Deciphering the Impact of Immunological Aging in CLL CAR T Cell Therapy

Chronic Lymphocytic Leukemia (CLL) primarily affects older adults, with a median diagnosis age range of 65-72 years. CAR T cell therapy shows high complete response (CR) rates >80% in pediatric acute lymphoid leukemia, yet only under 30% in CLL. CAR T cells in non-responding (NR) CLL patients upregulate pathways for effector differentiation and exhaustion. However, immune checkpoint blockade doesn't change CLL progression, suggesting exhaustion alone doesn't explain CLL T cell deficiency. We hypothesize that immunosenescence also contributes to CAR T cell defects in CLL.

19CAR-BBz T cells were generated from 11 treatment naïve CLL patients and age- and gender-matched healthy controls with a median age of 67. Immunophenotyping was performed with flow cytometric panels including 25 markers. We conducted a chronic antigen stress test: every five days, 1E6 CAR T-cells were counted, phenotyped, and stimulated with fresh CD19-expressing K562 cells at a 1:1 ratio. 24 hours after co-culture, supernatant was collected for cytokine analysis using LEGENDplex™ kits. Senescence signatures and clinical characteristics in CLL patients enrolled in CAR T cell trials were contrasted using ssGSEA analysis of bulk RNA sequencing data.

CAR T cells derived from treatment-naïve CLL patients exhibit a decrease in co-stimulatory receptors CD27 and CD28 expression alongside heightened levels of senescence markers p16, p21, p53, β-galactosidase and γ-H2AX. The ex vivo manufacturing process solidifies senescent trends seen in baseline CLL T cells compared to healthy control. CD28 is instrumental in predicting CLL from control in a random forest prediction model. In functional assays, the proliferative capacity of CLL patient derived CAR T cells was attenuated and production of IFN-γ was decreased. However, CLL-derived CAR T cells showed signatures of inflamm-aging such as increased expression of Granzyme A, B and perforin and production of pro-inflammatory factors such as sFas and IL-1β. In the clinical trial subjects, several hallmarks of immunosenescence were found in T cells from NR CLL patients at leukapheresis and ex vivo manufacturing harvest such as the loss of co-stimulatory receptors CD27/CD28, KLRG1 upregulation and decrease in TCR diversity. Enrichment of senescence pathways is also able to differentiate therapeutic outcome, and correlates negatively with expansion during CAR T cell manufacturing, maximum in vivo CAR T cell expansion and overall survival of patients. Furthermore, a machine learning model developed by Lu et al. (Nature Communications, 2022) predicted higher age in NR CLL patients compared to CR CLL patients based on their CAR T cell transcriptome.

Our study underscores the role of immunosenescence in reducing CAR T cell efficacy in CLL. Senescence signatures such as CD27/CD28 loss and enrichment in senescent pathways should be considered for patient stratification to enhance therapeutic success. Addressing immunosenescence is critical for improving CAR T cell therapy in CLL.