The index of suspicion for tuberculous pleural effusion should always be high.

Sir, Notwithstanding the implication that a high clinical suspicion of tuberculosis should be a precondition for 'a request for acid fast bacilli and tuberculosis culture' in suspected (but not confirmed) malignant pleural effusion, 1 stigmata documented in malignant pleural effusion, such as the involvement of the entire hemithorax, 1 bilateral effusions, 2 lymphocyte predominance, 1,2 haemothorax 3 and chylothorax, 4 are also the ones sometimes documented in tuber-culous pleural effusions where stigmata such as massive effusions, 5,6 bilateral effusions, 7 haemothorax 3 and chylothorax 4 have also been a feature. Accordingly, in the event that the initial presentation is pleural effusion without an identifiable primary malignant lesion, as may be the case in 7% of malignant pleural effusions, 2 stigmata common to malignancy as well as to tuberculosis should generate a high index of suspicion for tuberculosis , which should remain on the differential diagnosis even if the presentation is atypical of tuberculosis. The reason is that tuberculosis is eminently treatable if diagnosed early, and also because untreated patients, even those presenting with pleural effusion, generate the risk transmitting the disease to their immediate contacts. 8 There are few errors of omission to surpass that of mistaking tuberculosis for malignant disease, as was the case in a series from a cancer centre where, over the period 2001–2005, active tuberculosis was diagnosed in 26 patients, 18 with concomitant cancer and 8 without cancer but initially misdiagnosed as cancer. 9 The potential for misdiagnosis exists even with the use of sophisticated imaging modalities such as 18-fluorodeoxyglucose (FDG) positron emission tomography, which, in one reported instance, demonstrated multiple lesions of intense FDG uptake in the pleura, simulating malignant mesothelioma. Nevertheless, the eventual diagnosis, validated by histology of a specimen obtained from thoracoscopic biopsy (after non-diagnostic CT-guided needle biopsy), proved to be tuberculous pleurisy, 10 thanks to the high index of suspicion which prevailed against all odds. A massive haemorrhagic pleural effusion does not exclude the diagnosis of tuberculosis: a case report. The prevalence of pulmonary parenchymal tuberculosis in patients with tuberculous pleuritis. primary tuberculous pleurisy with intense 18-fluorodeoxyglucose uptake mimicking malignant mesothelioma.