Researchers at The University of Texas MD Anderson Cancer Center have identified two distinct mechanisms governing whether allogeneic CAR T cells are rejected or expand productively in patients — findings with direct implications for the design of off-the-shelf cell therapies, according to a study published in Cancer Discovery.
Allogeneic CAR T therapies are manufactured in advance from healthy donors and can be deployed immediately, unlike autologous approaches requiring patient-specific manufacturing. A central obstacle has been inconsistent outcomes after infusion — cells are sometimes cleared by the host immune system before exerting therapeutic effect, and the determinants of rejection versus productive expansion have remained poorly understood.
The study, led by corresponding author Pavan Bachireddy, analyzed 11 patients with large B-cell lymphoma (LBCL) treated with a single manufacturing lot of cemacabtagene ansegedleucel (cema-cel), an allogeneic anti-CD19 CAR T product under development by Allogene Therapeutics. Using a single lot across all patients reduced product-level variability and helped isolate recipient- and cell-level determinants.
On the host side, patients who failed to expand cema-cel had high pre-existing frequencies of alloreactive CD8+ T cells that rapidly cleared the infused product. On the product side, among patients where rejection did not occur, robust clonal expansion was driven by effector-like features within the cells — not by stem or central memory phenotypes, which had been assumed to be optimal for persistence.
The expansion-associated findings were reproduced in two independent cohorts: one receiving a separate cema-cel manufacturing lot, and one receiving Allogene's allogeneic anti-BCMA CAR T candidate ALLO-715 from a completed Phase I study (NCT04093596). The replication across a different product lot and a distinct construct targeting a different antigen strengthens the generalizability of the findings.