1552-P: Homeostasis Model Assessment of ß-Cell Function for Diagnosis of Insulinoma

Abstract

Insulinoma is characterized by uncontrolled hypersecretion of insulin and is currently diagnosed by the 72-h fasting test. Guidelines by Endocrine Society (ES), European (ENETS) and North American (NANETS) Neuroendocrine Tumor Societies provide divergent diagnostic criteria at the end of the fasting test, but parameters assessing ß-cell function noninvasively are not included. This study tested whether homeostasis model assessment of ß-cell function (HOMA-B) as a surrogate for insulin secretion at the end of the fasting test, provide comparable sensitivity and specificity for diagnosis of insulinoma and thus might help to improve current diagnostic guidelines. In 104 persons with suspected insulinoma, 72-h fasting tests were performed with frequent assessment of glucose, insulin, and C-peptide in venous blood. HOMA-B was calculated at time of lowest glucose concentration at the end of fasting test. Follow-up data of patients were collected to exclude false-negative diagnosis of insulinoma. The HOMA-B at the end of the fasting test was >6.5-fold higher in persons with than without insulinoma (insulin and C-peptide, both p<0.001). ENETS/NANETS and ES criteria (insulin and C-peptide) reached a diagnostic sensitivity of ≤0.83 (≤0.61-0.95) and ≤0.96 (≤0.78-1.00) as well as specificity of ≤1.00 (≤0.96-1.00) and ≤0.85 (≤0.76-0.92). The HOMA-B provided high diagnostic sensitivity (cut-off using insulin >253 a.u. and C-peptide >270 a.u.; both 0.96, 0.78-1.00) and specificity (≥0.96, ≥0.89-0.99) at the end of the fasting test for diagnosis of insulinoma (n=23). Using both insulin and C-peptide for HOMA-B, sensitivity even tended to be higher than that of ENETS/NANETS using insulin only (p=0.063) and specificity was higher than that of ES using both insulin and C-peptide (p≤0.002). HOMA-B with defined cut-offs using insulin and C-peptide at the end of the fasting test showed excellent diagnostic sensitivity and specificity and might represent an alternative and precise tool to diagnose insulinoma. Disclosure K.Bódis: None. M.Schön: None. L.Dauben: None. M.Wilker: None. K.Strassburger: None. V.Burkart: None. M.Roden: Advisory Panel; Eli Lilly and Company, Consultant; TARGET PharmaSolutions, Inc., Research Support; Boehringer-Ingelheim, Novo Nordisk, Novartis, Sanofi. K.Müssig: None.