Prospective comparison of brush cytology, viral culture, and histology for the diagnosis of ulcerative esophagitis in AIDS.

BACKGROUND & AIMS: The best method(s) for the diagnosis of esophageal ulcers in patients with acquired immune deficiency syndrome (AIDS) is unknown.

METHODS: Over an 8-year period, all human immunodeficiency virus (HIV)-infected patients with a CD4 lymphocyte count of < or =250/mm(3) and an esophageal ulcer underwent a standard protocol consisting of 2 biopsy examinations for viral culture (shell vial technique), brushing for cytology with immunohistochemistry (IHC) for viral antigens, and 10 biopsy examinations of the ulcer base/edge for routine histology and IHC. The cause of ulcer was based on histologic findings and clinical and endoscopic follow-up evaluation after therapy.

RESULTS: Forty patients with 48 episodes of ulcer were studied (mean age, 34 yr; 88% men; median CD4 lymphocyte count = 31 cells/mm(3); range, 0-250 cells/mm(3)). Six patients had more than one endoscopic examination documenting ulcer, and 2 patients had more than one cause of ulcer. The most common causes of ulcer were: idiopathic (IEU) 22, cytomegalovirus (CMV) 15, and gastroesophageal reflux disease 6. Viral culture was positive for herpes simplex in 3 of 3, but only 4 of 15 (27%) for CMV, and specificity of viral culture was 100%. Cytologic brushings were only positive in the patients with herpes simplex virus (HSV) or incidental Candida esophagitis. Viral cytopathic effect was apparent on routine histologic staining in each patient with viral esophagitis, and IHC confirmed the diagnosis in each patient. Follow-up evaluation disclosed no patient with a misdiagnosis based on histology.

CONCLUSIONS: Viral culture and cytologic brushings add little to the evaluation of esophageal ulcers in patients with AIDS over multiple biopsy specimens with routine histology.