Professional Documents
Culture Documents
Diagnostic Performance of Adenosine Deaminase Activity in Pleural Fluid A Single-Center
Diagnostic Performance of Adenosine Deaminase Activity in Pleural Fluid A Single-Center
Original article
a r t i c l e i n f o a b s t r a c t
Article history: Objective: To determine the diagnostic utility of adenosine deaminase (ADA) in a large series of pleural
Received 12 September 2009 effusions of different etiologies.
Received in revised form 6 December 2009 Methods: A retrospective study of 2104 consecutive patients presenting with pleural effusion was carried out
Accepted 11 March 2010 at a Spanish university hospital. ADA levels in pleural fluid were determined using a non-Giusti automatic
Available online 1 May 2010
kinetic assay, and a receiver operating characteristics curve analysis was applied to estimate their
discriminative properties.
Keywords:
Results: Pleural tuberculosis (TB) accounted for 221 (10.5%) effusions. Pleural fluid ADA N 35 U/L yielded 93%
Adenosine deaminase
Tuberculosis
sensitivity, 90% specificity, a positive likelihood ratio (LR) of 10.05 and a negative LR of 0.07 for the diagnosis of TB
Pleural effusion among lymphocytic exudates. The ADA activity was significantly higher in neutrophil- (111.6 U/L) than in
Empyema lymphocyte-rich (62.4 U/L; p = 0.002) TB effusions. Overall, more than 40% of parapneumonics and half of
Lymphoma lymphomatous effusions exceeded the cutoff set for TB. These were the only causes of ADA activity above 250 U/L.
When the prevalence of TB as a cause of exudative effusions is low (e.g., 1%), the estimated positive predictive
value of the ADA test may be as low as 7%, although the negative predictive value remains high (99.9%).
Conclusion: Where available, pleural ADA should be routinely used to rule TB in or out in areas with moderate to
high or low TB prevalence, respectively. A high ADA level is a characteristic not only of lymphocytic, but also of
neutrophilic TB effusions. An extremely high ADA activity should raise suspicion of empyema or lymphoma.
© 2010 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
1. Introduction examine pleural fluid has not been as widespread as would be expected,
particularly in countries with low incidence of TB.
Since its description 30 years ago [1], the measurement of pleural Some of the design flaws of the primary studies included in the meta-
fluid adenosine deaminase (ADA) activity is considered by some to be analyses can be avoided by simply using a large series of consecutive
an important reference standard for identifying pleural tuberculosis patients from a single center, as we did here. Our purpose was to
(TB) in clinical practice [2,3], while by others it is merely an aid to describe the operating characteristics of pleural fluid ADA in the
differential diagnosis of pleural effusion [4]. Central to this contro- differential diagnosis of pleural effusions. Our hypothesis was that in an
versy is whether pleural ADA can supplant pleural biopsy for the area with a moderate incidence of TB, such as Catalonia (Spain) [9], ADA
evaluation of suspected TB. has enough sensitivity and specificity to confirm or exclude TB, thus
Four meta-analyses on the diagnostic accuracy of pleural fluid ADA allowing immediate therapeutic decisions to be made in patients with
have been published [5–8]. However, differences in ADA measurement lymphocytic exudates. In addition, we estimated the predictive values of
methods (Giusti's colorimetric method vs non-Giusti), as well as prob- the pleural ADA test in different geographical areas, according to the
lems of publication bias (e.g., selection of English-language databases hypothetical prevalence of TB among patients presenting with an
and journals) and the modest methodological quality of some selected exudative effusion. Finally, we investigated whether or not the
studies raise questions concerning the reliability of the pooled measured predominant white blood cell count in pleural fluid, either neutrophil
efficacy of the test. Despite these studies having concluded that the or lymphocyte, influences ADA activity in TB patients.
test performance is reasonably good, incorporation of ADA to routinely
2. Methods
2.1. Subjects
⁎ Corresponding author. Department of Internal Medicine, Arnau de Vilanova
University Hospital, Avda Alcalde Rovira Roure 80, 25198 Lleida, Spain. We retrospectively reviewed all consecutive patients with pleural
E-mail address: jporcelp@yahoo.es (J.M. Porcel). effusion who underwent a diagnostic thoracentesis at the Arnau de
0953-6205/$ – see front matter © 2010 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.
