In order to localise neuroendocrine tumours of the foregut type (that is, of the stomach, duodenum, and pancreas), 18 patients were studied prospectively by endoscopic ultrasonography, computed tomography, transabdominal ultrasonography, magnetic resonance imaging, and somatostatin receptor scintigraphy. These 18 patients had a total of 25 primary tumour lesions which were verified histologically in tissue obtained by surgery or by ultrasound or endoscopy guided biopsy.Tumours were found in the stomach (n=-), duodenum (n=6), pancreas (n=17), and liver (n=l). Endoscopic ultrasonography had the highest sensitivity for tumour detection, followed by somatostatin receptor scintigraphy, computed tomography, transabdominal ultrasonography, and magnetic resonance imaging (88%, 52%, 36%, 32%, and 24% respectively). Endoscopic ultrasonography was especially sensitive in tumours smaller than 2 cm in diameter (88% v somatostatin receptor scintigraphy 35%; computed tomography 12%; transabdominal ultrasonography 6%; and magnetic resonance imaging 0%). Of 17 tumours located in the pancreas, endoscopic ultrasonography showed a sensitivity of 94% (somatostatin receptor scintigraphy 47%; computed tomography 47%; transabdominal ultrasonography 41%; and magnetic resonance imaging 29%). Of eight extrapancreatic tumours, six were identified by endoscopic ultrasonography, five by somatostatin receptor scintigraphy, and only one by computed tomography, transabdominal ultrasonography, and magnetic resonance imaging. One neuroendocrine tumour that was not detected by endoscopic ultrasonography was correctly identified by somatostatin receptor scintigraphy. Endoscopic ultrasound allowed correct determination ofthe tumour size and tumour spread into parapancreatic structures, especially the large vessels (T stage), in all 14 patients operated upon. The lymph node stage (N stage) was correctly determined in 10 of these 14 patients. In summary, endoscopic ultrasonography and somatostatin receptor scintigraphy were the most sensitive imaging methods for the localisation of these tumours and should be used as early diagnostic procedures to accurately stage neuroendocrine tumours of the foregut type. (Gut 1994; 35: 471-475)

The factors that determine the hormone and volume responses of pituitary adenomas to the somatostatin analog octreotide are poorly understood. We, therefore, studied the correlation between 111indium-pentetreotide somatostatin receptor scintigraphy (SRS) and the clinical and immunohistochemical classification of pituitary adenomas, on the one hand, and hormone and volume responses, on the other hand. Ten patients with GH-secreting (6 females and 4 males; age, 31-67 yr) and 14 patients with clinically nonfunctioning (NF) macroadenomas (5 females and 9 males; age, 22-79 yr) were preoperatively treated with 300 micrograms/day octreotide, which was increased to 600 and 1500 micrograms/day at weekly intervals and then continued for at least 3 months until surgery. SRS was performed before therapy. A sellar magnetic resonance imaging scan was performed before therapy; 1, 2, and 3 weeks and 3 months after start of therapy; and after surgery. Acromegalics also had an 8-h GH profile, insulin-like growth factor-I determination, and a 100-g oral glucose load at these time points. An attempt was made to identify NF adenomas as gonadotroph adenomas using their LH, FSH, and alpha-subunit responses to TRH. In acromegalic patients, octreotide suppressed mean GH (8-h profile) and insulin-like growth factor-I concentrations from 34.9 +/- 9.7 to 8.1 +/- 3.6 micrograms/L and from 2122 +/- 1025 to 701 +/- 208 micrograms/L, respectively, after 3 months. Significant (26-85% decline) tumor shrinkage occurred in 5 of 10 patients, mainly within the first week. Tumor shrinkage and GH suppression were not correlated. Four of 7 patients had increased pituitary 111indium-pentetreotide uptake, but this did not predict GH suppression or tumor shrinkage. Of the NF adenomas, 2 responded with shrinkage (57% and 96% decline). Four of 12 adenomas had increased 111indium-pentetreotide uptake, but this did not correlate with tumor shrinkage (2 adenomas; 1 gonadotroph and 1 null cell adenoma), immunohistochemistry, or clinical classification. We conclude that preoperative octreotide therapy suppresses GH in most patients and reduces tumor volume in up to 50% of acromegalic patients. It also induces shrinkage in some NF adenomas, although less frequently. SRS does not predict shrinkage of either tumor type. Shrinkage does not correlate with clinical classification or immunohistological characteristics. Further studies are needed to identify the factors that determine the hormone and volume responses of pituitary adenomas to octreotide therapy.

