The classic morphological techniques for the localization of insulinomas and gastrinomas are of limited value. Endoscopic ultrasonography and somatostatin receptor scintigraphy have shown high sensitivity for the detection of gastroenteropancreatic endocrine tumors. The aim of the study was to evaluate the sensitivity of endoscopic ultrasonography and that of somatostatin receptor scintigraphy in the localization of insulinomas and gastrinomas.

This retrospective study concerned 54 patients with insulinoma (n = 29) or gastrinoma (n = 26) operated on between March 1991 and March 2000 and who had at least one among the two tested examinations. Forty-two patients had scintigraphy (17 with insulinoma, 25 with gastrinoma), 47 had endoscopic ultrasonography (28 with insulinoma, 17 with gastrinoma). One of the ten patients with MEN 1 had both tumors. All diagnosis were confirmed by histologic examination.

The sensitivity of scintigraphy for the localization of insulinomas was 47%. There was one false positive. Sensitivity of endoscopic ultrasonography for insulinomas was 85%. The sensitivity of scintigraphy in the detection of gastrinomas was 65% for the tumors in the duodenopancreatic area, 20% for the tumors in the pancreatic tail and 71% for metastasis. The sensitivity of endoscopic ultrasonography was 46% for duodenal tumors, 75% for pancreatic tumors and 57% for lymph node metastasis. The combination of both localization studies increased sensitivity to 94%.

Endoscopic ultrasonography and somatostatin receptor scintigraphy are the gold standard for localization of gastrinomas. Association of both examinations increases the sensitivity. Scintigraphy for the detection of insulinomas should be performed when endoscopic ultrasonography is negative.