S Jain1, N Kumar, S K Jain. 1. Departments of Pathology and Medicine, Maulana Azad Medical College and Lok Nayak Hospital, New Delhi, India.
Abstract
OBJECTIVE: To highlight the utility of endoscopic brush smears in the diagnosis of gastric tuberculosis in clinically unsuspected cases. STUDY DESIGN: A retrospective analysis of endoscopic brush smears from 210 patients with gastric symptoms. In seven of these patients (3.3%) the possibility of gastric tuberculosis was suggested in Giemsa-stained smears. Biopsy was available in all cases. Ziehl-Neelsen stain to demonstrate tubercle bacilli was used in brush smears and biopsies in seven and three cases, respectively. RESULTS: Endoscopically the sites involved were antrum (two), pylorus (two), pylorus and duodenum (three). One patient had an ulcerative lesion, and six had growths. Granulomas and/or epithelioid cells were seen in brush smears in all cases. Tubercle bacilli could be demonstrated in cytologic smears in four cases. Endoscopic biopsy showed granulomas in five cases and non-specific gastritis in two. Tubercle bacilli could not be demonstrated in any of the biopsy sections. On further clinicoradiologic investigation, two patients were found to be follow-up cases of pulmonary and nodal tuberculosis. Enzyme-linked immunosorbent assay for HIV, done in three cases, was negative. A final diagnosis of primary gastric tuberculosis in five patients and secondary in two was considered. Six patients responded to antituberculosis treatment and showed healing of the lesions on repeat endoscopy after six months of therapy, while one was a recent case with four weeks' follow-up. CONCLUSION: Endoscopic brush cytology is a reliable modality for the diagnosis of gastric tuberculosis.
OBJECTIVE: To highlight the utility of endoscopic brush smears in the diagnosis of gastric tuberculosis in clinically unsuspected cases. STUDY DESIGN: A retrospective analysis of endoscopic brush smears from 210 patients with gastric symptoms. In seven of these patients (3.3%) the possibility of gastric tuberculosis was suggested in Giemsa-stained smears. Biopsy was available in all cases. Ziehl-Neelsen stain to demonstrate tubercle bacilli was used in brush smears and biopsies in seven and three cases, respectively. RESULTS: Endoscopically the sites involved were antrum (two), pylorus (two), pylorus and duodenum (three). One patient had an ulcerative lesion, and six had growths. Granulomas and/or epithelioid cells were seen in brush smears in all cases. Tubercle bacilli could be demonstrated in cytologic smears in four cases. Endoscopic biopsy showed granulomas in five cases and non-specific gastritis in two. Tubercle bacilli could not be demonstrated in any of the biopsy sections. On further clinicoradiologic investigation, two patients were found to be follow-up cases of pulmonary and nodal tuberculosis. Enzyme-linked immunosorbent assay for HIV, done in three cases, was negative. A final diagnosis of primary gastric tuberculosis in five patients and secondary in two was considered. Six patients responded to antituberculosis treatment and showed healing of the lesions on repeat endoscopy after six months of therapy, while one was a recent case with four weeks' follow-up. CONCLUSION: Endoscopic brush cytology is a reliable modality for the diagnosis of gastric tuberculosis.