A Modular Ex Vivo Human Vessel Perfusion Platform for Studying Vascular Reconstruction

Background Vascular tissue responses to mechanical and biochemical stimuli are central to many vascular diseases and interventions, including maturation and failure of arteriovenous fistulas (AVF). Experimental systems that preserve native human vascular architecture while allowing controlled perfusion, surgical manipulation, and vascular interventions remain limited. We developed a modular ex vivo perfusion platform that supports pulsatile flow across multiple vascular configurations and enables controlled study of vessel architecture, hemodynamics, and cell biology ex vivo. Methods Deidentified human iliac arteries and veins discarded from liver transplantation were obtained under an institutional review board-approved protocol and mounted in a closed-loop circuit driven by a pulsatile blood pump. Straight arterial vessels and surgically created ex vivo arteriovenous fistulas were perfused with endothelial cell culture medium supplemented with 1x penicillin-streptomycin, maintained at 37°C. Flow dynamics were assessed using Doppler ultrasound. Representative vessel segments were fixed for H&E, trichrome, TUNEL staining, and cleaved caspase-3 immunohistochemistry. TUNEL-positive area was quantified across whole vessel sections using FIJI/ImageJ with Otsu automated thresholding, with manual counts of TUNEL- and cleaved caspase-3-positive nuclei performed as confirmatory analyses. Results The platform accommodated straight arterial and surgically constructed arteriovenous fistula configurations during 2-hour pump perfusion. Pump settings of 40 strokes/min and 7 mL stroke volume generated pulsatile perfusion, confirmed by pressure transducers. In straight arterial runs, pulsed-wave Doppler confirmed measurable flow through the isolated human iliac artery, with velocities ranging from 4.81 to 25.18 cm/s and a mean velocity of 18.9 cm/s. In the AVF configuration, Doppler ultrasound demonstrated disturbed flow at the anastomosis, and venous outflow volume-flow rates ranged from 752.1 to 1442.5 mL/min, exceeding the 600 mL/min threshold commonly used to define functional fistula flow. Histology demonstrated close arteriovenous apposition in the constructed fistula. Pre- and post-perfusion TUNEL quantification showed no increase in TUNEL-positive area after 2-hour perfusion, whereas in a separate time-course analysis, TUNEL-positive area increased with longer intervals from vessel acquisition to fixation. Conclusions We describe a modular human vessel perfusion platform that supports pulsatile flow and clinically relevant arteriovenous fistula conditions ex vivo . Histologic assessment demonstrated preserved vessel architecture and no increase in apoptotic burden after short-term pump perfusion. This system provides a biologically viable human tissue platform for studying mechanical, biochemical, and device-related contributors to vascular reconstruction.