Hematology · Medical Literature
A distinct CAR-T cell phenotype mediates therapeutic response at limited doses.
Abstract
Chimeric antigen receptor (CAR) T-cell therapies are typically administered at high doses to maximize durable clinical responses. However, manufacturing constraints can limit the production of sufficient cells to achieve the intended dose. Although some patients experience durable responses after receiving lower CAR-T cell doses, the mechanisms underlying efficacy at these limited cell numbers remain poorly understood. To address this, we performed deep phenotyping of anti-CD19 CAR-T cell products and their corresponding leukapheresis starting materials from a phase I/II dose-escalation trial. We also report the primary and secondary clinical outcomes of the diffuse large B-cell lymphoma, follicular lymphoma, and mantle cell lymphoma cohorts of the HD-CAR-1 basket trial (NCT03676504; EudraCT 2016-004808-60). Patients responding to low-dose CAR-T therapy showed dose-dependent enrichment of functional effector and effector memory-like CAR-T cells. The absolute number of these cells emerged as a robust biomarker of therapeutic response, valid across dose levels and CAR-T targets. Effective low-dose products were associated with high T-cell numbers and T-cell-supportive myeloid states in leukapheresis materials, whereas regulatory myeloid states promoted dysfunctional CAR-T cells and treatment failure. Our study provides insights into the phenotype, mechanisms, and biomarkers of CAR-T cells that drive therapeutic responses at low doses, with medical and socioeconomic implications.
Clinical relevance
CAR-T cell therapy efficacy at sub-standard doses is poorly understood, yet manufacturing limitations frequently prevent patients from receiving full intended doses — a critical unmet need in hematologic malignancies including DLBCL, follicular lymphoma, and mantle cell lymphoma. This phase I/II basket trial (HD-CAR-1; NCT03676504) provides mechanistic insight into why a subset of patients achieves durable responses despite receiving low-dose anti-CD19 CAR-T products. These findings have immediate relevance for patient selection, product release criteria, and resource allocation in centers with constrained manufacturing capacity.
Key findings
1. Patients responding to low-dose CAR-T therapy demonstrated dose-dependent enrichment of functional effector (Teff) and effector memory-like (Tem) CAR-T cell subsets within their infusion products, identifying a phenotypically distinct responder population. 2. The absolute number of functional Teff/Tem CAR-T cells — rather than total CAR-T cell dose alone — emerged as a robust biomarker of therapeutic response, with predictive validity confirmed across multiple dose levels and different CAR-T target antigens. 3. Deep phenotyping of leukapheresis starting materials revealed that high baseline T-cell numbers and T-cell-supportive (non-regulatory) myeloid states were associated with manufacturing of effective low-dose CAR-T products. 4. Regulatory myeloid states in leukapheresis material were associated with dysfunctional CAR-T cell phenotypes in the final product and correlated with treatment failure, implicating the pre-collection immune microenvironment as a determinant of product quality. 5. Clinical outcomes across DLBCL, follicular lymphoma, and mantle cell lymphoma cohorts of the HD-CAR-1 basket trial are formally reported, providing multi-histology efficacy and safety data for this dose-escalation design. 6. Note: Specific quantitative effect sizes, confidence intervals, p-values, and NNT/NNH were not available in the provided abstract; the above findings reflect the directional and mechanistic conclusions as reported.
Practice implications
Clinicians should recognize that absolute Teff/Tem CAR-T cell count in the infusion product — not total cell dose — is a more informative biomarker of likely response, and incorporation of functional subset enumeration into product release criteria should be considered prospectively. Pre-collection leukapheresis immunophenotyping (T-cell counts and myeloid subset characterization) may identify patients at high risk of manufacturing failure, enabling earlier intervention strategies such as bridging immunomodulation, donor selection optimization, or enrollment in alternative therapeutic pathways. These findings support a paradigm shift toward quality-over-quantity dosing in CAR-T manufacturing, with potential to expand access and reduce costs by validating lower-dose products in appropriately selected patients.