Medizinische Universität Graz Austria/Österreich - Forschungsportal - Medical University of Graz
Gewählte Publikation:
Schmidt, P.
Sensitivity and Specificity of the automatic determination of Adenosine Deaminase Activity (ADA) on the COBAS-8000 system to exclude tuberculous pleurisy in a low prevalence area
Humanmedizin; [ Diplomarbeit ] Graz Medical University; 2015. pp.
[OPEN ACCESS]
FullText
- Autor*innen der Med Uni Graz:
- Betreuer*innen:
- Flick Holger
- Hönigl Martin
- Raggam Reinhard Bernd
- Altmetrics:
- Abstract:
- Aim Globally, tuberculosis (TB) is one of the most important diseases, being the second leading cause of death caused by infectious diseases after the human immunodeficiency virus. Tuberculous pleurisy occurs in 3% to 25% of the TB patients and is the second most common form of extrapulmonary TB. It is therefore important to consider TB when a pleural effusion of unknown origin is diagnosed. The adenosine deaminase activity (ADA) is found to be a sensitive and specific marker for the diagnosis of extrapulmonary TB in pleural effusions and other biological fluids. Its usefulness is well studied in countries with a high incidence of TB but not in areas with low prevalence. We therefore evaluated the sensitivity, specificity, PPV and NPV of automated determination of ADA on the Cobas® 8000 system in Styria, Austria. Methods This study was performed as a retrospective manner including patients with undiagnosed pleural effusions between 30th November 2011 and 15th February 2013. Participating hospitals were the University Hospital of Graz and the State Hospital of Hörgas-Enzenbach. In this period, ADA was analysed with the Cobas® 8000 system at the University Hospital of Graz. 65 patients with exudative pleural effusions were included and sensitivity, specificity, PPV and NPV of ADA were calculated for TB. Results The main leading cause for exudative pleural effusions among our patients (n= 65) was an underlying malignant disease (n=24; 36.9%). Second leading cause was pleuropneumonia (n=9; 13.9%), followed by unspecific chronic pleuritis (n=8; 12.3%), pleural empyema (n=7; 10.8%), cardiac decompensation (n=5; 7.7%), tuberculosis (n=4; 6.2%) and pulmonary embolism (n=3; 4.6%). The laboratory analysis of pleural fluid showed that the group of empyema had the highest levels in: the total number of cells (mean: 38.2 G/L) with highest percentage of neutrophils (mean: 79.8%), pleural LDH (mean: 4134 U/l), pleural-to-blood LDH ratio (mean: 20.3), pleural total protein (mean: 5.1 g/dl), ADA (mean: 51.5 U/l), together with lowest levels of glucose (mean: 6 mg/dl) and pH (mean: 7.17). Tuberculous pleural effusions had the highest level of lymphocytes (mean: 80.5%) as well as of lymphocyte-to-neutrophil ratio (mean: 50.4) and they followed the group of empyema with some distance in the number of total cells (mean: 4.7 G/L), LDH ratio (mean: 2.14), the level of pleural total protein (mean: 4.7 g/dl) and ADA (mean: 43.5 U/l). The highest pleural-to-blood total protein ratio were reached in the groups of pulmonary embolism (mean: 0.68), TB pleurisy (mean: 0.67) and empyema (mean: 0.65). Using a cut-off level of 25 U/l, the sensitivity, specificity, PPV and NPV of ADA for tuberculous pleural effusions were 100%, 87%, 33% and 100%, respectively. ROC analysis revealed an area under the curve of 0.926 (95% CI 0.86-0.99) for differentiating between patients with TB pleuritis from those with other etiologies for exudates. Summary In countries with a high or moderate TB prevalence pleural ADA is a routinely used marker to rule out pleural TB. In Austria M. tuberculosis is a rare cause of pleural effusions but pleural TB has to be considered in unexplained pleural exudates. However, there is only little evidence for the value of ADA determination in a low prevalence setting like in Austria. Our study showed that the sensitivity of ADA determination for TB is very high but the PPV is low, which can be explained by false positive results mainly in the group of empyema and NHL. The NPV is very high which means that results below the used ADA cut-off (25 U/l) virtually exclude the diagnosis of TB pleurisy. The study demonstrates that in low prevalence areas the ADA determination is a useful tool to rule out tuberculous pleurisy. To reduce the false positive rate and increase the PPV, ADA results have to be interpreted in conjunction with the differential cell counts or the lymphocyte-to-neutrophil ratio.