Glucagon-like peptide-1 receptor PET in indolent and aggressive insulinoma: Diagnostic performance, quantitative differences, and clinical implications
摘要
To evaluate the diagnostic performance, quantitative characteristics, and clinical implications of Glucagon-Like Peptide-1 Receptor (GLP-1R) PET using 68Ga-exendin-4 in patients with indolent and aggressive insulinomas, and to refine its role in comprehensive disease assessment.
MethodsIn this prospective study, patients presenting with hypoglycemia and fulfilling Whipple’s triad were enrolled. All patients underwent 68Ga-exendin-4 PET/CT, and a subset also received 68Ga-DOTANOC PET/CT. The pathology of surgery and biopsy served as the reference standard. Patients were stratified into indolent and aggressive insulinoma groups. Diagnostic performance metrics were calculated, and logistic regression analysis was performed to identify predictors of hypoglycemia control.
ResultsNinety-six patients with suspected insulinoma were enrolled in this study (64 indolent, 25 aggressive insulinomas, and 7 non-insulinomas). For indolent insulinomas, 68Ga-exendin-4 PET/CT demonstrated a sensitivity of 98.4%, specificity of 71.4%, and accuracy of 95.8%. Detection rate for aggressive insulinomas was significantly lower than that for indolent tumors (76.0% vs. 98.4%, P = 0.002). Conversely, 68Ga-DOTANOC PET/CT demonstrated a higher positivity rate in aggressive insulinomas compared with indolent cases (92.0% vs. 67.7%, P = 0.047). In the aggressive subgroup, the number of GLP-1R-avid lesions independently predicted poor hypoglycemia control (OR = 2.0, 95% CI: 1.1–3.7, P = 0.027).
ConclusionGLP-1R PET provides excellent diagnostic accuracy for indolent insulinomas but has reduced sensitivity in aggressive disease. In contrast, somatostatin receptor PET/CT is more effective for detecting aggressive tumors. GLP-1R-positive tumor burden may serve as a potential imaging biomarker for hypoglycemia control, supporting the complementary roles of these modalities in insulinoma management.