Chest ACCP Member Benefits
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Guest Access | Sign In via User Name/Password
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (12)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Porcel, J. M.
Right arrow Articles by Esquerda, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Porcel, J. M.
Right arrow Articles by Esquerda, A.
(Chest. 2004;125:160-164.)
© 2004 American College of Chest Physicians

Tumor Necrosis Factor-{alpha} in Pleural Fluid*

A Marker of Complicated Parapneumonic Effusions

José Manuel Porcel, MD, FCCP; Manuel Vives, MD and Aureli Esquerda, PharmD

* From the Departments of Internal Medicine (Dr. Porcel) and Clinical Laboratory (Dr. Esquerda), University Hospital Arnau de Vilanova, Lleida; and Division of Internal Medicine (Dr. Vives), Clínica Recoletas, Albacete, Spain.

Correspondence to: José Manuel Porcel, MD, FCCP, Department of Internal Medicine, University Hospital Arnau de Vilanova, Alcalde Rovira Roure 80, 25198 Lleida, Spain; e-mail: jporcelp{at}medynet.com


   Abstract
TOP
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Study objectives: We sought to determine whether pleural fluid tumor necrosis factor (TNF)-{alpha} is a more accurate parameter to identify nonpurulent complicated parapneumonic effusion (CPPE) than the classical chemistries, namely pH, glucose, or lactate dehydrogenase (LDH).

Methods: We studied 80 consecutive patients with parapneumonic effusions (35 with uncomplicated parapneumonic effusion [UPPE], 23 with nonpurulent CPPE, and 22 with empyema). Concentrations of standard biochemical parameters together with TNF-{alpha} were measured in pleural fluid, the latter by using an immunoenzymometric assay.

Results: Pleural TNF-{alpha} was significantly higher in CPPE (133.0 pg/mL) and empyema (142.2 pg/mL) than in UPPE (39.1 pg/mL). A cut-off value of 80 pg/mL for pleural TNF-{alpha} resulted in a sensitivity, specificity, and area under receiver operating characteristic curve (AUC) of 78%, 89%, and 0.87, respectively, for the diagnosis of nonpurulent CPPE. A multivariate analysis selected both pleural TNF-{alpha} >= 80 pg/mL and LDH >= 1,000 U/L (sensitivity, 74%; AUC = 0.86), but excluded pleural glucose <= 60 mg/dL (sensitivity, 39%; AUC = 0.82) and pH <= 7.20 (sensitivity, 41%; AUC = 0.78), for identifying the need for drainage. The combined sensitivity of pleural fluid TNF-{alpha} and LDH was found to be 91%.

Conclusions: Pleural TNF-{alpha} may contribute to the identification of patients with nonpurulent CPPE with at least the same diagnostic accuracy, if not better, than the use of pH, glucose, or LDH.

Key Words: empyema • parapneumonic effusion • tumor necrosis factor • pH • pleural effusion


   Introduction
TOP
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Parapneumonic effusions (PPEs) occur in approximately 40% of patients who require hospital admission for bacterial pneumonia,1 and represent the second most common cause of exudates, exceeded only by malignant effusions.2 Although most PPEs will resolve with antibiotic treatment alone (uncomplicated parapneumonic effusion [UPPE]), it is recommended that all patients with more than a minimal PPE undergo a thoracentesis to determine the gross appearance as well as the biochemical and microbiological characteristics of the pleural fluid. When pus is present (empyema), pleural space drainage is mandatory. However, some but not all patients found to have a clear pleural fluid may eventually need drainage for resolution of pleural sepsis (complicated parapneumonic effusion [CPPE]). Identification of this latter group represents a major challenge to clinicians. Traditionally, determination of pleural pH, glucose, and lactate dehydrogenase (LDH) has been claimed to assist in the decision to drain PPE3 4 ; however, these biochemical parameters lack both enough sensitivity and reliable discriminating cutoff values.5

As a PPE progresses from the exudative to the fibropurulent stage, polymorphonuclear leukocytes and soluble factors (eg, complement activation products, proinflammatory cytokines)6 7 become progressively higher within the pleural space. Tumor necrosis factor (TNF)-{alpha} plays a key role in the inflammatory response. A unique study by Odeh et al8 showed that pleural fluid levels of this cytokine were significantly higher in CPPE than in UPPE, suggesting a potential role of TNF-{alpha} to discriminate between these two types of PPE. In the present investigation, we aimed to clarify this potential clinical interest in analyzing TNF-{alpha} in pleural fluid for the identification of nonpurulent CPPE.


