Outcome of Patients with Mantle Cell Lymphoma after Failure of Anti-CD19 CAR T-Cell Therapy: A Descar-T Study By Lysa Group

Introduction: Since 2020, Chimeric antigen receptor T-cell (CAR-T), namely Brexucabtagene autoleucel (Brexu-Cel), is approved for mantle cell lymphoma (MCL) patients in relapsed or refractory (RR) post BTK inhibitors (BTKi), based on significant complete and duration of response (DOR) in ZUMA-2 trial (Wang et al, NEJM, 2020). Real-life studies confirmed the complete response rate achieved with Brexu-cel (72-82%) but with heterogeneous DOR, and no plateau on survival curves so far (Herbaux et al, Haematologica 2024). Outcome for patients relapsing post CAR-T cells remain largely unexplored. Using the French DESCAR-T registry (NCT04328298), we aimed to describe outcomes and salvage strategies for MCL patients in RR post CAR-T.

Method: After a median follow-up of 14.5 months for the 178 mantle cell lymphoma patients who received a Brexu-Cel infusion between July 2018 and 2023 and had an evaluable for response, 61 (34%) were either relapsed or refractory and 18 died without progression. We analyzed the clinical characteristics and outcomes of the RR patients based on clinical characteristics, timing of relapse and post-CAR-T salvage strategies. Overall survival (OS2) was defined as the time from Brexu-Cel relapse to death from any cause and progression-free survival (PFS2) as the time from failure to progression after salvage treatment.

Results: At infusion, 88.5% of the 61 patients were men with a median age of 66 years, 35.8% had a high MCL International Prognostic Index (MIPI), 76.2% (32/42) a Ki-67 index ≥ 30%, and 30.2% (13/43) a TP53 mutation. All patients were in RR after at least 2 prior lines, including a BTKi (97%) and auto/allo-transplant (44%). Most of them (56/61, 91%) had received bridging therapy with a response in 32%. At time of the first evaluation post CAR-T infusion, 50 (82%) patients presented an objective response, including 28 (46%) complete response. Timing of progression or relapse was 0-3 months in 28% (17 patients), 3-6 months in 42% (25 patients) and after 6 months for the remaining 30% (18 patients). After a median follow up of 15 months post CAR-T failure, median OS2 and PFS2 were 5.8 (CI95: 3.2-11.3) and 2 (CI95: 1.4-2.8) months respectively. Patients with an early relapse (< 3 months) had a median OS2 of 1.8 months, as compared to 6.7 months for those with a relapse within 3-6 months and 9 months for those with a relapse after 6 months. Forty-nine patients received post-CAR-T salvage (no treatment for 10 patients, steroids for one and missing data for 1). Of these, 16 (33%) received Lenalidomide +/- Rituximab (Len/R2), 13 (27%) (immuno)chemotherapy (CT/ICT), 8 (16%) a BTK inhibitor +/- Venetoclax (BTKi/Ven), 7 (14%) a CD3-CD20 bispecific antibody (TCE), 3 (6%) other targeted treatments, and 2 (4%) radiation. Ten patients (20%) had a response, with 9 complete response and 1 partial response. The ORR was 19%, 23%, 43% and 50% for patients treated with Len/R2, CT/ICT, TCE and radiation respectively, while no patients treated with BTKi/Ven or other targeted agent had a response. After a median follow up of 15 months, median OS2 was 6.7, 5.8, not reached, 11.3, 2.8, and 9.6 months for patients treated with Len/R2, CT/ICT, TCE, radiation, BTKi/Ven, and other targeted treatment respectively. Of note, none of the responding patients to TCE had a relapse so far, with a 1 year OS2 of 57% for the 7 treated patients, versus 36% for those treated with LEN/R2 or ICT and 0% for all the others. Within the 39 deaths, the leading cause was lymphoma (35/39, 90%). Interestingly, within all clinical characteristics at time of CAR-T infusion, MIPI score remained the only significant feature associated with OS2 in the multivariable model.

Conclusion : This analysis describes for the first time the poor outcome of Mantle cell lymphoma patients post CAR-T failure. Bispecific antibody (CD3-CD20) based salvage seems to offer promising ORR and prolonged OS2 whereas targeted strategies, including BTKi/Ven, do not provide any response. These data need to be confirmed in larger cohorts.