Single-cell multi-omics reveals type-2 function in CAR T cells associated with 8-year leukemia remission

Introduction: CD19-directed chimeric antigen receptor (CAR) T cell therapy has proven to be remarkably effective in treating relapsed or refractory (r/r) acute lymphocytic leukemia (ALL) in pediatric and young adult patients. Although early-phase trials have demonstrated an exceptional initial response rate, relapse has been frequently reported with varying patterns and approximately half of the patients fail to sustain event-free survival one year post-CAR T infusion, highlighting an emerging unmet need to investigate the molecular intricacies that govern long-term durable remission. Single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool to uncover the characteristic profile of CAR T cells, facilitating our recent identification of CAR T signatures mediating decade-long remission in chronic lymphocytic leukemia. However, until now, single-cell studies aiming to reveal the determinants of CAR T persistence have been constrained by small cohort sizes and limited cell numbers, thus impeding the faithful elucidation of deep biological mechanisms.

Materials and

Methods: To investigate the molecular determinants of ultra-long CAR T persistence, combining scRNA-seq and CITE-seq, we extensively sequenced more than one million pre-infusion CAR T cells at the basal level or upon CAR-specific stimulation from 82 pediatric ALL patients enrolled in the first two CAR T ALL clinical trials and 6 healthy donors (Fig.1a). Patient demographics were collected from September 2012 to July 2022, with a follow-up duration extending up to 106 months at the endpoint. To uncover the hallmarks of CAR T longevity, we correlated the single-cell multi-omics profiles with the duration of B-cell aplasia (BCA), a widely used pharmacodynamic measurement indicative of CAR T persistence. Consequently, we classified all patients into 5 persistence groups based on their BCA duration (Fig.1b). In our cohort, five patients have demonstrated a median BCA duration of 8.4 years (BCA-L group), representing a unique opportunity to identify molecular determinants of CAR T longevity. Patients in the BCA-O group continued to have BCA at the last data collection, with a median duration of 5.1 years. To mitigate potential confounding variables related to trial design, 42 patients including the 5 BCA-L patients from NCT01626495 were analyzed as the Discovery Cohort, while the other 40 patients including the 11 BCA-O from NCT02906371 constituted the Validation Cohort.

Results, Conclusions, and Discussions: Our analysis of CD19-specific stimulated CAR T cells revealed a prominent role of type-1 function, which was highly represented but had no discernible correlation with long-term persistence (Fig.1c). Unexpectedly, we identified that elevated type-2 functionality in CAR T infusion products is significantly associated with BCA-L and BCA-O patients maintaining ultra-long-term remission (Fig.1c). Higher type-2 cytokine level of long-term persistent CAR T cells was independently validated using both flow cytometry and multiplexed secretomic assay (Fig.1d, e). Analysis of ligand-receptor interactions unveiled that type-2 cells regulate a dysfunctional subset to maintain the whole population homeostasis (Fig.2a-h), and the addition of IL-4 during antigen-specific activation alleviates CAR T cell dysfunction while enhancing fitness at transcriptomic level (Fig.2i-l). To investigate the response post-CAR T infusion, we performed a comprehensive proteomic profiling of post-infusion serum proteins in a total of 41 patients (Fig.3a), revealing a higher level of circulating type-2 cytokines in 5-year or 8-year relapse-free responders (Fig.3b, c). Given the higher type-2 immunity observed in ultra-long persistent CAR T cells, our next inquiry was whether type-2 high CAR T cells could elicit enhanced antitumor efficacy in vivo (Fig.4a). In a leukemic mouse model, type-2 high CAR T products demonstrated superior expansion and antitumor activity particularly upon leukemia rechallenge (Fig.4b-g). Finally, we examined whether enhancing type-2 functionality could elevate the performance of the type-2 low CAR T sample. We proposed and evaluated two strategies to address this (Fig.5a), considering two scenarios: (1) adding IL-4 into the culture medium throughout the manufacturing process for products that have not yet been generated, and (2) priming CAR T products with IL-4 if re-manufacturing may not be feasible for practical clinical applications. Through rigorous in vivo studies utilizing a leukemia mouse model with follow-up exceeding 100 days and subjected to two tumor rechallenges, both methods proven to be highly effective in boosting the antitumor response of Type-2 Low CAR T cells (Fig.5b-d). Our findings provide key insights into the mediators of durable CAR T therapy response and suggest potential therapeutic strategies to sustain long-term remission by boosting type-2 functionality in CAR T cells.

Acknowledgements (Optional):