Evaluating the Effector Function of Individual CD19-Specific T Cells to Assess the Therapeutic Impact of a Manufactured Product

Adoptive cell therapy, based on the adoptive transfer of T cells genetically modified to enforce expression of a chimeric antigen receptor (CAR) has shown considerable promise in clinical trials, such as with CD19⁺ malignancies refractory to conventional treatments. However, efficiently predicting which patient may benefit from a given CAR design remains a challenge and undermines the therapeutic potential of such genetically modified T cells. Our approach to manufacturing CD19-specific T cells is based on electro-transfer of DNA plasmids from the Sleeping Beauty transposon/transposase system and selective outgrowth of CAR⁺ T cells on irradiated artificial antigen presenting cells (aAPC). The clinical-grade T cells prepared for infusion comprise a heterogeneous mixture and likely only a fraction of the genetically modified product contributes to therapeutic efficacy. Thus, we are developing an approach whereby the therapeutic potential of individual CAR⁺ T cell can be assessed. T cells are capable of different anti-tumor effector functions including cytotoxicity, cytokine secretion, and homing to target tissues, and currently no methodologies exist that can inform on all of these functions for a given (single) cell. Therefore, we have implemented an approach to interrogate the complexity of bulk populations of manufactured CAR⁺ T cells, so that the contribution of individual cells to the overall anti-tumor effect can be ascertained. CD19-specific CAR⁺ T cells currently being used in clinical trials were analyzed at a single cell level using technology that combines functional screening (cytotoxicity, cytokine production) with their molecular profiles. Bulk T cells which lysed tumor targets at 28% (1:1; effector:target ratio) in a standard 4 hour chromium lysis assay were observed at the single level to lyse multiple targets and function as serial killers (Figure). There was no correlation between IFN-g production and cytolysis. T cells that participated in killing and/or cytokine secretion were more susceptible to apoptosis than lymphocytes that did not participate in effector function. These data highlight that deconstructing the heterogeneity of a bulk population into individual cells can be used to determine whether subsets of T cells can participate in serial killing. The ability to serially image T-cell killing in massively parallel arrays will enable investigators to design CAR molecules that can support multi-hit kill kinetics. This lays the ground work for more complex studies evaluating cytokine production and genetic signatures of individual T cells bearing CAR species that comprise the infusion product. 

 

Figure: Mean serial killing of B-cell tumor (CD19^pos NALM-6) by CD19-specific CAR specific T cells (n=3) using single cell serial imaging.