In the present study, we found that BALF from TPE patients without pulmonary tuberculosis contained increased levels of IFN-γ, TNF-α, and VEGF. Moreover, we found an increase in the CD4+/CD8+ ratio of T lymphocytes in BALF from the TPE patients.
We measured the TNF-α level in BALF of newly diagnosed, untreated TPE patients without pulmonary tuberculosis. The TNF-α level was nearly 5-fold higher in TPE patients than in the controls. In previous studies, TNF-α level has been found to be higher in BALF in patients with pulmonary TB than that in healthy controls [10, 17]. Other studies demonstrated that TNF-α level in BALF in smear-negative active pulmonary TB patients were 4.4-fold higher than that in the controls , and higher TNF-α mRNA expression levels in the lungs of mice with latent tuberculosis infection (LTBI) compared to the controls . Thus, the increase in TNF-α levels in BALF observed in the present study could result from an inflammatory reaction in response to tuberculosis infection.
VEGF is an angiogenesis and inflammatory mediator, which has been associated with TB infection activity,[13, 20] and levels decline after successful TB treatment . In our study, we found that VEGF levels in BALF from patients with TPE without pulmonary tuberculosis were 31-fold higher than that in the controls. In previous studies, VEGF was measured in serum or in pleural effusion among TB patients [12, 21]. VEGF levels in serum were approximately 2–4 times higher in patients with active pulmonary TB than in the controls [21, 22], and in pleural effusion, it was approximately 15–39 times higher in patients with TPE compared to transudates secondary to congestive heart failure [12, 13]. Another study demonstrated that VEGF levels were higher in LTBI patients than in the controls, and VEGF measurement was one of the markers used to discriminate LTBI from the controls . As for TNF-α, we suggest that an increase in VEGF levels in BALF indicates an inflammatory reaction in response to tuberculosis infection. In our study, VEGF was present in the highest concentration among the cytokines measured. VEGF levels of BALF can be elevated in acute lung injury [24, 25] and sarcoidosis . Although VEGF is not a specific for tuberculosis, it could be a sensitive inflammatory marker.
IFN-γ is known to be a key cytokine in the host immune response to tuberculosis infection. Humans with IFN-γ receptor abnormalities and IFN-γ-deficient mice have increased susceptibility to M. tuberculosis infection [8, 27]. Previous studies have reported high levels of IFN-γ in pleural effusions in TPE patients [11, 28] and in BALF from patients with active pulmonary TB [29, 30] or smear-negative active pulmonary TB . An in vitro study has also found that the level of PPD-induced IFN-γ secretion by peripheral blood mononuclear cells (PBMC) from LTBI subjects was 20 times higher than the cut-off IFN-γ level (1.0 ng/ml) . Another study found that IFN-γ mRNA expression levels in the lungs of latent TB mice were 7-times higher than that in the control mice . In our study, we found that the levels of IFN-γ in BALF from TPE patients without pulmonary tuberculosis were 12-fold higher than that in the controls. We suggest that an increase in IFN-γ level in BALF indicates an inflammatory reaction in response to tuberculosis infection, similar to the IFN-γ test for LTBI.
It has been reported that cytokines associated with a T helper-2 (Th-2) response, IL-4, and IL-10 were higher than those associated with a Th-1 response in advanced TB patients; however, Th-2 responses were lower in early stages of TB infection . In vitro stimulation of PBMC with M. tuberculosis antigens elicited Th-2 responses . Cytokine mRNA expression levels in the lungs of mice indicated that Th-2 responses in LTBI mice were lower than Th-1 responses, and also IL-10 mRNA expression levels were lower than those of the control mice . The present study showed that there was no difference in Th-2 responses, defined by IL-4 and IL-10 levels, between TPE patients and the controls. However, IL-10 was the only cytokine whose levels in TPE patients were lower than those in the controls. We suggest that Th-2 responses might have a minor role in the inflammatory reactions evident in the present study.
Increases in the CD4+/CD8+ ratio of T lymphocytes in BALF from active pulmonary TB patients, as compared to controls, have previously been reported [32, 33]. Another study has demonstrated that the CD4+/CD8+ ratio was higher in susceptible TB mice . In the present study, we showed that the CD4+/CD8+ ratio of T lymphocytes changes in patients without active TB infection.
Diagnosis of TPE based on ADA levels in pleural effusion may have a limitation. The positive predictive value (PPV) of pleural fluid ADA measurement in patients with TPE is depends on local prevalence of the disease . The prevalence of TPE among all studied pleural effusions in our institution was about 23% between 2005 and 2010. This finding suggested that ADA has more than 80% of PPV in patients with TPE in our institution. In addition to high PPV value of ADA, all TPE patients in the present study were cured with anti-tuberculosis medication.
Another potential limitation of the present study is sensitivity of detecting active pulmonary tuberculosis by simple chest radiograph. High-resolution computed tomographies of patients with pleural tuberculosis were more sensitive than simple chest radiographs in detecting pulmonary tuberculosis activity . We did not perform computed tomography of the study subjects. Although the present cases were all negative of BAL fluid MTB-PCR, AFB stain, and AFB culture, there could be a little possibility of active tuberculosis in the different lung region other than BAL area.