OA-06: Robust in vivo expansion and long-term persistence of anti-BCMA CAR T cells, PHE885, manufactured in <2 days on the T-Charge platform

Introduction: BCMA targeted CAR-T cell therapy has clinical benefits for patients with r/r MM, but long manufacturing times and poor in vivo persistence remain unresolved problems. The T-charge platform is a novel rapid manufacturing process that reduces manufacturing time to < 2 days and preserves less-differentiated T cells in the CAR-T product. Here, we present correlative data from the ongoing Phase I trial in r/r MM (NCT04318327) of PHE885, a fully human product manufactured using the T-Charge platform.

Methods: We studied serial samples from 32 patients who received PHE885 at the Dana-Farber Cancer Institute. PHE885 manufacturing and CAR T cell persistence were assessed in apheresis samples (APH), final product (FP), and peripheral blood (PB) following infusion using flow cytometry and CyTOF. Bulk T cell receptor (TCR) sequencing was used to monitor PHE885 repertoire diversity over time. Serum cytokines and soluble BCMA were also measured.

Results: We have previously reported a 98% overall response rate (ORR) across all dose levels (2.5-20x106 CAR-T cells) and 100% ORR at >5x106 cell dose. Here we report immune assessment of the product and after in vivo expansion. The proportion of stem-like memory T cells (TSCM) in FP was preserved following manufacturing. CyTOF and TCR sequencing revealed that TSCM in FP had a highly activated and proliferative phenotype and significantly higher TCR repertoire diversity than memory T cells. PHE885 CAR T cells expanded rapidly after infusion reaching median peak levels of 3,118 cells/ul (range 373 to 17,865) in PB at a median of 14 days (range 10 to 27) after infusion. Cytokine release syndrome (CRS) occurred with expansion of CAR-T cells and was associated with increased serum levels of IFNg, TNFa, and IL-6. While the peak CAR-T expansion was not associated with the dose of PHE885, the frequency of TSCM in FP positively correlated with the expansion of CAR-T on day 14 after infusion. CAR-T cells at day 14 after infusion were predominately effector memory (TEM) and exhibited a highly activated and proliferative phenotype. Following peak expansion there was emergence of effector memory cells expressing CD45RA (TEMRA) over time. CAR-T cells during peak expansion in vivo had higher levels of TCR repertoire diversity than non-CAR-T cells, and post-infusion CAR-T cells shared significantly more TCR clonotypes with TSCM than with memory T cells in FP, suggesting that the highly heterogeneous CAR-T clones at peak expansion were derived from TSCM clones in FP. High levels of PHE885 persistence were observed with transgene detectable by qPCR in 67% of patients, and >10% of CD3+ T cells by flow cytometry in 39% of patients at 6 months. In patients with long-term persistence, CAR-T cells maintained a highly diverse TCR repertoire.

Conclusions: The T-charge rapid manufacturing platform successfully preserved TSCM clones in the FP leading to a highly heterogeneous and proliferative CAR-T expansion with durable persistence.