Glioblastoma, a WHO grade IV astrocytoma, constitutes approximately half of malignant tumors of the central nervous system. Despite technological advancements and aggressive multimodal treatment, prognosis remains dismal. The highly vascularized nature of glioblastoma enables the tumor cells to grow and invade the surrounding tissue, and vascular endothelial growth factor-A (VEGF-A) is a critical mediator of this process. Therefore, over the past decade, angiogenesis, and more specifically, the VEGF signaling pathway, has emerged as a therapeutic target for glioblastoma therapy. This led to the FDA approval of bevacizumab, a monoclonal antibody designed against VEGF-A, for treatment of recurrent glioblastoma. Despite the promising preclinical data and its theoretical effectiveness, bevacizumab has failed to improve patients’ overall survival. Furthermore, several other anti-angiogenic agents that target the VEGF signaling pathway have also not demonstrated survival improvement. This suggests the presence of other compensatory angiogenic signaling pathways that surpass the anti-angiogenic effects of these agents and facilitate vascularization despite ongoing VEGF signaling inhibition. Herein, we review the current state of anti-angiogenic agents, discuss potential mechanisms of anti-angiogenic resistance, and suggest potential avenues to increase the efficacy of this therapeutic approach.

Background: Granzyme B (GrB) is a key effector molecule, delivered by cytotoxic T lymphocytes and natural killer cells during immune surveillance to induce cell death. Fusion proteins and immunoconjugates represent an innovative therapeutic approach to specifically deliver a deadly payload to target cells. Epithelial membrane protein-2 (EMP2) is highly expressed in invasive breast cancer (BC), including triple-negative BC (TNBC), and represents an attractive therapeutic target. Methods: We designed a novel fusion protein (GrB-Fc-KS49) composed of an active GrB fused to an anti-EMP2 single-chain antibody tethered through the immunoglobulin G heavy chain (Fc) domain. We assessed the construct’s GrB enzymatic activity, anti-EMP2 binding affinity, and cytotoxicity against a panel of BC cells. The construct’s pharmacokinetics (PK), toxicity profile, and in vivo efficacy were also evaluated. Results: GrB-Fc-KS49 exhibited comparable GrB enzymatic activity to commercial GrB, as well as high affinity to an EMP2 peptide, with the dissociation constant in the picomolar range. The fusion protein rapidly internalized into EMP2+cancer cells and showed in vitro cytotoxicity to cell lines expressing surface EMP2, with half-maximal cytotoxicity (IC₅₀) values below 100 nM for most positive lines. Ex vivo stability at 37°C indicated a half-life exceeding 96 hours while in vivo PK indicated a biexponential plasma clearance, with a moderate initial clearance (t_1/2α=18.4 hours) and a much slower terminal clearance rate (t_1/2β=73.1 hours). No toxicity was measured in a Chem16 panel between the control and the GrB-Fc-KS49. In vivo, the GrB-Fc-KS49 showed efficacy against a TNBC syngeneic (4T1/_FLuc) mouse model, reducing tumor volume and cell proliferation and increasing cell death compared with controls. Treatment using an EMT6 mouse model confirmed these results. In addition to a significant impact on cell proliferation, GrB-Fc-KS49 treatment also resulted in a dramatic increase of tumor-infiltrating CD45+ cells and redistribution of tumor-associated macrophages. Transcriptomic analysis of tumors post-treatment confirmed the remodeling of the immune tumor microenvironment by the GrB-Fc-KS49 immunotoxin. Conclusions: GrB-Fc-KS49 showed high specificity and cytotoxicity towards EMP2-positive cells. In vivo, it reduced tumor burden and increased the recruitment of immune cells into the tumor, suggesting that GrB-Fc-KS49 is a promising therapeutic candidate against BC.