Chimeric Antigen T-Cell (CAR-T) Therapy Versus Autologous Stem Cell Transplantation (ASCT) in Secondary Central Nervous System Lymphoma (SCNSL): A Multicenter Retrospective Analysis

Introduction

Secondary central nervous system (CNS) lymphoma (SCNSL) can occur in a variety of non-Hodgkin lymphoma (NHL) subtypes and is most frequently observed in 2-10% of large B-cell lymphoma cases. After induction chemoimmunotherapy, many centers consider consolidation with autologous stem cell transplant (ASCT). In recent years, chimeric antigen T-cell therapy (CAR-T) has been proposed as a potential treatment strategy. However, comparative data between these two strategies is limited to help guide clinicians for the optimal treatments to achieve durable outcomes in SCNSL.

Methods

This study included 56 patients for analysis: ASCT (n=28) and CAR-T (n=28) across 2 academic institutions in the United States between 2010 and 2024. SCNSL was defined as history of parenchymal, leptomeningeal, and/or cerebrospinal fluid (CSF) involvement by lymphoma at any time prior to receiving cellular therapy. Systemic response was assessed via Lugano criteria and CNS response was assessed according to International PCNSL Collaborative Group criteria. Kaplan-Meier curves were generated for progression-free survival (PFS) and overall survival (OS). Patients with CNS involvement at the time of initial diagnosis and patients with subsequent CNS relapse were included for analysis. Exclusion criteria included: recipients of allogeneic stem cell bone marrow transplant, development of CNS disease after cellular therapy, and patients who had received both CAR-T and ASCT.

Results

Among the ASCT cohort: the most common histologic subtype was diffuse large B cell lymphoma (DLBCL) and high-grade B cell lymphoma (HGBCL) (n=25/28, 89%). The median age among ASCT recipients was 57. In the ASCT cohort: activated B-cell (ABC) subtype was seen in 4 patients (n=4/28) and a small number (n=3/28) patients had double or triple hit re-arrangements (c-MYC, BCL-2, BCL-6). At the time of ASCT 82% of patients were in complete response (CR) for systemic disease (n=23/28), and 79% of patients were in CR for CNS disease (n=22/28). BCNU/Thiotepa conditioning was the most common regimen (n=26/28, 93%). At a median time of follow up of 16 months, 61% of patients were alive (n=17/28). The median PFS was 47.1 months for the ASCT cohort and the median overall survival (OS) was 48 months. Among the CAR-T cohort: DLBCL and HGBCL were the most common histologic subtypes (n=25/28, 89%). Median age among CAR-T recipients was 63. 25% (n=7/28) were ABC subtype, and 36% (n=10/28) had double or triple hit gene re-arrangements. A smaller proportion of patients were in CR prior to infusion of CAR-T: 39% (n=11/28) were in CR for their systemic disease and 57% (n=16/28) were in CR for their CNS disease prior to CAR-T. At a median follow up of 4.6 months, 46% (n=13/28) of CAR-T recipients were alive and the median OS was 7.0 months.

Importantly, 71% (n=20/28) had received ≥ 1-2 lines of systemic therapy prior to ASCT compared to 46% (n=13/28) of patients in the CAR-T group. PFS was higher (p=0.025) in the ASCT cohort with median PFS of greater than 47 months compared with 7 months in the CAR-T cohort. This correlated with overall survival which was significantly longer (p=0.018) in ASCT compared to CAR-T group: 48 months versus 7.0 months, respectively.

Limitations of this study include retrospective analysis, disease heterogeneity, and differences in number of prior lines of therapy. Lastly, propensity score matching was not performed in analysis to account for heterogeneity.

Conclusions

ASCT remains a reasonable option for durable remissions in the management of SCNSL. Improved outcomes among the ASCT cohort may be explained by a higher proportion of patients in CR. The majority of patients with SCNSL present with acute neurologic symptoms that mandate urgent therapy and render urgent autologous CAR-T cell therapy logistically challenging. This data highlights two important points regarding consolidation that suggest a response-adapted approach: (1) those in CR after salvage appear to have excellent outcomes with ASCT; and (2) refractory patients may be salvaged with CAR-T cells, with expected lower rates of durable remissions. Additional research is needed in the management of SCNSL to better understand the nuances of disease biology, predict responses to therapy, and further develop personalized lymphoma care to improve outcomes.

Davis:Janssen Biotech: Speakers Bureau. Ahmed:Kite, a Gilead Company: Research Funding; Bristol Myers Squibb: Consultancy. McGuirk:Legend biotech: Consultancy; Autolus: Consultancy; NEKTAR therapeutics: Consultancy; Sana technologies: Consultancy; BMS: Consultancy; Kite: Consultancy; Novartis: Consultancy; Caribou bio: Consultancy; CRISPR therapeutics: Consultancy; Allo Vir: Consultancy; Envision: Consultancy. Hoffmann:ADC, Janssen, Pharmacyclics, BeiGene, Novartis, Astra-Zeneca, Abbvie, Kite, TG: Consultancy, Honoraria; Bristol Myers Squibb: Other: Travel; Genentech: Consultancy, Research Funding.