Combining ADA, protein and IFN-gamma best allows discrimination between tuberculous and malignant pleural effusion

Abstract

Background: The purpose of this study is to assess the usefulness of various enzymes, cytokines and biochemical studies of pleural fluid for the differential diagnosis of tuberculosis from malignant pleural effusions, and to clarify the role of combining diagnostic tests.

Methods: The study group included 39 cases with tuberculous effusions and 31 cases with malignant effusions, whose diagnoses were confirmed by pleural biopsy, cytology or microbiological methods. We compared pleural fluid levels of ADA, TNF-alpha, IFN-gamma, IL-2, IL-6, IL-8, pH, protein, glucose, cholesterol, triglyceride, amylase and lactic dehydrogenase between tuberculous and malignant effusions. Using stepwise logistic regression analysis, we evaluated the benefit of combining various parameters. Receiver operating characteristic(ROC) curves of ADA, cytokines and equations generated from regression analyses were plotted and compared with the area under curve(AUC). Cut-off values showing the best diagnostic accuracy were selected and compared.

Results: Compared to malignant effusion, tuberculous effusion showed significantly higher levels of ADA, IFN-gamma, TNF-alpha and IL-2. There was a good correlation between IFN-gamma and TNF-alpha. By stepwise logistic regression analysis, IFN-gamma, protein and ADA were independent variables predicting tuberculous from malignant effusions. The diagnostic accuracy and AUC of regression equation was greater than any other single parameters.

Conclusion: For the differential diagnosis of tuberculosis and malignant pleural effusions, combining ADA, protein and IFN-gamma best allows discrimination.

Fig. 1.

Receiver operating characteristics (ROC) curves for IFN-γ, TNF-α, ADA, IL-2, IL-6, Protein, and regression equation for the differential diagnosis of tuberculous from malignant effusion. Area under curve (AUC): IFN-γ (0.84), TNF-α (0.81), ADA (0.84), IL-2 (0.83), IL-6 (0.69), Protein (0.62), RE (0.92) Abbreviations: see Table 3., RE: regression equation (see Table 5).