Introduction: Transplantation of stem cell-derived pancreatic islets (SC-islets) could drastically improve disease control for subjects with type 1 diabetes. Engraftment of SC-islets is site dependent and has many challenges. No comparison of multiple implantation sites has been conducted so far. Therefore, we evaluated transcriptional maturity, C-peptide, glucose control, hypoxia and cell composition of SC-islets implanted to mice in the pancreas, liver, kidney capsule, omentum, and abdominal muscle.
Methods: In immunodeficient NOG-mice, SC-islets were implanted into the above organs. IVGTT and sampling of stimulated human C-peptide were carried out at one and three months. Then, implants were harvested for immunohistochemistry and RNA sequencing, which was compared to human islets for islet maturity. Furthermore, grafts were stained for the hypoxia marker pimonidazole, and insulin and glucagon as a fraction of IA2 for cell composition and quantified in FIJI (ImageJ 1.53q).
Results: Stimulated human C-peptide levels increased two to seven-fold from month one to three for all organ sites. At one month, stimulated C-peptide was higher in liver transplants than pancreas and omentum (p=0.0165 and 0.0004, respectively; ANOVA p=0.0008). No difference was seen between other sites. At three months, no difference in C-peptide was seen between groups, but at both time points, liver had the highest individual values.

RNA-sequencing data is being analyzed and will be ready before the conference.
IVGTT did not differ between sites at one or three months, except for lower basal glucose in liver compared to pancreas and omentum implants at three months. The liver group had the lowest AUC at three months, indicating the best clearance, despite having the highest AUC at 1 month.
Graft composition analysis showed liver implants having a higher β-cell fraction than all other sites except kidney, with no difference in α-cell fractions. In liver, the β-cell fraction was higher than the α-cell fraction (p= 0.0061), while in other sites, fractions did not differ.
No graft site was positive for pimonidazole, indicating no hypoxia was present after three months.
Discussion: In this, to our knowledge, first comparison of multiple sites for implantation of SC-islets, new findings emerged.
Firstly, SC-islets in all tested organ sites produced C-peptide already at one month, increasing two to seven-fold at three months. Seemingly, liver SC-islet grafts produce the most C-peptide at stimulation at one month, while other sites improved at three months. However, the highest individual C-peptide levels at both one and three months were from SC-islets implanted to liver.
Liver grafts also stood out in cell composition, with a higher β-cell fraction than other sites; interestingly not at the cost of α-cell area, which did not differ between sites.
In summary, all five sites seem functionally fitted for implantation of SC-islets, but with liver as the most advantageous site.