First results of a regenerative transcatheter heart valve implant from autologous tissue in a long-term animal model

Abstract Background The perfect heart valve replacement remains to be found regarding longevity, freedom of anticoagulation and availability in each size. Purpose Examination of a new technique for a regenerative, immunocompatible, transcatheter pulmonary valve implant from autologous pericardium was performed in an adult sheep model. Methods For each valve implant a pericardial patch was harvested by left antero-lateral mini-thoracotomy, placed on one of two slightly different shaped moulds either made of acrylic glass (AG) or printed in a 3D Printer (3D) and transferred into a container filled with a biological crosslinking agent at a non-toxic concentration to stabilize the given shape. After a median of 3 days (d) they were unpacked, sewn into a self-expandable nitinol stent with a diameter of 26 or 30mm and implanted via the jugular vein into pulmonary valve position of the same animal using a custom-made delivery system (1). Three groups (Gr) were planned: Gr 1) 4 animals with at least 6 d of crosslinking and AG mould; Gr 2) 5 animals with at least 3 d of crosslinking and AG mould; Gr 3) 6 animals with at least 3 d crosslinking and 3D mould. Follow ups (FU) were performed every 3 months (mo) evaluating valve function by intracardiac echocardiography (ICE) and cardiac MRI for up to 20.5 mo. All experiments were approved and conducted according to German federal law. Results In 11 of 13 animals minimally-invasive implantation was successful. One animal died because of ventricular fibrillation, in the other the implant dislocated into the right ventricle, so it had to be sacrificed. Direct post-implantation valve insufficiency was evaluated in 9 animals using ICE or angiography. In Gr 1 insufficiency was non-existent (n=2), in Gr 2 moderate or severe (n=2) and in Gr 3 non-existent or mild (n=5). Long term function as shown in figure 1 decreased rapidly in Gr 1 and 2 with median MRI regurgitation fractions (RF) in Gr 1 of 31% at 3 mo (n=3), 41% at 6 mo (n=3), 48% at 9 mo (n=2), 47% at 12 mo (n=3) and in Gr 2 of 43% at 3 and 54% at 6 mo (n=1) after implantation. Median RF in Gr 3 was small with 9% at 3 mo (n=4), 8% at 6 mo (n=3), 8% at 9 mo (n=3), 12% at 13 mo (n=3), 8% at 17 mo (n=2) and 20.5 mo (n=2). We never witnessed valve stenosis in any group at any point in time. Conclusion Gr 3 showed promising results regarding long time function of the implant which encourages further research with higher validity. The reason for failing of Gr 1 and 2 remains a