Endocrine Abstracts

Pre-operative localisation of insulinomas is essential prior to resection. Many imaging modalities have been used. Angiography is well described for localising neuroendocrine tumours and the use of selective intrarterial calcium injection to stimulate insulin release with simultaneous sampling of insulin levels in the hepatic vein has been shown to allow localisation of the tumour to particular vascular territories, even when other modalities have not been helpful.

We have audited our experience with this investigation since 1998. Of 21 patients identified all but three had biochemical proof of an insulinoma. In 2 cases there was subsequent evidence of factitious hypoglycaemia and one patient with MEN 1 had sampling during investigation of a gastrinoma. In all patients with proven insulinoma the Calcium Stimulation was able to localise a tumour, including in 8 patients where cross-sectional imaging had failed to do so.

Previous studies suggested that the dose of intra-arterial calcium should be adjusted for the patient weight and suggested that a rise in insulin levels to more than twice basal levels indicated a tumour in that vascular territory. We now use a fixed dose of 1ml 10% calcium gluconate. We have found that a ratio of >3.5 was found in at least one artery in all patients with proven insulinomas whilst the 3 patients without insulinomas did not have any ratios of >2.5. We would therefore suggest that the peak:basal ratio used to confirm the localisation of a tumour should be taken as 3.5.

In 9 patients C-peptide levels were measured in addition following Calcium Stimulation. These did not add any further diagnostic value.

In conclusion we have described the largest series of intra-arterial calcium stimulation for localising insulinomas and have confirmed its sensitivity even when other modalities have not identified a tumour. We have suggested refinements to the interpretation of this test.