Higher levels of circulating exhausted T-cells at pre- and post-infusion negatively associated with clinical efficacy in tisagenlecleucel (Tisa-cel) treated relapsed/refractory large B-cell lymphoma (r/r LBCL) patients (pts) in JULIET trial

Abstract Background: JULIET, a pivotal phase II trial of Tisa-cel, an anti-CD19 CAR-T cell therapy in pts with r/r LBCL demonstrated durable activity with a manageable safety profile (Schuster et al., 2021). We sequentially characterized peripheral blood T cell functionality by immunophenotyping and correlated with efficacy endpoints. Methods: Immunophenotypic characterization of both non-CAR-T and CAR-T cells was performed by flow cytometry in 378 blood samples from 111 pts collected at enrollment and within the first month post Tisa-cel infusion. Proportions of T cell subsets were identified by maturation (CCR7/CD45RA/CD45RO) and exhaustion (PD1/LAG3/TIM3) markers and then correlated with durable clinical response (Complete Response (CR) ≥ 6 mo vs. Non-response (NR)/Relapse), PFS and OS. Results: At enrollment, low T cell levels (<14.5% of PBMCs or <170 cells/uL) observed in 30% (31/105) pts were associated with NR/relapse (27/31) and with significantly shorter PFS and OS, compared to those with higher levels (median (m)PFS: 2.1 vs. 9 mo, p=0.0006; mOS: 6 vs. 28.5 mo, p=0.003). NR/relapse pts had higher levels of exhausted T cell subsets in both CD4 (LAG3+, LAG3+PD1+) and CD8 (PD1+, LAG3+, LAG3+PD1+) cells at enrollment. High levels (>median) of these subsets also had significantly shorter PFS and OS, with LAG+PD1+CD8+ ranked on top (mPFS: 2 vs. 34 mo; mOS: 4.6 mo vs. not reached, both p<0.0001) and remained significant in multivariate Cox models including LDH and Myc status (PFS: HR=3.1 [1.9 - 5.3]; OS: HR=2.8 [1.6 - 4.8]; both p<0.001). Hierarchical clustering showed concordant increase of all non-CAR-T exhausted subsets in NR/relapse pts, separating them from ongoing CRs. The trend continued at day 28 (D28), with almost all exhausted CD4 and CD8 subsets in both non-CAR-T and CAR-T cells being consistently higher in NR/Relapse compared to ongoing CRs. In addition, high levels of exhausted subsets in non-CAR-T cells at D28 were associated with shorter PFS and OS, with the strongest trend in PD1+, LAG3+, LAG3+PD1+ of CD8 and CD4 cells. Survival analyses for CAR-T exhausted subsets were less informative due to small sample size. Of note, several LAG3+ and LAG3+PD1+ subsets of CAR-T and non-CAR-T cells were also higher in NR/relapse at CAR-T peak expansion. In contrast, no obvious differences were observed for T cell memory subsets from both CAR-T and non-CAR-T cells across multiple timepoints. Conclusions: Decreased circulating T cells at enrollment, and increased circulating exhausted T cells in non-CAR-T (at enrollment and D28) and CAR-T (at D28) compartments, were associated with poor response and survival, suggesting these flow measurements may serve as surrogate for overall T cell function, which is critical for optimal Tisa-cel efficacy and early prognostication in LBCL. Citation Format: Ulrich Jäger, Xia Han, Davide Germano, David Quinn, Dalia Gaddis, Aisha Masood, Tanya Mulvey, Jennifer Mataraza, Michael R. Bishop, Gilles Salles, Richard Maziarz, Stephen Schuster. Higher levels of circulating exhausted T-cells at pre- and post-infusion negatively associated with clinical efficacy in tisagenlecleucel (Tisa-cel) treated relapsed/refractory large B-cell lymphoma (r/r LBCL) patients (pts) in JULIET trial [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 5186.