Safety and efficacy of CD3xCD20 bispecific antibody for the treatment of primary and secondary central nervous system lymphoma patients: A multicentric retrospective study

Abstract

Abstract Introduction: Relapsed/refractory (R/R) central nervous system involvement (CNS) of aggressive B cell lymphoma presents a dismal prognosis with an unmet medical need. CD3xCD20 bispecific antibody (BsAbs) has recently provided a major improvement in the context of R/R aggressive B cell lymphoma with response rates of 50 to 80%. Patients with primary or secondary CNS lymphoma (P/SCNSL) were excluded from trials with CD3xCD20 BsABs, because of concerns regarding their bioavailability in the CNS and the increased risk of neurologic toxicities (ICANS). However, one previous report showed that glofitamab could be measured at an effective concentration in the CSF of 3 patients with SCNSL. The aim of our study is to evaluate the safety and clinical efficacy of CD3xCD20 bispecific antibody in daily practice of R/R P/SCNSL. Methods: Adult patients were eligible for this retrospective study if they presented with a R/R PCNSL, or a R/R aggressive B cell SCNSL with CNS involvement at the start of the treatment with CD3xCD20 BsAbs. Patients and disease characteristics were collected from medical records. Other concomitant anti-lymphoma treatments were authorized. Descriptive statistics and Kaplan-Meier estimates were used for survival analysis. Patients responding to CD3xCD20 BsAbs and subsequently consolidated by intensive chemotherapy and autologous stem cell transplantation (IC + ASCT) or chimeric antigen receptor (CAR) T cell therapy were censored for the progression-free survival (PFS) at time of ASCT or CAR-T cell infusion. Results: 55 patients within 22 French and Belgium institutions were identified. We first present results from 41 patients from 13 institutions for whom sufficient data were collected. Median age was 57 years (range, 34 to 77 years), with 56.1% males. 37 patients had SCNSL and 4 patients had PCNSL. Among patients with SCNSL, 15 patients (40.5%) had isolated CNS relapse when starting BsAbs and 22 patients (59.5%) had simultaneous CNS and systemic relapse. Patients had received a median of 3 previous lines of therapy (range, 2 to 5). Nineteen (46.5%) patients received previous treatment with CAR-T cells and 10 (24.3%) patients received previous ASCT. Thirty (73.2%) and eleven (26.8%) patients received epcoritamab and glofitamab respectively. Seven (23.3%) patients received epcoritamab as monotherapy. Twenty-three (76.7%) patients received epcoritamab in combination with lenalidomide (n=14, 46.7%), ibrutinib (n=5, 16.7%), pomalidomide (n=5, 16.7%), rituximab (n=4, 13.2%) or with polychemotherapies (n=3, 10.7%). One (9.1%) patient received glofitamab as monotherapy. Ten (90.9%) patients received glofitamab in combination with ibrutinib (n=7, 63.6%), lenalidomide (n=6, 54.5%) or pomalidomide (n=2, 18.1%). Patients received a median of 3 cycles of BsAbs (range, 1 to 12). Overall response (ORR) and complete response (CR) rates were 65.9% (27/41) and 36.6% (15/41), respectively. Twelve patients achieved partial response (PR) and 14 patients had progressive disease (PD). CR rate after BsAbs monotherapy and combination was 25% (2/8) and 39.4% (13/33), respectively. Nine responder patients received a subsequent treatment, with axicabtagene ciloleucel CAR-T cell therapy (n = 7 including 3 CR and 4 PR), thiotepa based IC + ASCT (n = 1 PR) and allogenic stem cell transplant (n = 1 CR). With a median follow-up of 6.7 months (range, 0.5 to 17.1), median PFS was 8.5 months and median OS was not reach. The median duration of response (DOR) was not reached. Among the 18 responder patients who continued BsAbs, 2 (7.4%) patients relapsed. One patient relapsed after CAR-T cell therapy. The patient who underwent allogenic transplant relapsed and achieved a 2nd CR after restarting BsAbs. Cytokine release syndrome (CRS) developed in 19 (46.3%) patients including 4 grade 3/4 events. Nine (22.0%) patients experienced neurotoxicities including 3 grade 3/4 events. There were no grade 5 neurotoxicity event. Conclusions: CD3xCD20 BsAbs resulted in encouraging response rates with expected toxicities in heavily pretreated R/R P/SCNSL patients including patients treated with prior CAR-T cell therapy. BsAbs is an option to be considered for R/R SCNSL and these encouraging results highlight the need for prospective trials testing their role in this setting. A prospective trial testing epcoritamab in R/R PCNSL began in France (NCT06931652).