Recurrent Gastrointestinal Bleeding in a Patient with Recurrent Small Intestinal Neuroendocrine Tumors

Abstract

A 48 year-old man with a known history of duodenal carcinoid tumors presented to our institution with three days of melenic stools. The patient initially presented to our hospital in 2004 with melena and underwent esophagogastroduodenoscopy and enteroscopy which revealed four nonmucosal lesions in the second portion of the duodenum. There was no evidence of current or recent hemorrhage. Endoscopic ultrasonography demonstrated a submucosal location without invasion of the muscularis propria. Histology was consistent with gastrointestinal neuroendocrine tumor (NET), with immunostaining for chromogranin A and synaptophysin compatible with carcinoid. There were no symptoms of typical endocrine syndromes associated with gastrointestinal NET, such as the carcinoid syndrome. The patient underwent distal gastrectomy with gastrojejunostomy. He has since been admitted with recurrent episodes of melena, with negative upper and lower endoscopic evaluation, and normal enteroscopy and capsule endoscopy. He now presented to our emergency department with melena without other symptoms. He was hemodynamically stable and at his baseline hemoglobin level. Examination revealed a benign abdomen and hemoccult-positive dark brown stool. Upper endoscopy with enteroscopy of his Billroth II reconstruction revealed a normal efferent limb, and a 6mm nonmucosal lesion in the afferent limb, without stigmata of bleeding. Biopsies were consistent with a submucosal carcinoid. Colonoscopy was unrevealing. Abdominal CT did not reveal hepatic or distant metastases, and a small bowel series excluded other filling defects. Octreotide scintigraphy yielded no somatostatin receptor (SSTR)-positive foci. This case highlights the fact that gastrointestinal NETs may be multiple and recurrent. In addition, a bleeding source may be elusive despite serial endoscopic and radiologic testing. As carcinoids are not classically associated with significant luminal hemorrhage, it is possible that this gentleman continues to have obscure overt GI bleeding with concurrent (but coincidental) carcinoid disease. It is intriguing to speculate that his multiple submucosal NETs reflect the presence of other lesions within his luminal GI tract, or a syndrome associated with small bowel carcinoids. It would be interesting to pursue evaluation with MRI or octreotide scanning with other radionuclide tracers to increase the sensitivity for SSTR-positive lesions.