Redefining Tumor Vascular Permeability through Deep Learning-Guided Microneedle Delivery.
Journal:
ACS nano
Published Date:
Feb 9, 2026
Abstract
Low-permeability (LP) tumor vasculature constitutes a major barrier to efficient nanomedicine delivery, making quantitative assessment and mechanistic understanding of vascular permeability essential for the rational design of delivery strategies. Here, we introduce a deep learning-guided microneedle (MN) delivery platform that enables localized and spatiotemporally precise modulation of tumor vasculature to enhance nanoparticle extravasation. By integrating the MN system with an upgraded single-vessel analysis framework (nano-ISML 1.1), we quantitatively mapped vascular remodeling and nanoparticle transport across diverse tumor types and particle sizes. Localized histamine delivery via MNs selectively expanded endothelial junctions through VE-cadherin-mediated regulation, significantly increasing the frequency and length of interendothelial gaps, and thereby reprogramming LP tumors toward a high-permeability phenotype. This controlled vascular remodeling established a pronounced size-dependent permeability window, defined by locally induced gap dimensions that varied across tumor types, permitting efficient penetration of nanoparticles ≤200 nm while largely excluding particles >500 nm. By uniting nanotechnology, vascular biology, and artificial intelligence, this interdisciplinary framework provides a mechanistic and predictive paradigm for overcoming vascular barriers and advancing the rational design of tumor-targeted nanomedicines.
Authors
Keywords
No keywords available for this article.