Chest radiographic findings of pulmonary tuberculosis in severely immunocompromised patients with the human immunodeficiency virus

OBJECTIVE

We describe chest radiograph (CXR) findings in a population with a high prevalence of human immunodeficiency virus (HIV) and tuberculosis (TB) in order to identify radiological features associated with TB; to compare CXR features between HIV-seronegative and HIV-seropositive patients with TB; and to correlate CXR findings with CD4 T-cell count.

METHODS

Consecutive adult patients admitted to a national referral hospital with a cough of duration of 2 weeks or longer underwent diagnostic evaluation for TB and other pneumonias, including sputum examination and mycobacterial culture, bronchoscopy and CXR. Two radiologists blindly reviewed CXRs using a standardised interpretation form.

RESULTS

Smear or culture-positive TB was diagnosed in 214 of 403 (53%) patients. Median CD4+ T-cell count was 50 cells mm(-3) [interquartile range (IQR) 14-150]. TB patients were less likely than non-TB patients to have a normal CXR (12% vs 20%, p = 0.04), and more likely than non-TB patients to have a diffuse pattern of opacities (75% vs 60%, p = 0.003), reticulonodular opacities (45% vs 12%, p < 0.001), nodules (14% vs 6%, p = 0.008) or cavities (18% vs 7%, p = 0.001). HIV-seronegative TB patients more often had consolidation (70% vs 42%, p = 0.007) and cavities (48% vs 13%, p < 0.001) than HIV-seropositive TB patients. TB patients with a CD4+ T-cell count of ≤ 50 cells mm(-3) less often had consolidation (33% vs 54%, p = 0.006) and more often had hilar lymphadenopathy (30% vs 16%, p = 0.03) compared with patients with CD4 51-200 cells mm(-3).

CONCLUSION

Although different CXR patterns can be seen in TB and non-TB pneumonias there is considerable overlap in features, especially among HIV-seropositive and severely immunosuppressed patients. Providing clinical and immunological information to the radiologist might improve the accuracy of radiographic diagnosis of TB.

Serial chest radiographs in the management of children with a clinical suspicion of pulmonary tuberculosis

The aim of the study was to define the role of serial chest radiographs (SCR) in the management of children with a clinical suspicion of pulmonary tuberculosis (PTB) and to determine the interval at which they should be taken. Eighty children with a clinical suspicion of PTB were studied and followed-up for a duration of 18 months. SCR during the time of treatment were taken at monthly intervals for the first 3 months, then at 2-monthly intervals up to the end of therapy, and finally 2 months post-therapy. These were reviewed and the changes while on treatment noted and correlated with the clinical picture. Lung opacities were observed in 73 children (91 per cent) and were the most common radiological finding on the initial chest X-ray. These were followed by reduced chest wall muscle bulk present in 66 children (83 per cent). Mediastinal and/or hilar lymphadenopathy was noted in 47 children with a significant occurrence in the 0-4 age group (p = 0.004). Pleural effusions, cavities and calcification were rare. Human immunodeficiency virus (HIV) seropositive children with PTB accounted for 87 per cent and carried a poor prognosis (p = 0.0007). The common chest radiographic findings in children with PTB include lung opacities with hilar/mediastinal lymphadenopathy. Pleural effusions, cavitation, calcifications, miliary spread and normal chest X-rays were rare. SCR are useful in monitoring response to treatment, detection of onset of secondary infections and complications. HIV positive patients carry a poor prognosis. Based on the results of this study, pre-treatment, 2 months after onset of treatment, and end of therapy radiographs are recommended as routine in children with a clinical suspicion of PTB.