In vivo results of AGX101, a TM4SF1-directed tubulin inhibitor conjugate, in combination with immune checkpoint inhibitors

Introduction: TM4SF1 (Transmembrane-4 L-Six-Family-Member-1) is an endothelial marker with critical roles in angiogenesis, as well as a tumor cell antigen that contributes significantly to invasion and metastasis. TM4SF1 is upregulated 20-fold in angiogenic tumor vascular endothelium compared to normal vasculature endothelium and exhibits a unique nuclear internalization pathway. AGX101 is a novel tubulin inhibitor conjugate specifically directed against TM4SF1, delivering a potent maytansinoid payload directly to the nucleus of cells within the tumor microenvironment resulting in three mechanisms of action (MoAs): (1) activation of tumor immune surveillance, (2) tumor blood supply deprivation, and (3) direct tumor cell killing. Methods: TM4SF1 expression was assessed using immunohistochemistry. Safety and pharmacokinetics of AGX101 were evaluated in non-human primates (NHP), with escalating doses to determine the highest non-severely toxic dose (HNSTD). Efficacy studies were also conducted in mouse models, enabling assessment of the minimum effective dose (MED) needed to engage each of the three MoAs. The combination of HNSTD and MED enables calculation of a therapeutic index (TI). The potential for synergy with immune checkpoint inhibitors (ICIs) was also investigated. Results: Notable cancers with high TM4SF1 scoring intensity include lung cancer, kidney cancer, ovarian cancer, GI cancers, and breast cancer. In NHP, AGX101 exhibited a favorable safety profile.In preclinical efficacy studies, monotherapy demonstrated robust efficacy through each of the three MoAs in mouse syngeneic models including CT26 colon carcinoma and human tumor xenograft models including MIA PaCa-2 pancreatic cancer. Measured by exposure, TI was large. In syngeneic mouse models the effects of ICIs were potentiated by the AGX101 murine surrogate AGXB01, suggesting potential synergy. High response rates (RR) were achieved including CT26 with 71% RR for AGXB01+anti-CTLA-4 antibody vs 13% for anti-CTLA-4 alone; Renca mouse renal cancer model with 80% RR for AGXB01+anti-CTLA-4 antibody vs 20% for anti-CTLA-4 alone. In B16F10 mouse melanoma, a non-responsive model, AGXB01+ anti-PD-1 antibody or anti-CTLA-4 antibody ~doubled survival time vs either monotherapy of AGXB01 or anti-PD-1 antibody or anti-CTLA-4 antibody. In rechallenge experiments, 38 of 39 mice that had been rechallenged after tumor-free responses successfully eradicated new tumors without drug retreatment, demonstrating a durable immune response. Conclusion: AGX101 represents a promising new approach in cancer therapy. The preclinical data suggest that AGX101 could provide a significant therapeutic benefit by novel and differentiated mechanisms of action, namely selectively targeting the tumor vasculature and potentiating ICIs. Further clinical development of AGX101 is ongoing.