Superior Efficacy of CAR-T Cells Using Marrow-Infiltrating Lymphocytes (MILs^TM) As Compared to Peripheral Blood Lymphocytes (PBLs)

Background

The type of T cell used in generating chimeric antigen receptor (CAR) T cells is an important choice. Evidence suggests that T cells that are early in the effector/memory differentiation pathway with more stemness and greater potential to persist are better than more differentiated T cells with less stemness that are more readily exhausted and have less potential to persist.

Marrow-infiltrating Lymphocytes (MILs^TM) is a novel form of adoptive T cell therapy composed of patient-autologous, polyclonal CD4 and CD8 T cells that are activated and expanded from the bone marrow. Genetically unmodified MILs^TM have demonstrated antitumor activity in patients with multiple myeloma and are being developed for several other tumor types, including non-small cell lung cancer and other solid tumors. Distinguishing features of bone marrow T cells used to produce MILs^TM include their memory phenotype, inherent tumor antigen-specificity, higher CD8:CD4 ratio and ability to persist long-term when compared to peripheral blood lymphocytes (PBLs) which is the T cell source used to produce currently approved CAR-T therapies. Based on these differences, we hypothesize that MILs^TM provide a more robust and better fit platform for CAR-T therapy compared to PBLs.

Using a CD38-specific, 4-1BB/CD3z-signaling CAR as an initial model, we have demonstrated the feasibility of producing CAR-modified MILs^TM (CAR-MILs^TM) and showed that CAR-MILs^TM demonstrate superior killing in vitro compared to CAR-T cells generated from patient-matched PBLs (CAR-PBLs). Herein, we build on our previous data and add a second BCMA-specific CAR model. We use the two multiple myeloma model systems to compare cytolytic potential, functionality, and expression of phenotypic markers of memory, stemness and exhaustion between patient-matched CAR-MILs^TM and CAR-PBLs.

Methods

Matched pairs of CAR-MILs^TM and CAR-PBLs were produced from the bone marrow and blood of multiple myeloma patients. Two different in vitro cytotoxicity assays, the RTCA xCelligence real-time impedance and FACS assays, were used to evaluate antigen-specific killing of target tumor cells. Functionality of CD4 and CD8 CAR-T cells, at the single-cell level, was evaluated by measuring the secretion of 32 cytokines and chemokines following in vitro antigen-specific stimulation using IsoPlexis IsoCode chips and analyzed using IsoPeak. Expression of markers of T cell memory (CD45RO & CCR7/CD62L), stemness (CD27) and exhaustion (PD1 & TIM3) on CAR-MILs^TM and CAR-PBLs prior to and following antigen-specific stimulation was evaluated by flow-cytometry (FACS).

Results

CAR-MILs^TM demonstrated superior killing of tumor target cells in vitro, regardless of the antigen specificity of the CAR, when compared to matched CAR-PBLs and this superiority persisted even upon repeated antigen encounter - a factor that may be critical in guaranteeing better anti-tumor efficacy and persistence. CAR-MILs^TM demonstrated increased polyfunctionality (secretion of 2+ cytokines per cell) and an increased polyfunctional strength index (PSI) following antigen-stimulation compared to CAR-PBL in both CD4 and CD8 T cells. The enhanced PSI in CAR-MILs^TM was predominately mediated by effector, stimulatory and chemoattractive proteins associated with antitumor activity including Granzyme B, IFNg, IL-8, MIP1a and MIP1b. Coincidentally, increased PSI and enhanced secretion of these same proteins was reported to be associated with improved clinical responses in patients with Non-Hodgkin lymphoma treated with CD19-specific CAR-T therapy. Expression of memory markers on CD4 and CD8 T cells were similar in CAR-MILs^TM and CAR-PBLs both prior to and following antigen-stimulation. Although expression of CD27, PD1 and TIM3 were similar at baseline, CAR-MILs maintained higher levels of CD27 and lower levels of PD1 and TIM3 compared to CAR-PBLs following antigen-stimulation in both CD4 and CD8 T cells.

Conclusions

Collectively, our data suggest that CAR-MILs^TM have several advantages over CAR-PBLs, including increased cytolytic potential, enhanced polyfunctionality, increased stemness and less exhaustion. Based on these differences and the inherent antitumor properties of MILs^TM, we speculate that CAR-MILs^TM would be more potent and effective than currently approved CAR-T products derived from PBLs.