Background Although chimeric antigen receptor (CAR)-T cells are a viable treatment option for patients with relapsed or refractory (r/r) aggressive B-cell lymphomas, more than half of patients experience relapse within the first two years after product infusion. Successful autologous cell-based anti-cancer therapies require persistence of effector cells, features that highly depend on complex metabolic processes. We earlier demonstrated in chronic lymphocytic leukemia, that expansion and persistence of CD19-CAR T-cell infusion in vivo was correlated with increased mitochondrial mass and enhanced biogenesis(van Bruggen et al, Blood 2019). Metabolic features might therefore serve as early biomarker tools to predict efficacy and outcome of CAR T-cell therapy and might allow therapeutic targeting but did was never studied in a prospective clinical trial. Aims Perform extensive phenotyping and state-of-the-art metabolic state analyses of T cells prior and post CAR T cell production and investigate correlations with clinical outcomes and patient characteristics. Methods Samples were obtained from large B cell lymphoma (LBCL) patients that were part of the TRANSCEND-NHL-001 trial (NCT02631044, Abramson et al, Lancet 2020).Thirty-two patients with r/r B-cell lymphoma who received CAR-T cell treatment lisocabtagene maraleucel (16 responders (R) and 16 non-responders (NR)) were included in our analyses and (metabolically) analyzed by flow cytometry and extracellular flux analysis. Metabolic phenotyping was performed on 53 samples at baseline and 44 samples after an in vitro stimulation. Extracellular flux analysis was done on 14 DP samples. The cohort was split evenly between Responders and Non-responders, and balanced for a score calculated taking key clinical baseline characteristics that were found to be associated with response into account (NCT02631044, Abramson et al, Lancet 2020). Both T cells obtained at time of leukopheresis (starting material (SMAT)) and CAR-T drug products (DP), (following transduction of a CD19-targeted CAR T cell product with a defined CD4 and CD8 composition bearing a 4-1BB costimulatory domain and expansion) were analyzed directly after thawing and after stimulation with anti-CAR-antibody in case of CDP, or αCD3/αCD28 antibodies for SMAT samples. Patients consented to these studies and researchers were blinded for clinical outcomes at time of analysis. Results We found that the baseline clinical factors that associate with clinical response are related to specific metabolic cellular phenotypes, both in SMAT and DP, and with and without T cell stimulation. SMAT samples from patients with higher baseline scores show signs of an increased exhausted phenotype with regards to cellular ROS and exhaustion marker expression, whilst a more plentiful memory-like phenotype was seen in patients with a lower baseline score. When clustering by clinical response status, analysis of glucose and fatty-acid uptake, neutral lipid content, cellular ROS, mitochondrial mass and potential and flux data showed that responders, in both SMAT and DP samples, have an all-round increased metabolic activity upon αCD3/αCD28 or CAR T-receptor triggering, respectively, when compared to non-responders (Figure 1A). Furthermore, responding patients show enrichment of CD27 expression in the SMAT material, indicating a more abundant memory phenotype. Summary/Conclusions This is the first evidence of an association between a specific immune-metabolic phenotype of (CAR-) T cells and outcome of CAR-T cell therapy in patients with relapsed/refractory LBCL. More specifically, a higher abundance of memory phenotypes, metabolic activity and mitochondrial biogenesis of both harvested T cells and CAR-T product associates with improved clinical response. Combining metabolic, phenotypical and clinical characteristics and biomarkers of patient T cells (both pre- or post-production) shows the importance of the metabolic phenotype and fitness of T cell starting material, underlying the importance of early line CAR T therapy following diagnosis.
Read full abstract