Diagnosing gastro-intestinal tuberculosis: is the issue acid-fast bacilli?

To the Editor—I read with interest the case report on oesophageal tuberculosis mimicking carcinoma published in the December issue of the Hong Kong Medical Journal.1 We encountered a similar case where the ulcer was located in the proximal half of the oesophagus and had an elevated ulcerated edge. Our patient also had abnormalities in the stomach and the duodenum, an external compression of the gastric greater curve and an ulcerated mass in the duodenum. The biopsies of the oesophageal ulcer showed caseating granulomas without evidence of malignancy. However, the biopsies from the duodenum and the stomach lesions were positive for acid-fast bacilli (AFB). We have also encountered cases where the findings showed changes consistent with tuberculosis (TB), but yielded no evidence of AFB. Diagnosing TB infections has been reported to be difficult particularly in centres with little experience.2,3 In our experience with gastro-intestinal and peritoneal TB, only 47% and 50% respectively had AFB isolated, including isolation from other sites.4,5 Relying on AFB identification to make a diagnosis contributes to treatment delay. Evidence of previous TB infections and consistent histological changes are adequate for diagnosis. To date, we have only encountered one case where suspected TB of the biliary hilar strictures (presence of granuloma without AFB) turned out to be an inoperable cholangiocarcinoma. Despite the risk of hepatotoxicity, anti-TB treatment is generally safe, if monitored closely. Any delay in initiating therapy can lead to significant morbidity and even death.