The Impact of Lymphodepleting Chemotherapy Dose Modifications on Axicabtagene Ciloleucel Expansion and Treatment Outcome in Patients with Relapsed or Refractory CD19-Positive Diffuse Large B-Cell Lymphoma

Background: Lymphodepleting chemotherapy (LDC) administered before chimeric antigen receptor (CAR) T-cell infusion enhances tumor reactivity. However, the criteria for adequate lymphodepletion remain incompletely understood. Studies emphasize LDC dose intensity but often overlook dosing for patients with organ dysfunction or obesity, resulting in dosing uncertainty. This study uses a mixed-method approach to assess real-world clinical practice for LDC dosing and evaluate the impact of dose modifications on absolute lymphocyte count (ALC) kinetics, CAR T-cell expansion, and clinical outcomes.

Methods: To ascertain LDC dosing practices, a questionnaire was distributed in February 2023 to all 12 “first- and second-wave” commissioned UK CAR T-cell providers delivering axicabtagene ciloleucel (axi-cel). Data were analyzed using descriptive statistics and thematic analysis.

The impact of dose modifications was evaluated through a retrospective, single-center, multiple-cohort study at the Bristol Haematology and Oncology Centre, UK. All adult patients receiving fludarabine and cyclophosphamide (Flu/Cy) LDC followed by axi-cel infusion for the treatment of relapsed or refractory diffuse large B-cell lymphoma (r/r DLBCL) between January 2019 and September 2023 were included. Subjects were divided into two cohorts based on the requirement of a dose modification, defined as an adjustment or omission in LDC from axi-cel license recommendations. Outcomes included the best response at three months, duration of response (DOR), progression-free survival (PFS), and overall survival (OS). Patients were censored at the last follow-up. Categorical variables were analyzed using tests of independence, continuous variables with non-parametric tests, the area under the curve (AUC) by trapezoidal rule, PFS and OS with Kaplan-Meier and Log-rank tests, and multivariable analysis with Cox regression.

Results: 91.7% (N=11) of providers completed the survey. All providers used Flu/Cy and practiced dose modification in obesity (45.5%), renal (100%), and hepatic impairment (36.4%). The degree of reduction and initiation thresholds varied considerably; 72.7% utilized institutional protocols. 27.3% considered high-risk clinical features indicative of poorly responding disease as preclusion for dose modification.

65 patients met the study inclusion criteria; 39 (60%) received full-dose LDC while 26 (40%) received dose-modified LDC. Baseline characteristics were comparable. Fludarabine and cyclophosphamide doses were adjusted in 26.2% (N=17) and 21.5% (N=14) of patients, respectively. Dose-modified patients received median cumulative doses of 73.5 mg/m² fludarabine and 1378.5 mg/m² cyclophosphamide, compared to the full dose of 90 mg/m² and 1500 mg/m². ALC kinetics were strongly correlated between cohorts (r_s(18)=.99; p<.001). No significant differences were observed in axi-cel expansion (median AUC 171.51 versus 31.14 cells.day/µL; p=.13). Best response (p=.65) and DOR (p=.73) were comparable. Log-rank analysis demonstrated poorer PFS (p=.04) and OS (p=.006) in patients receiving dose-modified LDC; hazard ratios (HR) were 1.85 (95% confidence interval (CI) 1.01-3.37; p=.05) and 2.59 (95% CI 1.28-5.23; p=.008), respectively. When modeled with covariates related to patient fitness and disease burden, the adjusted HR was 2.10 (95% CI 1.09-4.03; p=.03) for PFS and 4.70 (95% CI: 2.08-10.61; p<.001) for OS. Significant increases in mortality (p=.03) and non-relapse mortality (p=.009) were observed. Relapse mortality was comparable (p=.99).

Conclusion: This survey provides a national service evaluation for UK LDC dosing practice. Although dose modifications are common, the lack of evidence-based guidelines results in practice heterogeneity. Identifying optimal management strategies and standardizing practices is essential.

This study provides novel, real-world insights into the effects of modifying LDC in a clinically well-characterized cohort. While these modifications seem to balance changes in Flu/Cy pharmacokinetics, resulting in comparable lymphocyte and axi-cel kinetics and treatment response, they may be associated with significantly poorer survival. Given the UK’s national criteria-led approval pathway, findings from this study have broad applicability and may inform clinical decision-making when weighing therapeutic options for r/r DLBCL.