doi:10.1016/j.ejim.2010.03.011
420 J.M. Porcel et al. / European Journal of Internal Medicine 21 (2010) 419–423
Vilanova University Hospital (Lleida, Spain) since January 1994 to July coefficient described the linear association between ADA and other
2009, for whom pleural ADA was available. We recorded demographic, pleural fluid biochemical parameters. Two-tailed p values ≤0.05 were
serum (protein, lactate dehydrogenase — LDH) and pleural fluid data considered statistically significant. All statistical analyses were con-
(cell count and differential, protein, LDH, ADA) as well as the final ducted with a statistical software package (SPSS, version 15.0; SPSS Inc;
diagnoses. If a patient had been submitted to repeated thoracentesis, Chicago, IL, USA).
only the results of the first were considered. The local ethics committee
approved the study protocol. 3. Results
A pleural effusion was categorized as malignant if malignant cells A total of 2193 patients with pleural effusion were identified from
were demonstrated in pleural fluid or pleural biopsy. A diagnosis of our computerized database during the study period, of which pleural
probable malignant effusion was made in patients with a known fluid ADA was available in 2104 cases. Of these, there were 591 (28%)
primary malignancy and negative pleural fluid cytological findings, malignant effusions, 493 (23.4%) transudates, 380 (18%) parapneumo-
after ruling out benign causes of the effusion. TB pleuritis was diag- nic effusions, 221 (10.5%) TB, and 419 (20%) miscellaneous exudates
nosed if Ziehl–Neelsen stains or Lowenstein–Jensen cultures of (Table 1). Among the 221 patients with TB effusions, 13 (5.8%) were HIV
pleural fluid, sputum or pleural biopsy specimens were positive or positive. Of the 80 patients with definite TB pleuritis, a pleural biopsy
the latter showed granulomas in the parietal pleura. TB was con- exhibited caseating granulomas in 17, a pleural fluid culture was
sidered probable in patients with lymphocytic exudates and high ADA positive in 40, and a sputum specimen was positive in 37.
levels (N40 U/L) in the pleural fluid, which cleared in response to The median (quartiles) age was significantly lower in patients with
antituberculous therapy. Neither probable malignant nor probable TB TB (31, 24.5 to 44.5 years) than in those with transudates (77, 70 to
subgroups were considered for most calculations because of the 83 years), malignancies (70, 59 to 78 years), parapneumonics (57, 41
different reference diagnostic standard applied to these participants. to 73 years) and miscellaneous exudates (69, 53 to 77 years; all
Parapneumonic effusion was defined as any exudative effusion p b 0.001).
associated with bacterial pneumonia, lung abscess or bronchiectasis.
Uncomplicated parapneumonic effusions described those effusions 3.2. Pleural fluid ADA in the etiological groups
that were successfully resolved with antibiotics alone. Complicated
parapneumonic effusions referred to those non-purulent-appearing Median ADA concentrations for each pleural effusion group are
effusions that required at least drainage and, eventually, further described in Table 2 and represented in Fig. 1. Patients with TB had
interventions whereas empyema described pus within the pleural higher median ADA levels (63.8 U/L, 53.3 to 78.2 U/L) than patients
space. The causes of other pleural effusions were determined by well- with non-TB effusions (16.3 U/L, 9.9 to 25.1 U/L; p b 0.001). However,
established clinical criteria. Finally, the terms transudate and exudate when parapneumonic effusions were categorized as non-complicated,
were based on the cause of the effusion (e.g., transudates encompass complicated and empyemas, the latter exhibited the highest ADA
those effusions that were clearly due to heart failure, cirrhosis or concentrations along with the TB group (Table 3). Notably, none of the
nephrosis). patients with pleural effusions due to TB had an ADA test exceeding
250 U/L, a feature that was only present in 19 empyemas and 3
2.3. Pleural fluid measurements
Table 1
Pleural fluid total ADA levels were determined immediately after Etiology of pleural effusions.
thoracentesis by using an automated ultraviolet kinetic assay (Roche
Causes N (%)
Diagnostics, Barcelona, Spain) on different discrete analyzers (Hitachi
717 and 911 or Hitachi Modular DP, Roche Diagnostics, Mannheim, Malignancy 591 (28)
Lung 222
Germany), according to the manufacturer's instructions. Other bio-
Breast 104
chemical measurements were performed using standard methodol- Unknown primary 54
ogies. White blood cells were manually counted in a Thoma chamber, Lymphoma 45
whereas differential cell counts were performed on pleural fluid Ovarian 44
smears after Wright's staining by counting, whenever possible, 200 Gastrointestinal 33
Mesothelioma 14
cells.
