Introduction: Transplantation of pancreatic islets into the liver is a curative therapy for patients with poorly controlled type 1 diabetes. So far, there is no common non-invasive method for longitudinal monitoring of the transplanted viable islet cell mass. Exendin-4 as a specific ligand of the glucagon-like peptide 1 receptor (GLP1R), which is localized at high levels on beta cells, might be useful as radionuclide tracer to monitor the viable beta-cell graft mass in the liver by PET/CT in vivo.
Methods: We first established imaging of the pancreatic beta-cell mass of pigs by [⁶⁸Ga]Ga-DOTA-Exendin-4 PET/CT, including a dynamic PET scan (60 min) with focus on the pancreas. INS^C94Y transgenic diabetic pigs (1) (n = 4; age = 3-4.5 months) received 3,300 - 9,700 IEQ wild-type neonatal pig islets (NPIs) per kg body weight by minimal invasive intraportal transplantation. NPIs were infused in the portal system preferentially of the left liver lobe, that right liver lobe could serve as imaging negative control. NPI recipient pigs got a glucocorticoid-free immunosuppression regime. [⁶⁸Ga]Ga -DOTA-Exendin-4 PET/CT was performed once prior to islet-cell transplantation, and up to 3-times in the post-transplant period (10-15 weeks). For imaging of the NPI graft mass in the liver, [⁶⁸Ga]Ga-DOTA-Exendin-4 was injected intravenously (effectively injected activity: 19.9-82.8 MBq; amount of peptide: 0.02 ± 0.005 µg/kg), followed by a dynamic PET/CT (60 min) of the liver region. The activity was analysed 60 min p.i. in the whole liver. Finally, histological analysis of systematically sampled liver tissue was performed to identify graft localization.
Results: The proof-of-concept [⁶⁸Ga]Ga -DOTA-Exendin-4 PET/CT imaging study showed a specific signal in pig pancreas (Figure 1), that was significantly reduced in a blocking study with cold Exendin-4. In livers of NPI recipient pigs which reached a normoglycemic status (<130 mg/dl Glc, 3 of 4 pigs), the standard uptake value (SUV) of the liver significantly increased compared to the control values (61% at 7-8 weeks and 94% after 12 weeks; Figure 2). The increased liver signal was multifocal with clusters of individual hotspots, matching the histological distribution pattern of islet in the liver, which contained mainly insulin positive beta cells. No signs of immune rejection were observed in the graft recipient animals reaching normoglycemia.


Conclusion: Viable beta-cell mass after intraportal islet allotransplantation can be detected in a diabetic large animal model by using [⁶⁸Ga]Ga -DOTA-Exendin-4 in PET/CT. As changes in the viable graft mass over time could be visualized, this new diagnostic tool should also be suitable for longitudinal monitoring of intraportal islet transplants in humans.