Characterizing transmural ingrowth spaces: a key step towards clinical in-situ endothelialization

Abstract

There is growing recognition in regenerative cardiovascular tissue engineering that transmural vessel ingrowth is the predominant—if not exclusive—mechanism for achieving in-situ endothelialization in prosthetic vascular grafts and heart valves in humans This process requires continuous ingrowth channels with dimensions sufficient to accommodate capillaries or even arterioles. While a variety of methods—such as electrospinning—exist to create porous scaffolds, current characterization techniques fail to determine whether the resulting structures offer well-defined and consistently continuous ingrowth spaces. Drawing on principles from geological porous media research, we applied a combination of nano-computer-tomography, deep-learning segmentation and super-resolution algorithms, and pore network modelling, to characterize the full thickness pore space morphology of electrospun scaffolds. Scaffolds were non-destructively reconstructed at high resolution (0.54 microns) and large fields of view, 57x faster than a brute-force approach, achieving total sample volumes greater than 1x10⁸ um³ in just a few hours. Electrospun scaffolds showed a median pore size and median pore volume of 5.51um (IQR: 5.15)/418.07um² (IQR: 1153.74), n = 15 698, for the 16% polymer weight percentage group; 5.40um (IQR: 6.23)/412.24um² (IQR: 1485.24), n = 13 437, for 18%; and 5.40um (IQR: 4.22)/356.34um² (IQR: 826.53), n = 28 620, for 20%. On deeper analysis, continuous, interconnected pore networks of <10 microns in minimum diameter were extracted, with the 18% group showcasing the most extensive, surface-to-surface networking. This analysis highlights the limited ability of single-needle electrospinning to produce sufficient growth space for reliable transmural capillary endothelialisation. With the advent of cutting-edge additive and reductive manufacturing techniques, alternative methods for porous scaffold construction show promise.