Others 75
Heart failure 390 (18.5)
2.4. Statistical analysis Parapneumonics 380 (18)
Uncomplicated 131
Complicated 133
Continuous and categorical data are presented as medians (quartiles)
Empyema 116
and percentages, respectively. Between-group comparisons of quanti- Tuberculosis 221 (10.5)
tative variables were performed with the nonparametric Kruskal–Wallis Definite 80
test. When significant differences were indicated by the Kruskal–Wallis Probable 141
test, a Mann Whitney U test with the Bonferroni correction was used to Non-cardiac transudates 103 (5)
Hepatic hydrothorax 60
determine which pair-wise differences were significant. The operating
Others 43
characteristics of pleural fluid ADA for identifying TB among patients Miscellaneous exudates 419 (20)
with lymphocytic exudates were assessed in terms of sensitivity, Idiopathic 74
specificity and likelihood ratios (LR). For these calculations only patients Abdominal surgery 71
Pericardial diseases 69
with definite TB or malignant effusions were considered. The diagnostic
Trauma 57
cutoff for ADA was optimized using receiver operating characteristic Pulmonary embolism 35
curves and selecting those values that provided the greatest sum of Cardiac surgery 23
sensitivity and specificity. Predictive values of ADA test according to the Connective tissue diseases 13
hypothetical prevalence of TB among exudative effusions were Others 77
Total 2104 (100)
calculated using the Bayes' formula [10]. The Spearman's correlation
J.M. Porcel et al. / European Journal of Internal Medicine 21 (2010) 419–423 421
Table 2 Table 3
Pleural fluid ADA levels in the study population. Pleural fluid ADA levels in parapneumonic and tuberculous effusions.
Etiology of pleural effusion No. of patients Pleural fluid ADA (U/L)a Pleural effusion categorization No. of patients Pleural fluid ADA (U/L)a
After excluding probable malignant effusions, there were 47 (9.5%) 3.6. Predictive values of the pleural ADA test related to the prevalence of
non-TB lymphocytic exudates that exceeded the cutoff established for tuberculosis
pleural TB, which included 27 malignant, 15 parapneumonic and 5
miscellaneous effusions (3 post-traumatic, 1 rheumatoid pleuritis and 1 The positive predictive value of the ADA test decreased in parallel
idiopathic). The primary tumors associated with false-positive ADA with the hypothetical prevalence of pleural TB among exudative
results were as follows: lymphoma (12 patients), lung (5 patients), effusions, whereas the negative predictive value actually remained
mesothelioma (4 patients), unknown primary (3 patients) and other high (Fig. 2). In our series, where TB represented about 14% of all
hematologic malignancies (2 lymphatic leukemias, 1 multiple myeloma). exudative effusions, the respective positive and negative predictive
Fig. 1. Box-plots showing the minimum, 25th, 50th, and 75th percentiles, and the maximum pleural fluid ADA levels (log scale) in different diagnostic groups. Footnote: Outliers
(open circles) and extreme values (stars) are plotted separately.
422 J.M. Porcel et al. / European Journal of Internal Medicine 21 (2010) 419–423
Table 5
Meta-analyses on the diagnostic accuracy of pleural fluid ADA in tuberculous pleural
effusions.
effusions. This fact lessens one of the study limitations, namely, its [3] Segura RM, Pascual C, Ocaña I, Martinez-Vázquez JM, Ribera E, Ruiz I, et al.
Adenosine deaminase in body fluids: a useful diagnostic tool in tuberculosis. Clin
retrospective design. Indeed, we only omitted 4% of all patients ad- Biochem 1989;22:141–8.
mitted with pleural effusion at our institution because ADA infor- [4] Frye MD, Sahn SA. Tuberculous pleural effusions in non-HIV infected patients;
mation was not available. Another limitation was the small number of 2009. Up to Date 17.3.
