Real World Comparison of Efficacy and Safety of Fludarabine-Versus Bendamustine-Based Lymphodepleting Chemotherapy for CD19 Chimeric Antigen Receptor (CAR) T-Cell Therapy in Relapse/Refractory (r/r) Large B-Cell Lymphoma (LBCL)

Introduction: Lymphodepleting chemotherapy enhances CAR T-cell expansion, persistence, and clinical activity. Although fludarabine-based lymphodepletion (LD) regimens are the most commonly used, other agents have recently been tested. Additionally, assessing the benefit of alternative LD regimens is important in the setting of drug shortages, which can disrupt patient care. This study aims to compare fludarabine to bendamustine-based LD in the real-world setting prior to CD19 CAR T cell for treatment of r/r LBCL.

Methods: This retrospective, multicenter study utilized Center for International Blood and Marrow Transplant Research (CIBMTR) data to assess outcomes of DLBCL patients treated with commercial CD19 CAR-T therapies for treatment of r/r LBCL from 2017 and 2023. Patient-, disease-, and CAR T-related factors were compared between LD regimens using Chi-square test for categorical variables and Wilcoxon two-sample test for continuous variables. The probability of overall survival (OS) and progression-free survival (PFS) were calculated using the Kaplan-Meier estimator. The probability of overall response rate (ORR) was estimated using cumulative incidence estimates with deaths before ORR as a competing risk. A Cox proportional hazards model was used to identify the significant factors for PFS, OS, and ORR. Logistic regression models were fitted for any grade and severe (grade >3) CRS and ICANS, and prolonged cytopenia.

Results: We identified 5,256 patients with DLBCL treated with axicabtagene ciloleucel (axi-cel), tisagenlecleucel (tisa-cel), or lisocabtagene maraleucel (liso-cel) during the study period. Of these, 4,809 (92.4%) received fludarabine LD and 447 (8.5%) received bendamustine LD. The median follow-up was 23.8 and 9.8 months for fludarabine and bendamustine, respectively. ORR at day 180 was 60% for all patients, with higher rates in the fludarabine group vs. bendamustine group (61.2% vs. 50.8%, p<0.01). The cumulative incidence of CR/PR at day 180 was also higher in the fludarabine group (79.1% vs. 70.2%, p<0.01). The 6- and 12-month PFS was 54.5% and 45.4% for fludarabine vs. 51.4% and 39.1% for bendamustine (p=0.04). No significant differences were observed in OS at 6 and 12 months (61% and 62.2% for fludarabine vs. 77.9% and 61.0% for bendamustine, p=0.65). Multivariate analyses indicated that bendamustine had an inferior ORR (hazard ratio [HR] 0.773, p=0.0013) but there was no difference in CR (HR 0.828, p=0.0606), PFS (HR 1.081, p=0.4990), or OS (HR 1.136, p=0.4194) compared to fludarabine. Age ≤ 65 years was associated with inferior ORR (HR 0.859, p<0.0001) and CR (HR 0.77, p<0.001) after adjusting for LD regimen. Axi-cel (HR 1.5) and liso-cel (HR 1.4) had higher ORR than tisa-cel (p<0.0001) after adjusting for LD regimen. A subset analysis during 2021-2023 adjusting for a more even distribution of LD regimens across CAR T products was performed and the main conclusions were the same.

Recipients of bendamustine experienced lower rates of severe CRS and ICANS compared to those who received the fludarabine (CRS: 4.5% vs. 7.7%, p=0.01; ICANS: 8.3% vs. 16.7%, p<0.01). Additionally, prolonged cytopenia was less common with bendamustine (11.6% vs. 20%, p<0.01). The multivariate analysis confirmed these findings, indicating that bendamustine was associated with lower CRS (odds ratio [OR] 0.445, p<0.0001), ICANS (OR 0.432, p<0.0001), and prolonged cytopenia (OR 0.479, p<0.0001) compared to fludarabine. Liso-cel was associated with lower CRS (OR 0.381, p <0.0001), ICANS (OR 0.507, p<0.0001), and prolonged cytopenia (OR 0.519, p=0.0002) compared to axi-cel and tisa-cel.

Conclusion: Patients receiving fludarabine-based LD had higher ORR, but similar CR without a significant impact on either PFS or OS compared to those receiving bendamustine across CAR T-cell products for treatment of LBCL. Bendamustine was associated with a reduced incidence of severe CRS, ICANS, and prolonged cytopenia. Relevant patient characteristics may inform the choice of LD, considering the trade-off between efficacy and safety. Bendamustine can be considered an alternative LD prior to CAR T for LBCL.