3417 - Patterns of CNS Failures in Relapse/Refractory Large B-Cell Lymphoma (LBCL) Patients with Secondary CNS Disease Following Chimeric Antigen Receptor T-Cell (CAR T) Therapy

J. Y. Nakashima¹, V. M. Khatri¹, R. J. Cruz-Chamorro¹, J. Zhou², P. Patra², S. Gaballa³, F. Khimani⁴, S. Mirza⁴, B. D. Shah³, H. Saeed³, J. C. Chavez³, F. L. Locke⁴, T. Nishihori⁴, H. D. Liu⁴, N. Dong³, A. Lazaryan⁴, T. J. Robinson⁵, C. Freeman⁴, M. D. Jain⁴, and N. B. Figura¹; ¹H. Lee Moffitt Cancer Center and Research Institute, Department of Radiation Oncology, Tampa, FL, ²University of South Florida, Tampa, FL, ³H. Lee Moffitt Cancer Center and Research Institute, Department of Malignant Hematology, Tampa, FL, ⁴H. Lee Moffitt Cancer Center and Research Institute, Department of Blood and Marrow Transplant and Cellular Immunotherapy, Tampa, FL, ⁵Yale University, New Haven, CT

Purpose/Objective(s): Relapsed/refractory large B-cell lymphoma (LBCL) patients with CNS involvement were excluded from the seminal prospective CAR T-cell trials. For those with CNS disease, the optimal CNS-directed bridging therapy remains unknown. Understanding the patterns of failure following CAR T for these patients may help direct an informed treatment strategy. We report a single-institution retrospective analysis of the clinical outcomes of patients with secondary CNS lymphoma following CAR T and analyze of their patterns of recurrence. Materials/

Methods: We retrospectively identified LBCL patients with secondary CNS disease who underwent CAR T therapy at our institution from 2018 to 2023. Patients with bridging radiotherapy (RT) were excluded. Baseline patient characteristics were categorized to assess potential risk factors. Time-to-event outcomes was estimated using Kaplan-Meier method and compared with log-rank tests. Patterns of failure analysis was performed by fusing three MRIs for each patient – 1) at time of initial CNS diagnosis, 2) at pre-CAR infusion, and 3) at post-CAR relapse – and comparing the location and disease extent. CNS recurrences were then categorized as either local, marginal (defined within 1cm), or distant.
Results: 21 patients with secondary CNS lymphoma were identified. The median age of patients was 63 (35-78) with the majority having DLBCL (n=18; 86%). 14 patients received axi-cel and 7 received tisa-cel. Seven patients had no evidence of systemic disease at time of CAR. The majority (n=15, 71%) had systemic bridging therapy. At a median follow-up of 16.7 months, 13 patients (62%) experienced disease progression. Seven patients experienced systemic-only progression, four had CNS-only progression, while two had concurrent CNS and systemic disease recurrence. The 12-month progression-free (PFS) and overall survival (OS) were 38.5% and 61.2%, respectively. The 12-month CNS-PFS was 64.2%. In the six patients with any CNS recurrence, 2 patients had strictly intracranial CNS failures, 2 had strictly extracranial CNS (i.e. spinal) failures, while 2 patients failed in both sites concurrently. All patients (n=6; 100%) experienced distant CNS failures when compared to their pre-CAR infusion MRIs. When comparing the MRIs at time of CNS diagnosis, the majority of patients (n=5; 88.8%) continue to have distant progression with one developing a marginal CNS recurrence. No patients experienced a local CNS failure when compared to either MR data set. Seven patients (33%) had grade 3+ neurotoxicity, while none had grade 3+ cytokine-release syndrome.
Conclusion: In patients with secondary CNS disease, patients tend to experience distant CNS progression following CAR T-cell therapy. This may suggest that global CNS-directed therapies, such as intrathecal, systemic therapy and/or whole brain RT, may be a preferable bridging strategy to prevent CNS failures post-CAR.