A Novel CD19-Anti-CD20 Bridging Protein Prevents and Reverses CD19-Negative Relapse from CAR19 T Cell Treatment In Vivo

Introduction

CAR T cells that recognize the antigen CD19 (CAR19s) have achieved remarkable success in otherwise untreatable B cell malignancies including refractory and relapsed ALL and NHL. However, clinical data from diverse CAR19 trials, and real-world experience with the approved CAR19 therapeutics (tisagenlecleucel and axicabtagene ciloleucel), highlight a critical issue, that of patient relapse due to the loss of expression of the target antigen (CD19) or the antigenic epitope. Antigen loss relapse rate of up to 50% have been reported across indications (adult ALL, pediatric ALL, adult NHL) irrespective of the specific CAR19 used. Attempts to treat patients who have relapsed from CAR19 treatment include provision of a CAR T cell to a second antigen, for example CD22. Such attempts have met with limited success, further, many patients cannot tolerate a second regimen of apheresis, consolidation, lymphodepletion and CAR T infusion.

Importantly, many of the patients relapsing with CD19-negative malignancies still have detectable levels of CAR19 T cells in circulation, since the CAR19s persist in the presence of normal B cells being produced by the bone marrow (these B cells are CD19-positive). Therefore, a technology that reactivates the patient-resident CAR19s to attack the relapsing tumor cell would be a highly attractive alternative to subsequent CAR T therapy. Here we present this technology and illustrate its' ability to prevent relapses and importantly, to reverse relapses in vivo.

Experimental Procedures

A stabilized form of the CD19 extracellular domain (ECD) was cloned in frame with an anti-CD20 scFv and an anti-albumin VHH, to create a monomeric CD19-ECD-anti-CD20 bridging protein with extended circulating half-life characteristics. The protein was purified from a mammalian cell expression system. Protein stability, binding affinities, and cytotoxic activity were analyzed in vitro. We utilized CD19-positive, CD20-positive and double positive cell lines to assess single and dual antigen activity. We utilized patient derived CD20-positive/CD19-negative cells to demonstrate translational relevance. Finally, we used single and dual flank in vivo models to assess the potency of the bridging protein in the relapse setting and in the prevention setting.

Results and Discussion

The CD19-anti-CD20 bridging protein was shown to be expressed at high levels, readily purified and highly stable (no aggregation or clipping, thermostable, and stable in media/serum at 37oC for extended periods). The purified bridging protein directed CAR19 cytotoxicity against CD19-negative/CD20-positive cells with superb potency (IC₅₀ = 23pM = 1.6 ng/ml). CAR19 T cells that were previously activated by a CD19-positive tumor cell could subsequently be activated by a CD19-negative tumor cell in the presence of the CD19-anti-CD20 bridging protein. In vitro, a CAR19 T cells found and eliminated CD19-negative cells "hidden" in a population of dual-positive cells in a mixing experiment but only if the bridging protein was present, otherwise, the CD19-negative cells invariably escaped from CAR19 T cells. The activity of the CD19-anti-CD20 bridging protein extended to CD19-negative/CD20-positive patient-derived cells tested in vitro. In vivo, using a dual flank model, CAR19 T cells plus the injected bridging protein controlled both CD19-positive/CD20-positive and CD19-negative/CD20-positive tumors, while CAR19 alone did not impact the latter tumor. In a relapse setting the growth of a mixture of CD19-positive and CD19-negative cells was merely delayed by CAR19 T cells alone but was eradicated when CAR19 cells were given along with the CD19-anti-CD20 bridging protein injected systemically. Importantly, CAR19 cells that had "lost" control over the mixed population could be restimulated to eliminate the CD19-negative population when the CD19-anti-CD20 bridging protein was added after those cells have begun to escape the initial (CAR19-only) treatment in vivo.

These results have led to the identification of a development candidate for the treatment of CD19-negative relapse from CAR19 treatment. The GMP production campaign is underway. The first-in-human trial will enroll patients relapsing from CAR19 therapy with CD19-negative malignancy, in whom CAR19 T cells are shown to still be present.