Tumor angiogenesis, the process by which tumors form new blood vessels, is a critical driver of cancer growth. This process is primarily fueled by vascular endothelial growth factor (VEGF) and its receptors (VEGFRs). While targeting tumor angiogenesis has shown promise in slowing cancer progression, no selective small-molecule inhibitors specifically targeting this pathway are currently available. At Bioptic, we are excited to share our recent advancements in developing a selective VEGFR-3 inhibitor with potential anti-tumor activity across multiple cancer types.
Innovative Approach to Drug Discovery
Our team at Bioptic employed a cutting-edge virtual screening pipeline to identify promising compounds. This pipeline integrates two advanced models:
- SMILES-Based Large Language Model (LLM): Fine-tuned on binding affinity data from contrastive molecular pairs, this model screened an ultra-large virtual library of 40 billion compounds from Enamine REAL.
- Graph Neural Network (GNN): Designed for selectivity, this secondary model re-scored top-ranked molecules to ensure specificity by distinguishing activity across similar kinases.
This dual-model approach allowed us to efficiently identify compounds with high potency and selectivity for VEGFR-3, a key receptor in tumor angiogenesis.
Identifying High-Potential Cancer Types
To determine which cancers would respond best to a VEGFR-3 inhibitor, we utilized the Oncobox algorithm. This powerful tool analyzed RNA-seq profiles from:
- TCGA Database: 11,428 profiles across 33 primary cancer sites.
- Internal Real-World Data (RWD) Cohort: 1,056 profiles from 89 cancer types.
The Oncobox algorithm evaluated VEGF(R) expression and tumor angiogenesis pathway activation to predict the efficacy of VEGFR-3 inhibitors. Our analysis identified papillary thyroid cancer as the most responsive, followed by clear-cell renal cancer, pancreatic cancer, ovarian cancer, and sarcomas. These findings guide our focus for further preclinical validation.
Promising Results from Compound Testing
From our screening efforts, we tested 110 compounds using the Eurofins VEGFRs KinaseProfiler, considering those with an IC50 < 10 μM as active. The results were encouraging:
- 1 compound was active against VEGFR-1.
- 1 compound was active against VEGFR-2.
- 4 compounds were active against VEGFR-3.
- 1 compound showed activity across all VEGFRs.
- 3 compounds were selective for a single VEGFR.
Notably, one VEGFR-3 inhibitor demonstrated >45-fold selectivity over VEGFR-1 and VEGFR-2, with minimal activity against 12 off-target kinases, including B-Raf, c-Raf, c-Kit, FGFR1-4, Flt3, Met, PDGFR-a, PDGFR-b, and Ret. This high selectivity suggests a lower risk of off-target effects, potentially reducing toxicity compared to existing pan-VEGFR inhibitors.
Implications for Cancer Treatment
The development of a selective VEGFR-3 inhibitor could transform cancer therapy by offering a more targeted approach. Unlike pan-VEGFR inhibitors, which can cause significant side effects due to their broad activity, a selective inhibitor may improve patient outcomes by minimizing toxicity. Additionally, our RNA-seq-based analysis paves the way for patient enrichment in clinical trials, enabling the identification of potential responders and the development of a companion diagnostic to guide treatment decisions.
Looking Ahead
Our findings highlight the power of Bioptic’s molecular search engine in identifying potent and selective inhibitors for specific targets. When combined with the Oncobox algorithm, this approach streamlines the development of novel anti-cancer drugs. We are now advancing our selective VEGFR-3 inhibitor toward preclinical validation, with the goal of bringing a safer, more effective therapy to patients with cancers driven by tumor angiogenesis.
This research was conducted by Maxim Sorokin, Simon Steshin, Vladislav Vinogradov, Kong Nguyen, Ivan Izmailov, Andrew Garazha, and Andrey Doronichev, affiliated with OmicsWay Corp. (San Mateo, CA; Covina, CA), Optic, Inc. (San Francisco, CA), and ORIC (South San Francisco, CA). No funding sources were reported.
©2025 American Society of Clinical Oncology, all rights reserved. Licensing available upon request. For more information, please contact licensing@asco.org.
