Chimeric Antigen Receptor T-Cell (CAR-T) Therapy in Secondary Central Nervous System Large B-Cell Lymphoma (SCNSL): A Multicenter Retrospective Analysis
Gulrayz Ahmed,Aseel Alsouqi,Aniko Szabo,Alexandra E. Rojek,Peter A. Riedell,Farrukh T. Awan,Laura Samples,Mazyar Shadman,Marie Hu,Veronika Bachanova,William Wesson,Nausheen Ahmed,Madiha Iqbal,Mohamed A. Kharfan-Dabaja,Michael Scordo,Patrick Connor Johnson,Yi-Bin Chen,Sawa Ito,Mehdi Hamadani,Matthew J. Frigault
DOI: https://doi.org/10.1182/blood-2023-184345
IF: 20.3
2023-11-28
Blood
Abstract:Introduction Secondary central nervous system (CNS) involvement develops in 2-10% of large B-cell lymphoma (LBCL) cases. Current treatment approaches for relapsed or refractory CNS disease generally do not provide durable responses. Limited data suggest efficacy and feasibility of CAR-T therapy in SCNSL. We report here the largest multicenter retrospective analysis of the safety and efficacy of CD19-directed CAR-T therapy in SCNSL patients. Methods This study included SCNSL patients who received commercial CD19-directed CAR-T therapy at 9 academic institutions in the United States between 2016 and 2022. SCNSL was defined as history of parenchymal and/or leptomeningeal/cerebrospinal fluid (CSF) involvement by lymphoma at any time prior to receiving CAR-T therapy. Systemic response was assessed using the Lugano criteria while CNS response was assessed as per International PCNSL Collaborative Group criteria. Kaplan-Meier curves were generated for progression-free survival (PFS) and overall survival (OS) for entire cohort, as well as subsets of patients with and without active CNS disease at the time of CAR-T infusion. Results Ninety patients were included in the analysis, with baseline characteristics shown in Table 1. CAR-T products utilized include axicabtagene ciloleucel (n= 38, 42%), tisagenlecleucel (n=37, 41%) and lisocabtagene maraleucel (n= 15, 14%). Fludarabine/cyclophosphamide was used as lymphodepletion in 84 patients (93%). 28 patients received CNS directed radiation therapy (RT) prior to CAR-T infusion (31%) within a median time of 28.5 days (range 5 - 906) from last day of RT. Amongst all patients at the time of CAR-T infusion 82 (91.1%) had active CNS and/or systemic involvement while 8 (8.9%) had no evidence of disease; 68 (75.5%) had active CNS involvement while 22 (24.5%) had a prior history of CNS involvement but no active at the time of CAR-T infusion. ( Table 1) Median follow up all patients was 0.5 year (range 0.01 - 4.7). 71 patients (79%) developed cytokine release syndrome (CRS; median time to onset 3 days) with 3.3% patients developing Grade 3-4. Immune effector cell-associated neurotoxicity syndrome (ICANS; median time to onset 5 days) was reported in 55 patients (61%), with 28.8% developing Grade 3-4. No patients died due to CRS or ICANS. At 1-month post CAR-T, 80 patients were evaluable for CNS response and 71 patients in the entire cohort were evaluable for systemic response. Similarly, at 3-months these numbers were 66 and 60, respectively. Objective response rates (ORR) were 75% and 80% at 1-month for CNS and systemic disease respectively, while corresponding ORRs at 3-months were 68% and 76.7%. At 1- and 3-months timepoint 36 (45%) and 33 (50%) patients, respectively achieved CNS complete remission (CR), while 44 (62%) and 42 (70%) patients respectively achieved systemic CR. 1- and 2-year PFS for entire cohort were 25% (95%Cl 15% - 34%) and 16% (95%Cl 8.5% - 28%) respectively, while 1- and 2-year OS were 46% (95%Cl 36% - 59%) and 31% (95%Cl 22% - 46%). Non-relapse mortality (NRM) were 4.6% (95%CI, 1.8% -12%) and 7.2% (95%CI 2.9% - 18%) at 1- and 2-year. Cumulative incidence of relapse at 1- and 2-year were 73% (95%Cl 63% - 83%) and 77% (95%Cl 68% - 88%), respectively. ORRs for CNS response in patients with active CNS disease at 1-month and 3-month interval were 71% (n=45) and 60% (n=31). Similarly, ORR for systemic response with active CNS disease at 1-month and 3-month interval were 85% (n=43) and 74.3% (n=32).The 2-year PFS for patients with and without active CNS disease at CAR-T infusion were 12% (95%Cl 5.4% - 27%) and 28% (95%Cl 12% - 65%) respectively, with a corresponding OS of 25% (95%Cl 15% - 41%) and 58% (95%Cl 37% - 91%). ( Figure 1) The 2-year NRM rates for patients with and without active CNS disease at CAR-T infusion were 4.6% (95%Cl 1.5% - 14%) and 14% (95%Cl 3.6% - 53%) respectively, while the cumulative incidence of relapse in similar order were 83% (95%Cl 73% - 95%) and 58% (95%Cl 39% - 88%). Conclusion In this largest known case series to date, CAR-T therapy appeared safe and feasible in patients with SCNSL, but is associated with high rates of treatment failure, particularly in patients with active CNS disease at CAR-T infusion. This warrants further study of optimal CNS-directed treatment strategies prior to CAR-T, development of novel cellular therapy platforms or consideration of post CAR-T maintenance therapy in these high-risk patients.
hematology