Endocrine Abstracts

Objective: to evaluate the role of pre-operative insulinoma localisation using selective intra-arterial calcium gluconate stimulated hepatic venous sampling.

Method: Calcium gluconate (0.025 milliEquivalents Ca per kilogram) is used as a secretagogue to promote insulin secretion. Simultaneous catheterisation of right hepatic vein and pancreatic arterial supply were performed. The secretagogue was infused into the gastroduodenal, splenic, common hepatic and superior mesenteric arteries. Right hepatic venous samples were collected before and 30, 60, 120 seconds after each injection. Plasma insulin peak of at least two-fold the baseline is indicative of insulinoma in the region supplied by the selected artery. Ten patients with biochemically proven inappropriately high insulin levels when hypoglycaemic during fasting underwent selective intra-arterial calcium gluconate stimulated hepatic venous sampling between January 2000 and October 2003. Results obtained were compared with pre-operative findings of CT pancreas (n=10), endoscopic ultrasonography (n=3) and with the intraoperative findings.

Results: A definitive localisation by this technique was obtained in eight patients which correlated with intraoperative findings (80% sensitivity). Two patients showed no localisation. One of the two had unusual pancreatic arterial anatomy making interpretation of sampling results impossible and had decided against surgery. The other patient had a positive endoscopic ultrasonography, negative CT but has so far refused surgery and has improvement of symptoms on medical treatment. No adverse effects of sampling were encountered. CT pancreas was positive in two of ten. Endoscopic ultrasonography was suggestive in two of three.

Conclusion: Selective intra-arterial calcium gluconate hepatic venous sampling in our series was a highly sensitive and safe method for pre-operative localisation of insulinoma, proving positive in a number of cases where other imaging techniques had failed to locate the tumour.