632 HPN601 is a protease-activated EpCAM-targeting T cell engager with an improved safety profile for the treatment of solid tumors. (9th November 2020)

• Record Type: Journal Article
• Authors: Lin, Jack; Rocha, Sony; Kwant, Kathryn; Dayao, Maria; Ng, Tessie; Aaron, Wade; Callihan, Evan; Gamez-Guerrero, Maria; Hemmati, Golzar; Wright, Kevin; Barath, Manasi; Xiao, Yinghua; Degree, Master; Yu, Timothy; Chew, Patrick; Evans, Thomas; O'Rear, Jessica; Gatto, Scott; Cremin, Michael; Yu, Stephen; Patnaik, Purbasa; Hundal, Avneel; Austin, Richard; Lemon, Bryan; Wesche, Holger
• Is Part Of: Journal for immunotherapy of cancer. Volume 8(2020)Supplement 3
• Page Start: A379
• Page End: A379
• Publication Date: 2020-11-09
• Subjects: Cancer -- Immunotherapy -- Periodicals; Cancer -- Immunological aspects -- Periodicals; Tumors -- Immunological aspects -- Periodicals; Immunotherapy -- Periodicals; 616.99406105
• DOI: 10.1136/jitc-2020-SITC2020.0632
• Languages: English
• ISSNs: 2051-1426

Abstract

Abstract : Background: Epithelial cell adhesion molecule (EpCAM) is highly expressed in many solid tumors. However, therapeutics targeting EpCAM have had limited clinical utility or failed in clinical development likely due to the expression of EpCAM in normal tissues. For example, clinical testing of solitomab, an EpCAM-targeting T cell engager, resulted in severe dose-limiting toxicities, including elevated liver transaminases, hyperbilirubinemia, and diarrhea. Designing an EpCAM-targeting T cell engager that is only active in the tumor would expand its therapeutic window and improve its safety profile. Methods: Using a T cell engager prodrug platform named ProTriTAC that couples therapeutic half-life extension with functional masking, we have engineered HPN601, a protease-activated EpCAM-targeting T cell engager. HPN601 is a single polypeptide with three binding domains: anti-albumin for half-life extension, anti-CD3e for T cell engagement, and anti-EpCAM for tumor cell engagement. The anti-albumin domain contains a masking moiety and a protease-cleavable linker and keeps the molecule inert outside the tumor microenvironment. Activation by tumor-associated proteases removes the anti-albumin domain along with the masking moiety to reveal a potently active drug inside the tumor. This active drug has minimal activity outside of tumor because, without an albumin binding domain, it is rapidly cleared in circulation. Results: A humanized rodent tumor model was used to simultaneously measure anti-tumor efficacy and clinically relevant toxicity endpoints. In this model, a surrogate molecule of HPN601 was safely administered at a dose ten-fold higher than the minimal efficacious dose required for durable tumor regression. Higher doses produced toxicities including elevated ALT/AST and bilirubin, body weight loss, and evidence of tissue damage by histopathology. In contrast, a constitutively active EpCAM-targeting T cell engager could only be dosed safely up to its minimal efficacious dose. The improved safety profile of HPN601 is further supported by a toxicokinetic study in non-human primates: compared to a constitutively active EpCAM-targeting T cell engager, HPN601 had significantly attenuated cytokine production, including IFN-g, IL-2, IL-6, and IL-10. Conclusions: HPN601 is a conditionally active EpCAM-targeting T cell engager with a ten-fold improved therapeutic window compared to a constitutively active EpCAM-targeting T cell engager. An EpCAM-specific T cell engager with an improved safety profile could address unmet needs in many solid tumors and demonstrate the feasibility of using conditionally active T cell engagers to target more solid tumor antigens. Ethics Approval: The study was reviewed and approved by Harpoon's Institutional Animal Care and Use Committee.