Pitfalls in the Detection of Insulinomas With Glucagon-Like Peptide-1 Receptor Imaging
Kwadwo Antwi, Matthias Hepprich, Natasha Anouschka Müller, Jean Claude Reubi, Melpomeni Fani, Christof Rottenburger, Guillaume Nicolas, Felix Kaul, Emanuel Christ, Damian Wild
Abstract
PURPOSE: Physiological pancreaticoduodenal uptake of radiolabeled exendin-4 in Brunner glands of the proximal duodenum is the most common pitfall for false interpretation of glucagon-like peptide-1 receptor (GLP-1R) imaging. The aim of this study was to analyze the pancreaticoduodenal uptake in GLP-1R PET/CT and SPECT/CT images and to identify additional potential reading pitfalls in patients with suspected insulinoma. METHODS: A post hoc analysis of a prospective study, including 52 consecutive patients, was performed. All patients underwent 1 Ga-exendin-4 PET/CT and 2 In-exendin-4 SPECT/CT scans (4 and 72 hours postinjection) in a randomized crossover order. Three board-certified nuclear medicine physicians read all scans independently. They were unaware of other results. Reference standard was surgery with histopathological confirmation of an insulinoma/nesidioblastosis and normalization of blood glucose levels after surgery. RESULTS: There were no false-positive readings. However, there were a number of false-negative PET/CT and SPECT/CT readings, respectively: (1) due to false interpretation of uptake in the pancreaticoduodenal region (falsely interpreted as physiological uptake in Brunner glands instead of an insulinoma in 0.6% vs 9.0%), (2) due to ectopic insulinoma (0% vs 2.6%), (3) due to small insulinoma (1.9% vs 5.1%), (4) due to insulinoma overlap with kidneys (1.9% vs 4.5%), and (5) due to nesidioblastosis (0.6% and 1.9%). Pitfalls were identified in all GLP-1R PET/CT and SPECT/CT scans. CONCLUSIONS: Peripancreatic uptake, small size of an insulinoma, insulinoma overlap with kidneys, and presence of nesidioblastosis are potential pitfalls in GLP-1R imaging, which can lead to false reading results.