IntroductionRepeated access to the circulation is essential for providing successful hemodialysis. Currently, the preferred vascular access to facilitate this is an autologous arteriovenous (AV) fistula. However, their use is limited by pre-existing conditions, prior surgeries, unsuitable autologous vessels, or maturation failure1. Although currently available synthetic alternatives like expanded polytetrafluoroethylene (ePTFE) can be used immediately without the need for maturation, they lack the healing properties of autologous tissue and are susceptible to infection, intimal hyperplasia, and thrombosis2,3.Recurrent complications in synthetic vascular access grafts are related to the repeated insertion of large-bore needles required for hemodialysis4,5. The disruption of the graft wall caused by repetitive punctures results in loss of graft integrity6; hematomas, (pseudo)aneurysm formation, and eventual graft failure7. The non-healing nature of ePTFE grafts further exacerbates these problems by providing niches for bacterial accumulation and impeding the host’s immune response to infections8,9. Furthermore, the mismatch in compliance between the rigid synthetic grafts and the outflow veins frequently leads to hemodynamic stress and the development of venous neointimal hyperplasia10,11.Vascular tissue engineering (TE) may offer many benefits over traditional synthetic grafts, as TE vascular access grafts, most importantly, are able to heal. Furthermore, they can adapt to local conditions and flow patterns, resembling natural vessels, and allow the immune system to manage infections effectively. Our proposed solution aims to bridge the gap between autologous AV fistulas and synthetic grafts by leveraging in situ tissue engineering to create a functional, living vascular access. We recently reported that a synthetic, biodegradable supramolecular electrospun scaffold fully transforms into a living, compliant vascular graft within 12 weeks after implantation in a large animal AV-shunt model12. The grafts could be successfully cannulated after termination. These grafts consisted of Polycarbonate-Bis Urea (PC-BU), a polymer with supramolecular properties that was also previously successfully applied in vitro13,14 and in vivo12,15,16. Due to its supramolecular nature, this material enables a modular design, allowing for the incorporation of BU-functionalized additives to fine-tune mechanical properties14 and facilitate functionalization with bioactive components17, anti-microbial18, and/or anti-fouling additives19. These highly adaptable characteristics could help address key challenges associated with fully synthetic grafts, such as infection, intimal hyperplasia, remodeling, and thrombosis. However, we observed no improved patency of these grafts compared to the gold standard ePTFE grafts12. This may have been related to the presence of a 3D-printed anti-kinking coil made of polycaprolactone (PCL), which tends to fragment under continuous (hemodynamic) stress20. The resulting coil fragments protruding inside the lumen may increase the risk of thrombosis.In this proof-of-concept study, we pursued a modular approach, combining the previously developed PC-BU electrospun grafts with a newly designed 3D-printed coil with supramolecular properties. This was achieved by blending PCL with bis urea-modified PCL to mitigate fracturing induced by the high crystallinity of PCL and enhance patency.We then examined the in vivo performance of this adapted TE vascular access graft in goats while concurrently assessing the impacts of cannulation on the various stages of remodeling, starting at 2 weeks after implantation. For this purpose, we evaluated scaffold resorption, mechanical behavior, vascular neo-tissue formation, healing capacity, and graft patency over a 12-week follow-up period.ResultsSynthetic scaffold designThe luminal layer was made of electrospun PC-BU (PC-BU: Mn = 21 kg/mol) with a thickness of 450 ± 50 µm. Directly onto this luminal electrospun layer, an anti-kinking spiral made of PCL/PCL-BU 60:40 w/w (PCL: Mn = 45 kg/mol, PCL-BU: Mn = 2.7 kg/mol) (Fig. 1A) was 3D printed, with a thickness of 0.52 ± 0.06 mm and an interspacing of 