Chronic destructive pulmonary tuberculosis: assessment of disease activity by computed tomography

Abstract

Determination of disease activity of chronic destructive pulmonary tuberculosis (TB) on imaging studies can be difficult because several imaging findings due to disease chronicity such as a residual cavity can be misinterpreted as an active disease. To evaluate computed tomography (CT) findings to predict active disease in patients with chronic destructive pulmonary TB. CT findings of 36 patients with chronic active destructive pulmonary TB and 78 patients with chronic inactive destructive pulmonary TB were reviewed and their patterns of lung lesions were compared. Statistical comparisons were performed using chi-square and Student's T tests for univariate analyses, and a stepwise logistic regression method was used for multivariate analysis. Based on univariate analyses, cavitary destruction (P = 0.015), non-branching centrilobular nodules (P < 0.001), tree-in-bud pattern (P < 0.001), airspace nodules (P < 0.001), and cavities in other lobes (P = 0.001) were more frequently seen in chronic active destructive pulmonary TB. A stepwise logistic regression analysis demonstrated that tree-in-bud pattern (odds ratio, 52.3; 95% confidence interval, 6.2-437.2; P < 0.001) were significant CT findings associated with active disease. Tree-in-bud pattern were the most characteristic CT findings to predict active disease in patients with chronic destructive pulmonary TB.