GC	1 (6%)	1 (8%)	0 (0%)
Extranodal involvement	14 (70%)	10 (83%)	4 (50%)	0.16
CNS involvement	3 (15%)	2 (17%)	1 (13%)	1
Median serum IgM level (g/L, range)	10.9 [1–59.7]	11.9 [1–59.7]	9.8 [1.2–20]	0.93
Ann Arbor stage III–IV	15 (75%)	9 (75%)	6 (75%)	1
IPI ≥ 3	11/18 (61%)	7/11 (64%)	4/7 (57%)	1
Median number of lines for HT (range)	2 [0–5]	2 [0‐5]	3 [2‐5)	0.01
Prior autologous HSCT	3 (15%)	0 (0%)	3 (38%)	0.04
Characteristics at BsAb
Median age at BsAb (years, range)	71 [46–86]	81 [46–86]	62 [55–71]	0.009
>80 years	7 (37%)	7 (58%)	0 (0%)
Median time from HT to BsAb (months)	16 [1–53]	10 [1–53]	28 [8–48]	0.11
Refractory to last therapy	13 (65%)	10 (83%)	3 (38%)	0.06
Refractory to ≥2 consecutive therapiesa	11 (55%)	7 (58%)	4 (50%)	1
ECOG performance status ≥2	4 (20%)	3 (25%)	1 (13%)	0.62
Elevated LDH	13 (68%)	7 (58%)	6 (86%)	0.33
Ann Arbor stage III–IV	14 (74%)	8 (73%)	6 (75%)	1
CNS involvement	1 (6%)	1 (13%)	0 (0%)	1
CD3xCD20 BsAb				1
Epcoritamab	13 (65%)	8 (67%)	5 (63%)
Glofitamab	7 (35%)	4 (33%	3 (37%)

Abbreviations: BsAb, bispecific antibody; CAR, chimeric antigen receptor; CNS, central nervous system; ECOG, Eastern Cooperative Oncology Group; GC, germinal center; HSCT, hematopoietic stem cell transplantation; HT, histological transformation; IPI, International Prognostic Index; LBCL, large B‐cell lymphoma; LDH, lactate dehydrogenase; WM, Waldenström macroglobulinemia.

^aDisease progression or stable disease as best response or disease progression within 6 months after completion of therapy. Applies only for HT‐directed regimens.

Patients received epcoritamab (n = 13) or glofitamab (n = 7) BsAb, used as monotherapy in 16 patients (80%) or combination (20%). Among the 13 patients treated with epcoritamab, 2 received zanubrutinib in combination (one for concurrent CNS involvement by LBCL and the other for concurrent BNS) and 2 received lenalidomide in combination. The median number of cycles administered was 3 (range, 1–21) for epcoritamab and 12 (range, 3–12) for glofitamab. The best ORR was 80% (95% confidence interval [CI], 58%–92%), including 45% CR (95% CI, 26%–66%) (Supporting Information S1: Table S2). The best ORR was 83% and 75%, and the best CR was 42% and 50% in CAR‐T‐naïve and CAR‐T‐exposed patients, respectively. The median time to response and CR was 44 days (range, 22–174) and 80 days (range, 37–253), respectively. At last follow‐up (FU), 12 patients (60%) were alive, of whom 67% (8/12) remained in CR (7/8 achieved with BsAb) and 25% in partial response (PR), and 5 patients (25%) still received BsAb (Figure 1A). The reasons for discontinuing BsAb included relapsed or progressive disease (PD) in seven patients, planned end of treatment in five patients, consolidation with CAR‐T or allogeneic SCT in 2 patients (one each), and death due to infection in one patient. Regarding WM disease evaluation, serum IgM levels before and after BsAb were available for 11 patients, and four patients had undetectable IgM peak before BsAb therapy (Supporting Information S1: Figure S1). The mean IgM level before BsAb was 3.8 g/L (IQR, 1–10). For the four patients with IgM level >4 g/L before BsAb, two CR, one PR, and one minor response were observed after treatment with BsAb.

Figure 1.

Outcomes after bispecific antibodies. Swimmer's plot depicting time to response, duration of response, and type of BsAb received (A). Kaplan–Meier estimates of progression‐free survival and overall survival from BsAb initiation (B).

With a median FU of 11 months (95% CI, 7–21), the estimated PFS rates at 6 months and 1 year were 59% (95% CI, 41–86) and 51% (95% CI, 31–82), respectively (Figure 1B). The estimated OS rates at 6 months and 1 year were 79% (95% CI, 62–100) and 57% (95% CI, 37–89), respectively (Figure 1B). The 1‐year PFS was 66% (95% CI, 43–100) and 33% (95% CI, 12–96), and the 1‐year OS was 71% (95% CI, 48–100) and 47% (95% CI, 22–100) for CAR‐T‐naïve and CAR‐T‐exposed patients, respectively. On univariate analysis, only elevated LDH was correlated with worse PFS (P = 0.02) and OS (P = 0.008) in an exploratory analysis. Eight patients experienced relapse/progression during FU, all related to LBCL. Salvage therapy consisted of various regimens (ifosfamide + etoposide, ibrutinib, obinutuzumab associated with GEMOX, or ibrutinib + lenalidomide). One patient, in CR after three cycles of epcoritamab and lenalidomide, was retreated with this combination after relapsing 10 months after therapy discontinuation. At the time of data cut‐off, eight patients have died, five from progressive LBCL and three from infection.

