OA2-AM24-TU-06 - Developing An Engineered T Cell Therapy To Eliminate Inhibitors In Hemophilia A

Hemophilia A is a bleeding disorder caused by deficiency of coagulation factor VIII (FVIII). Management often involves replacement with FVIII protein. About 10-30% of patients develop neutralizing anti-FVIII antibodies (inhibitors) that complicate their clinical management. Chimeric antigen receptor (CAR) T cell therapy has traditionally used an antibody-derived single-chain variable fragment to target T cells to surface antigens on malignant cells. We previously modified this platform using an autoantigen on the CAR to selectively target autoantibody producing B-cells. Here, we employ a similar strategy using FVIII domain to eliminate FVIII inhibitor producing B-cells.  

Study

Design/Methods:

We designed a panel of FVIII-based Chimeric AlloAntibody Receptors (CALLARs) employing different domains of FVIII. As a B cell model, the Nalm6 cell line was transduced to express a panel of inhibitory FVIII-specific surface IgGs. Human CALLAR-T cells were evaluated for efficacy, specificity, and toxicity, using Nalm6 target cells in vitro and in a xenograft mouse model. 

Results/Findings:

A lead CALLAR candidate comprised all of the FVIII domains except A1 and employed the CD28 and CD3z signaling domains (designated A2-LC-28z); this constitutes most of the immunodominant B cell epitopes of FVIII. A2-LC-28z CALLAR T cells robustly eliminated Nalm6 cells expressing FVIII inhibitors in vitro and in vivo (Fig. 1A). CALLAR T cells were effective in vitro even in the presence of soluble anti-FVIII antibodies. However, mice treated with A2-LC-28z CALLAR T cells developed rapid weight loss resulting in early death (Fig. 1B). The toxicity was found to be Nalm6 independent implicating an off-target interaction. Necropsy demonstrated infiltration of the human T cells into lung parenchyma. Additionally, mouse serum demonstrated markedly elevated human TNFa in A2-LC-28z treated mice, with no increase in other mouse and human cytokines tested. We investigated several potential off-target interactions based on known protein interactions with native FVIII. We have ruled out interaction of CALLAR T cells with von Willebrand Factor, factor IX, and LRP1. Additionally, treatment of mice with anti-TNFa using infliximab did not rescue mice.    

Conclusions:

These data demonstrate that while a FVIII-CALLAR is successful in targeting FVIII-inhibitor expressing cells, an off-target interaction results in toxicity. The underlying cause of the toxicity remains unknown. Strategies to identify the mechanism and mitigate toxicity are ongoing. Our findings underscore the unique complexity inherent to developing non-antibody, native protein-based immune-receptors for cell therapy.