Elsa Hélène Smith 1,2, Fenja Klima 1,2, Matthias Anzböck 3, Alessandra Holzem 4,5, Marion Subklewe 4,5, Oliver Scherf-Clavel 3, Wilhelm Huisinga 2,6, Charlotte Kloft 1,2
1 Freie Universität Berlin, Institute of Pharmacy, Department of Clinical Pharmacy and Biochemistry (, Germany), 2 Graduate Research Training Program PharMetrX (Berlin/Potsdam, Germany), 3 Department of Pharmacy, Ludwig-Maximilians-Universität München (, Germany), 4 Department of Medicine III, University Hospital, Ludwig Maximilian University Munich, Munich (, Germany), 5 German Cancer Consortium (DKTK), Partner Site Munich, Munich (, Germany), 6 Institute of Mathematics, University of Potsdam (, Germany)
Introduction/objectives:
Chimeric antigen receptor (CAR) T-cell therapy is an immunotherapy that has transformed the treatment of haematological malignancies by enabling the patient’s immune system to eliminate tumour cells. While CAR-T cells induce deep and durable remissions and improve treatment outcomes, a substantial proportion of patients fail to respond to treatment or experience relapse [1]. A key, yet understudied factor contributing to treatment efficacy, is the lymphodepleting regimen administered before CAR-T cell infusion [2,3,4], most often comprising cyclophosphamide and fludarabine in a variety of conditioning regimes.
Lymphodepletion promotes CAR-T cell expansion, persistence and clinical efficacy by depleting endogenous lymphocytes, enhancing cytokine concentration, and fostering a pro-inflammatory microenvironment [5,6,7]. Gaining insights into the pharmacokinetics (PK) and pharmacodynamics (PD) of lymphodepleting agents and factors influencing their variability, is therefore of major interest. Pharmacometric modelling and simulation enables the analysis of dose-exposure-response relationships at the population and individual level, thus providing a suitable framework for improving understanding of the PK and PD of lymphodepleting regimens [8]. This project aims to review and compare published population PK models for fludarabine and cyclophosphamide as well as PD modelling approaches describing cell-depleting effects of chemotherapies. Ultimately, this work seeks to support optimisation of conditioning strategies to enhance CAR-T therapy outcomes.
Methods:
A literature search was conducted in the PubMed database following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) guidelines to identify published PK models of fludarabine and cyclophosphamide. Studies were excluded if they did not present a clinical PK model of the drugs, reproduced previously published models without modifications, or were not available as full-text articles. To investigate the PD effect of lymphodepletion, a broader literature review was conducted to identify PD models of lymphodepletion and general modelling approaches for cell depletion, using key words related to the lymphodepleting agents and their PD effects. Studies were excluded if they lacked a PD model or were unrelated to oncology or CAR-T therapy settings.
From each eligible study, data was extracted to enable model comparison, including bibliographic information, study population demographics (sample size, sex, age, and body weight), clinical indication, dosing and sampling details (single dose, infusion duration, sampling times, matrix and metabolites analysed), model characteristics (model type, structural model, and evaluated covariates), and key PK/PD input and output requirements.
Results:
The PK model review identified 10 PK models for fludarabine and 30 for cyclophosphamide. Across studies, variability was observed in study population demographics, clinical indications, dosing strategies, and model characteristics. Several models were developed for specific populations, including paediatric, adult, elderly or female-only populations. Models were mostly established in the context of haematopoietic stem cell transplantation conditioning or chemotherapy for a range of malignancies. Only two studies addressed conditioning prior to CAR-T therapy, including one kinetic-pharmacodynamic (KPD) model describing the combined effects of fludarabine and cyclophosphamide and one nonlinear mixed-effects (NLME) PK model focusing on fludarabine. Nonetheless, despite these differences, structural models were comparable, with for instance most fludarabine models being two compartmental, indicating a potential translatability to the setting of CAR-T therapy.
Given the scarcity of PD models of lymphodepletion related to fludarabine and cyclophosphamide, the scope of the review was expanded to include general modelling approaches for cell depletion. The PD models exhibited heterogeneity in population demographics, disease characteristics, PD metrics modelled, data requirements, and modelling approaches with the majority being mechanistic models. Altogether, these studies used different strategies to model the effect of cytotoxic and lymphodepleting agents on immune cell populations, and provided suitable frameworks depending on the focus of specific analyses.
Conclusion:
The literature provides a solid foundation for the PK modelling of fludarabine and cyclophosphamide and outlines approaches for modelling immune cell depletion and the types of data that are required. However, PK and PD models developed for lymphodepletion in the context of CAR-T therapy remain limited. To better understand factors driving effective lymphodepletion, future work will focus on adapting existing PK and PD models to this therapeutic context. Linking the PK/PD relationship of fludarabine and cyclophosphamide to CAR-T cell expansion and efficacy could provide insights. While PK/PD modelling of lymphodepletion is a powerful tool, its full potential has yet to be leveraged for tailoring lymphodepleting regimens, guiding dosing strategies, and ultimately improving the safety and effectiveness of CAR-T cell therapies.
References:
[1] A. Suchiita, S.C. Sonkar. Revolutionizing immunotherapy: the next frontier in CAR T-cell engineering. Crit. Rev. Oncol. Hematol. 211: 104751 (2025).
[2] A.V. Hirayama, J. Gauthier, K.A. Hay et al. The response to lymphodepletion impacts PFS in patients with aggressive non-Hodgkin lymphoma treated with CD19 CAR T cells. Blood 133: 1876–1887 (2019).
[3] M. Scordo, J.R. Flynn, S.M. Devlin et al. Population Pharmacokinetic Model Identifies an Optimal Fludarabine Exposure for Improved Outcomes after CD19-Directed CAR T Cell Therapy for Aggressive B-NHL: Analysis from the Cell Therapy Consortium. Blood 140: 1588–1591 (2022).
[4] S.S. Neelapu. CAR-T efficacy: is conditioning the key? Blood 133: 1799–1800 (2019).
[5] M. Canelo-Vilaseca, M. Sabbah, R. Di Blasi et al. Lymphodepletion chemotherapy in chimeric antigen receptor-engineered T (CAR-T) cell therapy in lymphoma. Bone Marrow Transplant. 60: 559–567 (2025).
[6] L. Amini, S.K. Silbert, S.L. Maude et al. Preparing for CAR T cell therapy: patient selection, bridging therapies and lymphodepletion. Nat. Rev. Clin. Oncol. 19: 342–355 (2022).
[7] B. Lickefett, L. Chu, V. Ortiz-Maldonado et al. Lymphodepletion – an essential but undervalued part of the chimeric antigen receptor T-cell therapy cycle. Front. Immunol. 14: 1303935 (2023).
[8] D.R. Mould, R.N. Upton. Basic concepts in population modeling, simulation, and model-based drug development. CPT Pharmacomet. Syst. Pharmacol. 1: e6 (2012).
Reference: PAGE 34 (2026) Abstr 12304 [www.page-meeting.org/?abstract=12304]
Poster: Drug/Disease Modelling - Oncology