[5] Goto M, Noguchi Y, Koya ma H, Hira K, Shimbo T, Fukui T. Diagnostic value of
patients with confirmed TB who presented with a neutrophilic pleural adenosine deaminase in tuberculous pleural effusion: a meta-analysis. Ann Clin
fluid. A future prospective study comparing ADA and its isoenzymes Biochem 2003;40:374–81.
between neutrophil- and lymphocyte-rich TB effusions would be of [6] Greco S, Girardi E, Masciangelo R, Capoccetta GB, Saltini C. Adenosine deaminase
and interferon gamma measurements for the diagnosis of tuberculous pleurisy: a
interest. meta-analysis. Int J Tuberc Lung Dis 2003;7:777–86.
In conclusion, the use of pleural ADA levels provides a rapid and [7] Morisson P, Neves DD. Evaluation of adenosine deaminase in the diagnosis of
accurate means of detecting TB pleurisy, thereby expediting the initial pleural tuberculosis: a Brazilian meta-analysis. J Bras Pneumol 2008;34:217–24.
[8] Liang QL, Shi HZ, Wang K, Qin SM, Qin XJ. Diagnostic accuracy of adenosine
decision-making process. We advocate a strategy of using ADA as a
deaminase in tuberculous pleurisy: a meta-analysis. Respir Med 2008;102:744–54.
routine screening test in all patients presenting with an exudative [9] Porcel JM. Tuberculous pleural effusion. Lung 2009;187:263–70.
pleural effusion of uncertain origin, regardless of the prevalence of TB [10] Eisenberg MJ. Accuracy and predictive values in clinical decision-making. Cleve
in the population. Clin J Med 1995;62:311–6.
[11] Valdés L, San José E, Alvarez D, Sarandeses A, Pose A, Chomón B, et al. Diagnosis of
tuberculous pleurisy using the biologic parameters adenosine deaminase, lysozyme,
5. Learning points and interferon gamma. Chest 1993;103:458–65.
[12] Burgess LJ, Maritz FJ, Le Roux I, Taljaard JJ. Use of adenosine deaminase as a
diagnostic tool for tuberculous pleurisy. Thorax 1995;50:672–4.
• Elevated concentrations of pleural ADA (N35U/L) are a hallmark of [13] Villena V, Navarro-Gonzálvez JA, García-Benayas C, Manzanos JA, Echave J, López-
TB effusions, whether lymphocytes or neutrophils predominate in Encuentra A, et al. Rapid automated determination of adenosine deaminase and
the pleural fluid. lysozyme for differentiating tuberculous and nontuberculous pleural effusions.
Clin Chem 1996;42:218–21.
• The diagnostic work-up of exudative effusions should include a [14] Valdés L, Alvarez D, San José E, Penela P, Valle JM, García-Pazos JM, et al.
pleural fluid ADA measurement because, even in low TB prevalence Tuberculous pleurisy: a study of 254 patients. Arch Intern Med 1998;158:2017–21.
areas, an ADA value b35 U/L argues strongly against this disease. [15] Diacon AH, Van de Wal BW, Wyser C, Smedema JP, Bezuidenhout J, Bolliger CT,
et al. Diagnostic tools in tuberculous pleurisy: a direct comparative study. Eur
• An ADA activity in pleural fluid N250 U/L is highly suggestive of Respir J 2003;22:589–91.
empyema or lymphoma rather than TB. [16] Lin MT, Wang JY, Yu CJ, Lee LN, Yang PC, TAMI Group. Mycobacterium tuberculosis
and polymorphonuclear pleural effusion: incidence and clinical pointers. Respir
Med 2009;103:820–6.
References
[17] Zemlin AE, Burgess LJ, Carstens ME. The diagnostic utility of adenosine deaminase
isoenzymes in tuberculous pleural effusions. Int J Tuberc Lung Dis 2009;13:214–20.
[1] Piras MA, Gakis C, Budroni M, Andreoni G. Adenosine deaminase activity in pleural
[18] Antony VB, Sahn SA, Antony AC, Repine JE. Bacillus Calmette–Guérin-stimulated
effusions: an aid to differential diagnosis. Br Med J 1978;2:1751–2.
neutrophils release chemotaxins for monocytes in rabbit pleural spaces and in
[2] Valdés L, Pose A, San José E, Martinez-Vázquez JM. Tuberculous pleural effusions.
vitro. J Clin Invest 1985;76:1514–21.
Eur J Intern Med 2003;14:77–88.