No abstract

T HE Zollinger-Ellison syndrome manifests itself in 90 to 95 percent of cases as severe peptic ulcer disease. 1 About half the time, the ulcer disease is associated with diarrhea, and in approximately 10 percent of patients, diarrhea is the only clinical manifestation. 2,3 Basal serum gastrin concentrations are usually elevated in patients with this syndrome. 4 Normal basal serum gastrin values with abnormal results of secretin tests have been reported, however, in a few patients. [5][6][7] We describe a patient with a duodenal gastrinoma, a two-year history of diarrhea, and no peptic ulcer disease. The patient had persistently normal serum gastrin concentrations, but abnormal results of a secretin test and increased serum concentrations of the total progastrin product. C ASE R EPORTA 61-year-old man was admitted to the hospital in February 1993 with a two-year history of diarrhea (8 to 10 watery stools per day) and episodes of epigastric pain, but no weight loss. Before this admission, upper and lower gastrointestinal endoscopy, as well as biopsies and an extensive search for infections and parasitic organisms, had been performed. No abnormalities had been found. Treatment with pancreatin, loperamide, and cimetidine by the referring physician had had no effect on the diarrhea.On admission, the stool weight was 1010 g per day; the stool contained normal amounts of fat (6.7 g per day) and chymotrypsin (17.5 U per gram of stool weight). The concentrations of hemoglobin and of serum creatinine, total bilirubin, alkaline phosphatase, aminotransferases, total protein, calcium, and phosphorus were normal, as were the white-cell count and the results of serum protein electrophoresis. Additional laboratory tests, such as those for serum vitamin B 12 , folic acid, thyroid hormones, thyrotropin, gastrin (on three occasions), vasoactive intestinal peptide, and serotonin and for urinary 5-hydroxyindoleacetic acid, were normal, as were the results of tests for malabsorption, including a Schilling test and breath tests for D -xylose, and H 2 -lactose, H 2 -lactulose, and H 2 -glucose. Transabdominal ultrasonography, computed tomography of the abdomen, radiographic examination of the small intestine, and colonoscopy, including histologic examination of biopsy specimens, were also normal. Endoscopy of the upper gastrointestinal tract showed prominent duodenal mucosal folds. Histologic examination of duodenal-biopsy specimens showed shortening of the villi, with moderate infiltration by inflammatory cells. Celiac disease was suspected, and the patient began to follow a gluten-free diet.The patient was sent home in March 1993 and readmitted in October 1993. Despite a gluten-free diet, diarrhea and histologic abnormalities on subsequent duodenal biopsies persisted. Parenteral nutri-tion was started. Nevertheless, the stool weight was 1070 g per day, and the fat and chymotrypsin content was again found to be normal, indicating secretory diarrhea. Basal serum gastrin concentrations were normal on three occasions (20, 32, and 46 pg pe...

In a prospective study of 13 patients (three males and 10 females; mean age 53 [8-82] years) the value of somatostatin receptor scintigraphy (SRS) and endoscopic ultrasonography (EUS) was compared with transabdominal ultrasound (US), computed tomography (CT) and magnetic resonance imaging (MRI) in the diagnosis of insulinoma (six patients) or gastrinoma (seven patients). There were ten separate primary lesions of each tumour type, proven histologically. For insulinoma the sensitivity of EUS was 90%, SRS 10% and CT, US and MRI together 20%. For gastrinoma the sensitivity of EUS was 90%, SRS 100%, and 30% for the other three methods together. Thus EUS had a high diagnostic localizing sensitivity for both tumours, while SRS was highly sensitive only in the diagnosis of gastrinoma, not insulinoma. The value of CT, MRI and conventional ultrasonography lies in their ability to visualize distant metastases.