Fifteen patients (75%) experienced CRS (5% grade 3), and 15% experienced ICANS (0% grade 3–4). The median duration was 1 day (IQR, 1–1) for CRS and 6 days (IQR, 1–20) for ICANS. Only one patient (5%) with grade 3 CRS and grade 1 ICANS was admitted to the intensive care unit. Infections occurred in 45% of patients, including four grade 3 events (20%). Between 5% and 20% of the patients developed grade ≥3 prolonged cytopenias (Supporting Information S1: Table S3).

Although cross‐trial comparisons need to be interpreted with caution, the response rates observed in our study are consistent with those observed with epcoritamab and glofitamab in the pivotal studies, especially for patients with transformed lymphoma (67.5% ORR and 45% CR with epcoritamab, 50% CR with glofitamab) and patients with activated B‐cell sub‐type lymphoma (66% ORR and 48% CR with epcoritamab, 59% CR with glofitamab).9, 10, 11 The population of our study was comparable in terms of number of previous lines of therapy, refractoriness, and previous exposure to autologous SCT and CAR‐T therapy but included older patients (71 years vs. 64 and 66 years in the pivotal studies). Recent real‐world data analyses have confirmed the efficacy and manageable safety profile of BsAb in R/R LBCL; however, patients with transformed lymphoma usually represent a minority of patients in these studies and are mainly transformed FL.13, 14, 15 Interestingly, one study suggested that transformed lymphomas could carry a better prognosis than de novo LBCL following BsAb treatment. ¹⁵

Although we report a limited, retrospective, and potentially biased case series, CD3xCD20 BsAb are associated with high response rates in R/R HT‐WM, including for patients relapsing after CAR‐T therapy. These results need to be confirmed in a larger cohort and with a longer FU, but they already support the use of BsAb in this highly refractory and difficult‐to‐treat population.

AUTHOR CONTRIBUTIONS

Morgane Brocard: Conceptualization; investigation; writing—original draft; methodology; writing—review and editing. Damien Roos‐Weil: Writing—review and editing; supervision. Lukshe Kanagaratnam: Conceptualization; methodology; writing—review and editing; supervision. Thomas Hueso: Writing—review and editing. Lucile Bussot: Writing—review and editing. Benjamin Papoular: Funding acquisition; writing—review and editing. Cécile Tomowiak: Writing—review and editing. Steeve Chevreux: Writing—review and editing. Gilles Crochet: Writing—review and editing. Tristan Vaugeois: Writing—review and editing. Elise Toussaint: Writing—review and editing. Simone Ferrero: Writing—review and editing. Charalampia Kyriakou: Writing—review and editing. Anne Quinquenel: Writing—review and editing. Prashant Kapoor: Writing—review and editing. Eric Durot: Supervision; writing—review and editing; methodology; writing—original draft; conceptualization; investigation.

CONFLICT OF INTEREST STATEMENT

D.R.W. reports funds and/or consulting fees from AbbVie, AstraZeneca, BeOne, and Johnson and Johnson. G.C. reports consultancy/travel fees from Roche, AbbVie, Kite/Gilead, Sobi, Amgen, Johnson and Johnson, and BMS. S.F. is a consultant for Janssen, EUSA Pharma, AbbVie, and Sandoz, is on the advisory board of Janssen, EUSA Pharma, Recordati, Incyte, Roche, AstraZeneca, CSL Behring, and Italfarmaco, received speaker's honoraria from Janssen, EUSA Pharma, Recordati, Lilly, BeOne, Gilead, and Gentili, and received research funding from Gilead, BeOne, and Morphosys. C.T. reports consulting fees from Johnson and Johnson, AstraZeneca, BeOne, AbbVie, and Lilly. P.K. is the principal investigator of trials for which Mayo Clinic has received research funding from Amgen, Regeneron, Bristol Myers Squibb, Loxo Pharmaceuticals, Ichnos, Karyopharm, Sanofi, AbbVie, and GlaxoSmithKline, has received honorarium from Keosys, and served on the Advisory Boards of BeOne, Mustang Bio, Janssen, Pharmacyclics, X4 Pharmaceuticals, Kite, Oncopeptides, Ascentage, Angitia Bio, GlaxoSmithKline, Sanofi, and AbbVie.

ETHICS STATEMENT

This retrospective study was conducted in accordance with the Declaration of Helsinki. This study was authorized by the “Commission Nationale Informatique et Libertés” (authorization number 2206749v0), allowing the computerized management of the medical data (MR004 methodology) for French centers, and by each institutional review board for other European and US centers. The participants were informed of the research purposes and had a right of opposition.

FUNDING

This research received no funding.

Supporting information

supporting information.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request. For original data, please contact edurot@chu‐reims.fr.