In vivo validation of multimodality pore-network modeling to identify angio-permissive scaffold porosity.

BACKGROUND: Successful vascular tissue regeneration in vascular and soft-tissue biomaterials is governed not by bulk volumetric porosity, but by the existence of continuous, ingrowth-permissive pathways traversing the full scaffold thickness, termed angio-permissivity. Conventional structural metrics often fail to capture these functional conduits, leading to unpredictable in vivo outcomes and a disconnect between scaffold design and biological integration. METHODS: We developed a transmural space characterization workflow integrating micro-computed tomography, deep-learning-assisted super-resolution reconstruction and segmentation, and pore-network modeling. Three architecturally distinct electrospun scaffold groups were thresholded for continuous pathways (>10 μm) and analyzed for vascular ingrowth permissivity. Findings were validated in a subcutaneous rat model (7 and 21 days) to correlate architectural parameters with extracellular matrix remodeling and neovascularization. RESULTS: Quantitative modeling identified a "porosity paradox," where architectures with the highest total volumetric void space remained functionally isolated due to internal partitioning and sub-critical bottlenecks (<10 μm). Only scaffolds exhibiting a dense, continuous network of surface-to-surface growth tunnels supported robust transmural integration. In vivo, these continuous spatial configurations facilitated deep neovascularization and a rapid maturation "catch-up" to native glycosaminoglycan levels by day 21. In contrast, partitioned architectures restricted vessel recruitment and demonstrated impaired extracellular matrix preservation, regardless of high initial porosity. CONCLUSIONS: Transmural connectivity is a primary architectural determinant of tissue and vascular integration. This non-destructive, scalable framework enables the engineering-led design of tissue-engineered biomaterials by prioritizing the specific spatial and architectural requirements of the target physiological niche over stochastic volumetric metrics. Validation here uses a subcutaneous model that isolates architecture from cardiovascular hemodynamics - the next requirement for translation.