A biological alternative to alloplastic grafts in dialysis therapy: evaluation of an autologised bioartificial haemodialysis shunt vessel in a sheep model.

OBJECTIVES

To evaluate bioartificial haemodialysis access grafts in a sheep model with respect to patency and morphology.

MATERIAL AND METHODS

Bovine internal thoracic arteries (n=28) were decellularised. Fourteen grafts (DC grafts) were directly implanted as cervical AV shunts, the remaining were re-seeded with endothelial cells (ECs) derived from blood samples of the later ovine recipient (EC grafts) first. Following simulated punctures and duplex ultrasound scans to determine patency, grafts were explanted for immunohistochemical characterisation after 3 and 6 months, respectively. DC grafts underwent biomechanical testing for compliance (C), suture retention strength (SRT), and burst pressure (BP) before (n=6) and after (n=6) implantation.

RESULTS

Following 3 and 6 months, the majority of EC (n=6/6; n=6/7) and DC grafts (n=5/6; n=5/7) were patent and not relevantly stenosed (peak systolic velocity: EC grafts=76 cm s(-1)±4; DC grafts=77 cm s(-1)±5). Simulated haemodialysis punctures revealed significantly shorter bleeding times in all bioartificial grafts than in native jugular veins (P>0.001). Comparing native carotid arteries with DC grafts prior to and post-implantation, the latter differed significantly with respect to C (P>0.001; P=0.005), whereas only pre-implant DC grafts differed regarding BP (P=0.002); no differences were observed for SRT. Histology revealed complete endothelial surface coverage of EC, but not DC grafts. Furthermore, DC grafts exhibited areas of pronounced tissue calcification.

CONCLUSION

The preclinical development of a bioartificial haemodialysis access graft with promising mechanical and morphological properties in a sheep model is feasible.