Endoscopic evaluation of chronic human immunodeficiency virus-related diarrhea: is colonoscopy superior to flexible sigmoidoscopy?

Abstract

In patients with chronic human immuno-deficiency virus (HIV)-related diarrhea undergoing lower endoscopy, the decision to perform flexible sigmoidoscopy or colonoscopy is controversial. The purpose of this study is twofold: 1) to evaluate the diagnostic yield of colonoscopy in a large group of patients with chronic HIV-related diarrhea and negative stool studies, and 2) to determine whether colonoscopy is superior to flexible sigmoidoscopy in this setting. Methods: All HIV-infected patients with chronic diarrhea who were referred for diagnostic colonoscopy at Bellevue Hospital Center between January 1992 and December 1996 were identified. Patient charts, pathology reports, and endoscopy records were reviewed. Results: During the 5-yr study period, 317 consecutive patients with chronic unexplained diarrhea undergoing colonoscopy were identified. A potential cause of diarrhea was found in 116 patients (36.6%). Cytomegalovirus was the most common pathogen detected (24%). The yield of colonoscopy was significantly higher in patients with a CD4 count of <100 cells/mm 3 than in those with higher CD4 counts (44.8% vs 6.4, p < 0.0001). Thirty percent of pathogens and 75% of lymphomas were identified only on biopsies taken from the proximal colon, well beyond the reach of the flexible sigmoidoscope. Importantly, 94% of the pathogens that were found only in the proximal colon were organisms for which effective therapy is currently available. Conclusions: Colonoscopy is superior to flexible sigmoidoscopy in HIV-infected patients with chronic unexplained diarrhea. If flexible sigmoidoscopy had been performed instead of colonoscopy, 30% of pathogens would have been missed and 75% of lymphomas would have escaped detection.