Comparison of Bendamustine and Fludarabine/Cyclophosphamide As Lymphodepletion before CAR-T in Patients with Diffuse Large B-Cell Lymphoma

CART-19 therapies have advanced treatment options for relapsed/refractory diffuse large B-cell lymphoma (r/r DLBCL). Lymphodepletion (LD) with fludarabine and cyclophosphamide (Flu/Cy) has been the standard practice, but recent Flu shortages necessitated alternatives like bendamustine (Benda). We compared the efficacy and toxicity of Benda vs Flu/Cy in patients undergoing CART-19 therapy.

This is a retrospective single-center study of 90 patients with r/r DLBCL treated with axi-cel (74.4%) or liso-cel (25.6%) at UPMC Shadyside from 06/2018 - 10/2023. Our initial cohort had 88 Flu/Cy and 45 Benda patients. A propensity score matched data set was created to balance baseline covariates between LD groups. Given the smaller number of subjects in the Benda group, 1:1 (n=45 for each) matching was used. Covariates used in the matching were age, stage, bridging therapy, number of prior therapies, transformation from indolent lymphoma, CAR-T product, pre-LD C-reactive protein and lactate dehydrogenase. Data were collected via chart review and assessed with descriptive statistics and Kaplan-Meier analyses. Hematologic parameters included: absolute neutrophil (ANC), platelet (PLT) and lymphocyte counts (ALC), and hemoglobin (Hb). Responses were assessed with Lugano 2014 criteria, and toxicities graded based on the ASTCT and Common Terminology Criteria for Adverse Events standards.

ALCs were decreased by a mean of 85.8% (± SD 98.5%) in Flu/Cy, while by 98.7% (± 316%) in Benda (p=0.8). Response rates were similar in the two groups, with a 3-month ORR (3-month CR) of 64.4% (60%) with Flu/Cy at a median follow-up of 22 months and 64.4% (51.1%) with Benda (ORR: p=1; CR: p=0.39) at 10.9 months. The duration of CR was also similar (respondents with 6-month duration of response; Flu/Cy 67.6%, Benda 73.3%; p=0.63). No difference was noted in progression-free survival (PFS: median PFS in months/percent PFS at 6-months; Flu/Cy: 16.4/62.2%; Benda: 8.5/58.2%; p=0.37) or overall survival (median OS in months/percent OS at 6-months; Flu/Cy: not reached (NR) /86.7%; Benda: NR/79.9%; p=0.835).

We evaluated cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS). No differences in CRS and ICANS rates were noted between the groups (Flu/Cy: CRS 79.1%; G≥3: 2.3%; ICANS 42.2%; G≥3: 15.6% - Benda CRS:71.1%; G≥3: 4.4%; ICANS: 28.9%; G≥3: 4.4%; p>0.05). Also, no difference was noted when stratifying by CAR-T product (axi-cel: Flu/Cy: CRS 81.8%; G≥3: 3%; ICANS 47.1%; G≥3: 20.6% - Benda CRS:81.8%; G≥3: 6.1%; ICANS: 30.3%; G≥3: 6.1%; p>0.05). CRS and ICANS occurred sooner in Flu/Cy (median days to CRS peak: 3 [1-7 and ICANS peak: 6.5 [3-10]) than Benda (CRS: 5 [0-11], p<0.01; ICANS: 9 [3-16], p<0.05).

We evaluated ANC, PLT and Hb nadirs and cytopenia grades for each LD. Flu/Cy resulted in median ANC and PLT nadirs (x10⁹/L) of 0 [0 - 0.9] and 43 [7 - 193], and a Hb nadir (g/dL) of 7.8 [5.6 - 12.5]. Benda led to ANC and PLT nadirs of 0.8 [0 - 5.2] (p<0.001) and 100 [3 - 265] (p=0.001), respectively, and a Hb nadir of 10.2 [3.9 - 12.7] (p<0.001). Flu/Cy resulted in G≥3 neutropenia in all patients, G≥3 anemia in 54.5% and G≥3 thrombocytopenia in 57.5%. The respective values for Benda were 73% (p<0.001), 24.4% (p=0.01), and 29.7% (p<0.05). Flu/Cy led to febrile neutropenia in 46.7%, G-CSF use in 8.9%, red blood cell transfusion in 42.2%, and PLT transfusion in 26.7%, while the respective values for the Benda group were 17.8% (p<0.001), 0% (p<0.05), 20% (p=0.053), and 8.9% (p<0.05).

This study has limitations. It is retrospective with data from a single institution. The small sample size limits the statistical power. We aimed to match subject characteristics, but propensity matching has limitations especially in such a diverse group of subjects. The shorter follow-up for the Benda may impact accurate survival outcomes reporting. Additionally, the study lacks data on CAR-T expansion.

In conclusion, Benda is a viable alternative LD for CAR-T therapy in r/r DLBCL. We observed a trend towards better RRs and OS with Flu/Cy, and a trend towards less CRS and ICANS with Benda, as previously reported. Regarding cytopenias, Benda resulted in lower rates of severe cytopenias and their associated complications. These findings could support the use of Benda in patients with poor performance status. Further research with larger cohorts and mechanistic studies are warranted to better define the use of Benda in place of Flu/Cy.

Shlomchik:Orca bio: Consultancy, Current holder of stock options in a privately-held company; Bluesphere bio: Consultancy, Current Employment, Current holder of stock options in a privately-held company, Patents & Royalties. Dorritie:Genmab: Research Funding; Hoffman La-Roche: Research Funding; Kite-Gilead: Research Funding; Janssen: Honoraria, Research Funding; BMS: Honoraria, Research Funding.