   Materials and Methods
TOP
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Patients
All consecutive patients with a diagnosis of PPE who were admitted to the Department of Internal Medicine at the University Hospital Arnau de Vilanova (Lleida, Spain) from 1996 to 2002 entered the study. Typical or UPPE described those effusions that were successfully resolved with antibiotics alone. CPPE referred to those nonpurulent-appearing effusions that did not resolve without chest tube drainage, whereas empyema described frank pus within the pleural space, the end stage of a complicated effusion. The final decision concerning pleural space drainage was at the discretion of the attending physician.

Measurements
The samples obtained by thoracentesis were immediately analyzed for total cell counts, differential cell count, pH, glucose, protein, LDH, adenosine deaminase (ADA), cytology, and both aerobic and anaerobic bacterial cultures. Separated specimens were collected in tubes containing ethylenediamine tetra-acetic acid, centrifuged at 1,500g for 15 min at 4°C, and the supernatants were stored at – 70°C until TNF-{alpha} analysis. Only the results of the first thoracentesis were considered.

All biochemical measurements were performed on a selective, discrete, multichannel analyzer (Hitachi 717 and 917; Hitachi; Tokyo, Japan) using standard methodology. Specifically, pleural ADA activity and pH were assessed with an automated ultraviolet kinetic test (Roche Diagnostics; Barcelona, Spain) and through a blood-gas machine, respectively. Cellular counting was performed in a Thoma chamber (Weber Scientific International; Middlesex, UK). Human TNF-{alpha} in pleural fluid was measured by an immunoenzymometric assay (Biosource Europe, S.A; Nivelles, Belgium), according to the instructions of the manufacturer. The investigator running the TNF-{alpha} assay was blinded of clinical information, and similarly clinicians who made the decision to institute pleural drainage did not know the results of the cytokine test.

Statistical Analysis
Continuous data are reported as medians (quartiles). The {chi}2 and Kruskal-Wallis tests were used to compare groups for qualitative and quantitative variables, respectively. To compare performances of biochemical pleural fluid analytes, receiver operating characteristic curves were constructed. The cutoff value for TNF-{alpha} with the highest diagnostic accuracy was selected in order to discriminate UPPE from CPPE. To identify predictors of CPPE, we used a backward conditional stepwise logistic regression analysis on dichotomized biochemical variables of pleural fluid. Finally, a bivariate correlation analysis (Pearson) was used to test for significant linear relationship between polymorphonuclear leukocyte counts and levels of TNF-{alpha} in pleural fluid. A two-tailed p value <= 0.05 was considered significant. Data were analyzed by using the SPSS statistical package (version 10.0; SPSS; Chicago, IL).


   Results
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
The study population comprised 80 patients (60 men and 20 women), with a median age of 62 years (quartiles, 42 to 75) years. Of these, 35 patients had UPPE, 23 patients had CPPE, and 22 patients had empyema. Overall, pleural fluid culture findings were positive in 23 of 74 patients (31%), aerobic Gram-positive bacteria being isolated in 13 cases. Categorization of PPE among the patients with culture-positive fluids was as follows: UPPE (n = 2), CPPE (n = 7), and empyema (n = 14).

Table 1 shows clinical and pleural fluid characteristics for the three different types of PPE. As expected, patients with CPPE presented more frequently with large effusions and positive culture findings, and their pleural fluids had lower pH and glucose contents, and higher LDH concentrations than those observed in patients with UPPE. Empyema fluids exhibited particularly high values of leukocytes, ADA, and pleural fluid-to-serum (P/S) LDH ratios when compared to the remainder groups.


View this table:
[in this window]
[in a new window]

  		Table 1. Characteristics of the Study Population*

 
Median TNF-{alpha} levels in pleural fluid were significantly higher in CPPE (133.0 pg/mL) and empyema (142.2 ng/mL) than in UPPE (39.1 ng/mL, p < 0.001), but no statistical difference was found between the first two groups (Fig 1 ). Various cutoff levels of TNF-{alpha} were tested for the identification of CPPE, and 80 pg/mL yielded the best results (Table 2 ). This corresponded to a sensitivity, specificity, and accuracy of 78%, 89%, and 84%, respectively.



View larger version (16K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Distribution of pleural TNF-{alpha} in the three different classes of parapneumonic effusions.

