Localization of insulinomas - Comparison of conventional arterial stimulation with venous sampling (ASVS) and superselective ASVS

Purpose: To examine the value of superselective arterial stimulation venous sampling (ASVS) to localize insulinomas. Material and Methods: Superselective ASVS (SS-ASVS) was performed in 9 patients with insulinoma. Injection of secretagogue (calcium gluconate: 0.01 mEq Ca++/kg) was performed into the gastroduodenal, splenic (proximal and distal), and superior mesenteric arteries in 9 patients and additionally into the dorsal pancreatic artery in 6 patients. Sampling from the hepatic vein was performed to measure serum insulin concentrations at 30, 60 and 120 s after each injection of secretagogue into these arteries. SS-ASVS results were correlated with surgical findings, compared to those of conventional ASVS. Results: Insulinomas were correctly localized to the head, body or tail of the pancreas by SS-ASVS in 8 patients (89%). Conventional ASVS detected insulinomas in 7 patients (78%), although it could not distinguish whether the insulinoma was located in the pancreatic body or tail in 4 of the 7 patients. There were eight-fold or more increases in serum insulin levels in hepatic venous samples related to the artery supplying the. tumor in 8 patients. Localization of the insulinomas was verified at surgery in all patients. Conclusion: SS-ASVS is a useful method for detailed evaluation of overproduction of insulin from pancreatic insulinomas and their localization. When the pancreatic insulinoma is situated in the pancreatic body or tail, the localization is more accurately made by SS-ASVS than by conventional ASVS.