Toward a 3D Printed Perfusable Islet Embedding Structure: Technical Notes and Preliminary Results.

Abstract

To date, islet transplantation to treat type 1 diabetes mellitus remains unsuccessful in long-term follow-up, mainly due to failed engraftment and reconstruction of the islet niche. Alternative approaches, such as islet embedding structures (IESs) based on 3D printing have been developed. However, most of them have been implanted subcutaneously and only a few are intended for direct integration into the vascular system through anastomosis. In this study, we 3D printed a proof-of-concept IES using gelatin methacrylate biocompatible ink. This structure consisted of a branched vascular system surrounding both sides of a central cavity dedicated to islets of Langerhans. Furthermore, we designed a bioreactor optimized for these biological structures. This bioreactor allows seeding and perfusion experiments under sterile and physiological conditions. Preliminary experiments aimed to analyze if the vascular channel could successfully be seeded with mature endothelial cells and the central cavity with rat islets. Subsequently, the structures were used for a humanized model seeding human endothelial progenitor cells (huEPC) within the vascular architecture and human islets co-cultured with huEPC within the central cavity. The constructs were tested for hemocompatibility, suture strength, and anastomosability. The 3D printed IES appeared to be hemocompatible and anastomosable using an alternative cuff anastomosis in a simple ex vivo perfusion model. While rat islets alone could not successfully be embedded within the 3D printed structure for 3 days, human islets co-cultivated with huEPC successfully engrafted within the same time. This result emphasizes the importance of co-culture, nursing cells, and islet niche. In conclusion, we constructed a proof-of-concept 3D printed islet embedding device consisting of a vascular channel that is hemocompatible and perspectively anastomosable to clinical scale blood vessels. However, there are numerous limitations in this model that need to be overcome to transfer this technology to the bedside.