2.73 ± 0.26 mm (Fig. 1B). On top of that, the outer layer, with a thickness of 104 ± 20 µm, made of electrospun PC-BU was applied, sandwiching the 3D-printed spiral between the luminal and outer electrospun layer. The spiral design displayed local kinking resistance and dimensional stability in the lumen; no folds were observed. No morphological changes were observed after 3D printing on top of the electrospun layer. The fiber diameter at the luminal side of the bare PC-BU grafts was 3.6 ± 0.4 µm. The fiber diameter of the outer layer made of PC-BU grafts was 3.6 ± 0.5 µm. The diameter of the lumen of the grafts was 5.9 ± 0.2 mm.Fig. 1: Graft composition.Chemical structures of PCL-BU, PCL, and PC-BU (A). Image of the grafts, including coils as well as a schematic representation (B). Stress-strain curves of individual polymers and polymer mixtures (C).Full size imageStress-strain curves for 3D-printed coils (Fig. 1C) of different polymers and their mixtures revealed that the modulus of pure PCL (107.4 ± 0.8 MPa) was two orders of magnitude higher than that of the 80:20 w/w PCL-BU:PCL mixture (8.1 ± 0.06 MPa), pure PCL-BU (7.9 ± 0.09 MPa), and PC-BU (5.5 ± 0.03 MPa), all of which did not provide sufficient resistance to kinking. Increasing the PCL content to 40% led to an increase in the modulus to 42.1 ± 0.4 MPa. The thermal properties of the materials revealed that in the second heating run for the mixtures, only a melting transition was observed between 50 and 55 °C, which is similar to pure PCL. The enthalpy of this transition increases with higher PCL content, with 20% having 5.2 J/g, 40% 23.7 J/, and pure PCL 67.2 J/g. Pure PCL-BU has a melting transition at 117.5 °C, corresponding to the melting of the bis urea hard blocks. In the first heating run, the 60:40 w/w PCL-BU:PCL showed a second melting transition at 116.1 °C corresponding to the bis urea stacks. Upon melting, these BU do not form crystals again, which might be caused by the PCL hindering their interaction by interacting with the PCL chains of the PCL-BU co-polymer. The glass transition was around −60 °C for all the polymers and mixtures containing PCL. These results indicate that the PCL chains crystallize and dominate the interactions. The amount of crystallization increases with higher PCL content.Survival and graft patencyGrafts were explanted in 6 animals at 12 weeks (PC-BU n = 8; ePTFE n = 2) and 2 animals at 4 weeks (PC-BU n = 3 [cannulated n = 2, non-cannulated n = 1]; ePTFE n = 1) due to early termination because of bilateral stenosis in the jugular outflow veins, causing brain edema. No deformations (dilation, aneurysms, and/or ruptures) were observed at explanation, and PC-BU explants showed remodeling from synthetic to neo-tissue (Fig. 2A). No complications were found in sham-operated vessels. Coil fragmentation was not observed at any point during the study. Angiograms at 4 weeks showed that 62.5% of TE grafts and 66.7% of ePTFE grafts had a patent lumen (Fig. 9B). However, at this time point, two animals were prematurely sacrificed due to complications, resulting in the explantation of one ePTFE graft, two cannulated TE grafts, and one non-cannulated TE graft. At 8 weeks, 66.7% of ePTFE grafts and 56% of TE grafts remained patent. By 12 weeks, 45% of TE grafts and 66.7% of ePTFE grafts were still patent (Fig. 2B). Occlusions occurred in all groups, indicating no specific impact of cannulation on graft patency. Specifically, 2 out of 8 involved occlusion of the jugular vein itself, while 6 out of 8 cases showed occlusion of the graft at the anastomosis site. No stenosis was observed near the arterial anastomoses. Wall thickness was observed to increase over time (Fig. 2C), from the initial 0.6 mm at implantation to 0.76 ± 0.06 mm on average after 4 weeks increased to 2.64 ± 0.18 mm after 12 weeks. The wall thickness was also thicker when compared to the native CCA (0.78 ± 0.04 mm, p 0.01) higher stress of 74.64 ± 0.63 MPa, compared to the TE grafts with an average of 16.5 ± 2.54 MPa and 6.9 ± 3.25 MPa for the 4- and 12-week explants, respectively. The native carotid was measured at 50% strain to have a stress of 0.48 ± 0.15 MPa, significantly different (p