Immunotherapy of Acute Myeloid Leukemia Enabled By Targeting of Pan-Cancer Molecular Patterns with a Novel Trispecific T Cell Engager

Acute myeloid leukemia (AML) is a blood cancer which remains difficult to treat, especially in the setting of relapse or refractory status. Despite the success of immunotherapies such as CAR T and T cell engager therapies for the treatment of B cell malignancies, the same approaches have failed to shift the treatment paradigm for myeloid malignancies. One of the limitations to using antigen specific immunotherapies for AML is the concern that due to the shared surface antigen profile of AML cells and hematopoietic stem cells, targeting AML cells can lead to myeloablation. As such, many of the protein targets that have been the focus of AML immunotherapy research for years pose significant risks for on-target off-tumor toxicity. Identifying cancer-specific targets would be ideal to develop safe immunotherapies for AML, and non-protein targets have the advantage of not being subject to some of the common mechanisms of drug resistance development such as genetic mutation or downregulation.

Our study introduces high mannose (Man9) oligosaccharides (Man9 glycans are absent from healthy cells but they arise in malignant cells from an error in normal glycan assembly) and phosphatidylserine (PS) (PS molecules are located in the inner leaflet of the phospholipid bilayer cell membrane, but PS flips to the outer leaflet during apoptosis or malignant transformation) as promising targets for immunotherapy of AML. Our biotech partner Vitruviae developed fully synthetic receptor decoy molecules utilizing the lectin receptor DC-SIGN for Man9 binding and T cell immunoglobulin and mucin domain 1 (TIM-1) for PS binding.

In flow cytometric assays utilizing these DC SIGN and TIM-1 derived homodimeric as well as heterodimeric molecules, we demonstrated highly specific binding to AML cells (nine AML cell lines and eight samples of newly diagnosed AML patients) while sparing healthy cells such as peripheral blood and bone marrow derived mononuclear cells. Glycan microarray confirmed that the therapeutic molecules were able to bind to several abnormal glycans but to none of the glycans found in healthy cells. We next created a molecule with high affinity to Man9, PS and human CD3 thus functioning as a trispecific T cell engager directed against Man9 and/or PS positive cancers while lacking affinity to healthy cells. Initial toxicology studies in immunocompetent mice (including human CD3 transgenic mice) showed that high doses of the molecule (up to 4.8 mg/kg) were well tolerated. We evaluated the binding of our trispecific molecule to patient matched normal and cancerous tissue by FFPE immunohistochemistry and observed excellent tumor specificity, which predicts a favorable human safety profile. In vitro efficacy studies (coculture of luciferase transduced target cells with activated CD8+ T cells in the presence and absence of the T cell engager) demonstrated excellent antileukemia activity of the molecule against THP-1 and SET-2 AML cell lines with an IC50 in the 5-10 pM range. Additional structure activity relationship testing revealed that the potency of the molecule could be further improved when employing a more proximal anti-CD3 domain (to facilitate synapse formation between T cells and tumor cells) and thus, this optimized structure (VTRU200) was selected as the lead molecule for further preclinical and clinical development. We recently conducted a preliminary in vivo efficacy study to evaluate C1498 AML progression in immunocompetent human CD3 transgenic mice treated with three intravenous doses of VTRU200 and observed an impressive response (almost 100-fold decreased tumor burden) based on in vivo bioluminescence imaging. Additional preclinical efficacy and dose finding studies are currently underway, and translational studies in a patient-derived xenograft model of AML as well as investigational new drug (IND) enabling studies are in preparation.

In conclusion, we validated two highly promising non-protein targets for AML immunotherapy and developed innovative T cell engager-based Man9/PS dual targeting strategies characterized by the absence of on-target off-tumor toxicity and antigen negative relapse potential, which are major safety and efficacy limitations of currently available immunotherapies. Our goal is to develop a first-in-class, breakthrough Man9 x PS x CD3 trispecific T cell engager for AML patients with an anticipated timeline to IND of 18 months.