P012 | BRIDGING THERAPY TO CAR-T CELLS IN RELAPSED/REFRACTORY SECONDARY CENTRAL NERVOUS SYSTEM B NON-HODGKIN LYMPHOMAS.

Anti-CD19 CAR-T therapy is a successful treatment for relapsed/refractory B-cell lymphomas, rarely used in secondary central nervous system lymphomas (SCNSL). Bridging therapy (BT) role on outcome is not consensual. At Saint-Louis Hospital (2018–2025), 25 out of 428 B-NHL patients treated with CAR-T were SCNSL. Pts had leukapheresis (LK): Axi-cel (n=9/25,36%), Tisa-cel (n=7/25,28%), Liso-cel 7/25,28%), and Brexu-cel (n=2/25,8%). BT, administered between leukapheresis and CAR-T infusion, was given in 92% (n=23/25) of pts and categorized as chemotherapy (CT), targeted therapy (TT), or a combination (CT+TT). At data cut-off (03/04/2025), histologies were: large B cell (n=20/25), follicular (n=3/25) and mantle cell lymphoma (n=2/25). At treatment decision (TD), median age was 66y, median performance status (PS) was 2. CNS involvement was present in 20 pts (n=9 isolated, n=11 systemic+CNS), and 5 pts became CNS+ before CAR-T infusion. Median n of prior lines was 2. Most received CT (n=15/23), while 4 pts received TT or CT+TT (n=4/23, each). CT were: Etoposide-Ifosfamide (HLX-VP16)+/-Rituximab(R)(n=10), R-Methotrexate-Cytarabine (R-MTX-ARAC)(n=2), R-Gemcitabine-Oxaliplatin (n=1), R-HLX-Cisplatin-VP16 (n=1), R-Dexamethasone-ARAC-Oxaliplatin (R-DHAX)(n=1). CT+TT (concomitant and/or sequential) included R-HLX-VP16+Ibrutinib (Ibru)(n=1) or Lenalidomide (Len)(n=1), R-MTX-ARAC/R-HLX-VP16+Zanubrutinib (n=1), R-DHAX/Ibru (n=1). TT were: Venetoclax (n=1), Ibru (n=1), Idelalisib (n=1), Len (n=1). BT was safe with low morbidity, and no mortality. Grade(G)>2 anaemia was the most common adverse event. BT allowed an improvement of PS in 40% (n=9/23) of pts. Overall Response Rate (ORR) to BT was 43.5%, with 21.7% of CR (46.5% and 26.5% for CT, 75% and 25% for CT+TT). No response was observed after TT. Median vein to vein time was 41 days (range:28-84). All LK pts were injected. CRS was observed in 84% of pts (n=21/25)(G1=13; G2=6; G3=1; G5=1). ICANS occurred in 48% of pts (n=12/25)(G1=2; G2=5; G3=2; G 4=2; G5=1). After a median f-up of 6 mo, median PFS and OS were 6 and 7 mo, respectively. PFS was higher in responders (7 vs 5 mo; p=0.38) and in earlier treated pts (15 vs 5 mo for 2L vs 2L+, p=0.04)(Fig.1). In conclusion, BT in SCNSL is safe and can improve disease control as well as PS leading to a high injection rate in this population characterized by a dismal outcome. Despite a small population, our data suggest encouraging results for combination strategies of BT.