Abstract 4236: A subtype-specific T-cell transcriptomic atlas reveals determinants of T-cell dysfunction in CAR T-cell therapy resistance

Abstract

Abstract Chimeric Antigen Receptor (CAR) T-cell therapy has been a major success for B-cell Acute Lymphoblastic Leukemia (B-ALL), yet recent clinical trials suggest that approximately 20% of patients do not respond to therapy and 40% of initial responders relapse within a year. We hypothesize that intrinsic T-cell factors are a major contributor to early therapeutic failure to CAR T-cell therapy. In particular, since most protocols do not select for T-cell subtypes, the heterogeneity in T-cell composition likely plays a major role in CAR T-cell engineering and confounds functional and clinical correlative studies. In order to characterize the molecular phenotypes of T-cells underlying therapeutic response, we performed subtype-specific transcriptomic profiling of clinical T-cell populations in 71 patients treated with CAR T-cell therapy at the Children's Hospital of Philadelphia. We obtained clinical T-cells via clinical leukapheresis prior to CAR T-cell engineering, used FACS to separate T-cells into five subtypes per patient (Naive, Stem Cell Memory, Central Memory, Effector Memory, and Effector), and performed RNA-Sequencing on all 355 samples. To our knowledge, this is the largest and most comprehensive transcriptomic profiling of clinical T-cells in CAR T-cell therapy to date. We found that a higher proportion of CCR7+CD62L+ naive and early memory T-cell phenotypes associates with longer clinical T-cell persistence (HR: 3.25 [1.37-7.89], P = 0.006), and that combining proportions of these populations with CCR7-CD62L-CD45RO+CD95+ effector memory T-cells has the greatest prognostic value (HR: 4.82 [1.89-12.32], P < 0.001). Apoptotic signaling pathways were enriched in effector and effector memory T-cell lineages (FDR < 0.001), suggesting that programmed cell death limits the contribution of these phenotypes to long-term CAR T-cell persistence. Maintenance of naive and early memory T-cell phenotypes associates positively with expression of transcription factors TCF7 and LEF1 (FDR < 0.001) and negatively with TBX21, EOMES, and ZEB2 (FDR < 0.001). Within T-cell subtypes, mixed-effects regression and integrative network analysis identified genes and network modules that discriminate between clinical responders and non-responders. We found that higher expression of CRTAM and SLAMF6 in naive and stem cell memory T-cells associates positively with clinical CAR T-cell persistence (FDR = 0.0088-0.019), suggesting that these lineage-determining factors direct T-cell fates toward effective CAR T-cell lineages. Finally, we identify an enrichment of type I interferon response pathways among patients with poor CAR T-cell persistence (F-test P < 0.001), suggesting a role of the chronic interferon response network in T-cell dysfunction. Together, these data shed light on the critical role of pre-infusion T-cell populations in CAR T-cell therapy, and may inform clinical prognosis and the development of improved CAR T-cell therapies. Citation Format: Gregory M. Chen, Changya Chen, Rajat K. Das, Yang-Yang Ding, Bing He, Hannah Kim, David M. Barrett, Kai Tan. A subtype-specific T-cell transcriptomic atlas reveals determinants of T-cell dysfunction in CAR T-cell therapy resistance [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 4236.