Use of the PET ligand florbetapir for in vivo imaging of pancreatic islet amyloid deposits in hIAPP transgenic mice.

AIMS/HYPOTHESIS: Islet amyloid deposits contribute to beta cell dysfunction and death in most individuals with type 2 diabetes but non-invasive methods to determine the presence of these pathological protein aggregates are currently not available. Therefore, we examined whether florbetapir, a radiopharmaceutical agent used for detection of amyloid-ß deposits in the brain, also allows identification of islet amyloid in the pancreas. METHODS: Saturation binding assays were used to determine the affinity of florbetapir for human islet amyloid polypeptide (hIAPP) aggregates in vitro. Islet amyloid-prone transgenic mice that express hIAPP in their beta cells and amyloid-free non-transgenic control mice were used to examine the ability of florbetapir to detect islet amyloid deposits in vitro, in vivo and ex vivo. Mice or mouse pancreases were subjected to autoradiographic, histochemical and/or positron emission tomography (PET) analyses to assess the utility of florbetapir in identifying islet amyloid. RESULTS: In vitro, florbetapir bound synthetic hIAPP fibrils with a dissociation constant of 7.9 nmol/l. Additionally, florbetapir bound preferentially to amyloid-containing hIAPP transgenic vs amyloid-free non-transgenic mouse pancreas sections in vitro, as determined by autoradiography (16,475 ± 5581 vs 5762 ± 575 density/unit area, p < 0.05). In hIAPP transgenic and non-transgenic mice fed a high-fat diet for 1 year, intravenous administration of florbetapir followed by PET scanning showed that the florbetapir signal was significantly higher in amyloid-laden hIAPP transgenic vs amyloid-free non-transgenic pancreases in vivo during the first 5 min of the scan (36.83 ± 2.22 vs 29.34 ± 2.03 standardised uptake value × min, p < 0.05). Following PET, pancreases were excised and florbetapir uptake was determined ex vivo by γ counting. Pancreatic uptake of florbetapir was significantly correlated with the degree of islet amyloid deposition, the latter assessed by histochemistry (r = 0.74, p < 0.001). CONCLUSIONS/INTERPRETATION: Florbetapir binds to islet amyloid deposits in a specific and quantitative manner. In the future, florbetapir may be useful as a non-invasive tool to identify islet amyloid deposits in humans.