Colon in and Polyp Out

Abstract

Introduction: Tagged red blood cell (RBC) scan has a limited but unique role in the evaluation of overt gastrointestinal bleeding. We present a case where a patient was offered total colectomy at an outside hospital solely based on a tagged RBC scan that suggested a major colonic bleed, when in fact the bleeding was from a duodenal polyp on enteroscopic evaluation. Case Report: 87 year old man with a history of Billroth II surgery for a bleeding peptic ulcer in the 1980s was transferred for further evaluation of gastrointestinal bleeding. He had history of dark stools for 3 weeks. At outside facility, a tagged RBC scan was performed that identified the bleed was localized to his left upper quadrant. An endoscopy was not performed, instead a colectomy was offered. Upon his refusal to surgery, he was transferred to our facility. On our examination, the patient had melena. An esophagogastroduodenoscopy was performed that revealed Billroth II anatomy and a 2cm bleeding polyp with a thick stalk in the afferent limb. A snare-cautery polypectomy was performed. The pathology revealed that the polyp was a benign polypoid duodenal vascular malformation. The tagged RBC scan was later reviewed with radiology, which showed two areas of tracer activity. One was near the mid-abdomen, which corresponded to the polypectomy site, and the other was in the left upper quadrant, which was suspicious of a simultaneous colonic bleed. The next day a colonoscopy was performed which showed only melenic stool in the splenic flexure without evidence of an active bleed. Discussion: Tagged RBC scan has many merits but it does have significant drawbacks as well. Chief among them are non-localization and erroneous localization of bleeding. In a retrospective studyinvolving80 patients, tagged RBC scans were positive in 36% of overt GI bleed cases and negative in 64% cases. Among the positive cases, 10% were determined to be “incorrect positive” and led to incorrect surgeries in 5 out of 8 cases, 15% cases were “truly positive”, and 11% were “unconfirmed positive”. This led to a positive predictive value of 41%. Among the negative cases (64%), 21% were “false negatives” and 34% were “unconfirmed negative”. Its accuracy diminished proximally, with more false negatives reported, when bleeding was in the gastroduodenal region. Conclusion: Results of RBC scan should be interpreted with caution with endoscopic correlation.Figure 1Figure 2Figure 3