Paradoxical Angiogenic Activation Under Anti-Vegf Therapy: Tip Cell Hyper-Sprouting and Vessel Remodeling Drive Inefficient Vascularization in Multiscale Tumor Simulations Publisher Pubmed



Manshadi M D ; Soltani M
Source: BMC Cancer Published:2026

Abstract

Background: Anti-angiogenic therapy targeting VEGF is designed to suppress tumor vascularization. Paradoxically, clinical and preclinical studies report transient increases in vessel density or abnormal vascular patterning under treatment, phenomena poorly understood mechanistically. Methods: Using a multiscale 3D agent-based model of solid tumor growth, we simulated the effects of three anti-VEGF agents — Bevacizumab, Ranibizumab, and Brolucizumab — initiated at iteration 1320 (day 40) over a 60-day period (1980 iterations). We quantified vascular cell density, tip cell dynamics, interstitial pressure, and Tumor Angiogenesis Factor (TAF) to dissect vascular remodeling patterns. Results: All three drugs increased total vessel cell count compared to control (Ranibizumab: +67%, Bevacizumab: +11%, Brolucizumab: +25%). Strikingly, tip cell numbers surged under therapy — Ranibizumab induced a 94% increase over control, Brolucizumab + 50%, Bevacizumab + 33%. Despite increased vessel and tip counts, tumors under Ranibizumab and Brolucizumab showed suppressed growth, indicating non-functional or chaotic angiogenesis. Bevacizumab was associated with an increased peak TAF relative to control (1.65 vs. 1.58), and high interstitial pressure (0.75) suggest compensatory signaling and vascular leakage while endpoint TAF values were comparable across conditions. Conclusion: Anti-VEGF therapy does not simply prune vessels — it triggers dysregulated sprouting and architectural instability. The disconnect between vessel quantity and functionality reveals a state of “angiogenic inefficiency,” where increased vascular metrics mask underlying dysfunction. These findings redefine vascular response to therapy and highlight the need to evaluate vessel quality — not just density — in treatment assessment. © The Author(s) 2026.