Localization of Pancreatic Endocrine Tumors by Endoscopic Ultrasonography

Abstract
After a pancreatic endocrine tumor has been diagnosed on the basis of clinical signs and the results of laboratory tests, localization of the tumor by the usual imaging procedures fails in as many as 40 to 60 percent of patients. Endoscopic ultrasonography, a sensitive test for small carcinomas of the pancreas, might also be useful in patients with endocrine tumors of the pancreas that cannot be localized by conventional methods. We studied 37 patients later shown to have 39 endocrine tumors of the pancreas who had negative results on transabdominal ultrasonography and CT. All the patients underwent endoscopic ultrasonography, and 22 also underwent selective angiography. All the tumors were confirmed by surgical excision and immunohistologic examination; they consisted of 31 insulinomas, 7 gastrinomas, and 1 glucagonoma, 0.5 to 2.5 cm (mean, 1.4 cm) in diameter. All but one of the patients were cured of their disease, as ascertained by at least six months of clinical and laboratory follow-up. Using endoscopic ultrasonography, we were able to localize 32 of the 39 tumors (sensitivity, 82 percent); no tumor was incorrectly localized. The size of the tumors was very similar (within 2 mm) to that predicted by endoscopic ultrasonography. Among the 22 patients who underwent both angiography and endoscopic ultrasonography, ultrasonography was significantly more sensitive than angiography for tumor localization (sensitivity, 82 percent vs. 27 percent). Among 19 control patients without pancreatic endocrine tumors, endoscopic ultrasonography was negative in 18 (specificity, 95 percent). Endoscopic ultrasonography is a highly sensitive and specific procedure for the localization of pancreatic endocrine tumors. It should be considered for the preoperative localization of such tumors once the clinical and laboratory diagnosis has been established. (N Engl J Med 1992;326:1721–6.)