 

View this table:
[in this window]
[in a new window]

  		Table 2. Operating Characteristics of Pleural TNF-{alpha} at Different Cutoff Values for Diagnosing Nonpurulent CPPEs*

 
We next compared the relative diagnostic accuracies of pleural fluid pH, glucose, LDH, and TNF-{alpha} to categorize a PPE as complicated (Table 3 ). Decision thresholds for pH, glucose, and LDH were those widely accepted by consensus statements and expert opinions.3 4 We noted that pleural fluid TNF-{alpha} had the highest diagnostic accuracy as measured by the area under the receiver operating characteristic curve (AUC), 0.87, although overlapping confidence intervals (CIs) indicated that no single test could be clearly identified as being superior to others. Notably, pleural TNF-{alpha} offered a sensitivity that doubled that of pleural pH for predicting a CPPE. Only one patient (3%) with pleural pH <= 7.20 and two patients (6%) with pleural glucose levels <= 60 mg/dL exhibited concentrations of TNF-{alpha} < 80 pg/mL.


View this table:
[in this window]
[in a new window]

  		Table 3. Measures of Diagnostic Accuracy for Tests That Identify a Nonpurulent CPPE*

 
Logistic regression analysis from the biochemical pleural fluid variables showed that TNF-{alpha} at the prespecified threshold value was the most useful diagnostic indicator for CPPE (adjusted likelihood ratio [LR]+, 7.6; 95% CI, 3.1 to 10.8; adjusted LR-, 0.06; 95% CI, 0.01 to 0.27). The second best test was LDH (adjusted LR+, 3.6; 95% CI, 1.0 to 8.2; adjusted LR-, 0.23; 95% CI, 0.05 to 1.04). Combination of pleural TNF-{alpha} and LDH in an "or" rule, wherein the pleural space would be drained if any one of the two results are "positive," yielded measures for sensitivity (91%; 95% CI, 78 to 100%), specificity (77%; 95% CI, 62 to 92%), and accuracy (83%; 95% CI, 72 to 93).

A significant but weak correlation was found between levels of TNF-{alpha} and percentage of neutrophils in pleural fluid (r = 0.26, p = 0.02) when considering the whole population of PPE. However, subgroup analysis showed no correlation between neither percentages nor absolute neutrophil counts and pleural TNF-{alpha}.


   Discussion
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
The present study suggests that pleural TNF-{alpha} may be a good biochemical marker of inflammation in patients with PPE. Elevated levels of pleural TNF-{alpha} identified more reliably the subgroup of patients with nonpurulent-appearing PPE who required invasive management with tube thoracostomy than traditional fluid chemistries.

Determining the need to drain a PPE can be a complex decision based not only on the microbiological and chemical fluid characteristics, but also on host factors and radiographic data.9 This decision should generally favor drainage, because the severe morbidity associated with progression to an empyema justifies the placement of a few extra chest tubes. Undoubtedly, results of thoracentesis provide essential information for patient management. Thus, detection of gross pus means that patient has an empyema, which is an absolute indication for drainage. Pleural drainage will also most likely be needed for patients with positive pleural fluid culture findings, yet up to 90%, 70%, and 25% of UPPEs, CPPEs, and empyemas, respectively, have negative bacteriologic fluid examination findings.10 In addition, it is our experience that approximately 10% of patients with positive pleural culture findings respond to antibiotics alone.10

In clinical practice, the biochemical characteristics of pleural fluid are the cornerstone to discriminate which patients have CPPE. A guideline4 for the treatment of PPE published by the American College of Chest Physicians stated that the presence of a low pleural fluid pH (< 7.20) or glucose (< 60 mg/dL) are usually associated with a poor prognosis, thus needing more aggressive therapeutic maneuvers. Light3 also included pleural fluid LDH in his classification scheme for PPE in that levels > 1,000 U/L define a borderline CPPE. A meta-analysis5 of these three biochemical indexes revealed that pleural fluid pH was a little better for identifying nonpurulent CPPE (AUC = 0.89) than was the pleural fluid glucose or LDH (AUC = 0.71), although multiple design problems in the primary studies caution the interpretation of data. We found considerable overlap between the AUC values for pH (AUC = 0.78), glucose (AUC = 0.82), and LDH (AUC = 0.86), but there was a trend toward a better accuracy of the latter, mainly due to its superior sensitivity.

Even though the above-mentioned decision thresholds for pleural pH or glucose are commonly used to screen for CPPE, the evidence supporting this practice comes from consensus of clinical opinion.4 In fact, the diagnostic yield of these parameters in the present study was disappointing in that 55% (12 of 22 patients) with PPE who ultimately underwent drainage exhibited pleural levels of pH and glucose > 7.20 and > 60 mg/dL, respectively. Unfortunately, the inadequate sensitivity of the preferred tests (ie, pH or glucose) contrasts with the critical decision to initiate prompt drainage when indicated. For this reason, alternative biochemical markers, such as pleural SC5b-9,6 myeloperoxidase,11 or TNF-{alpha},8 has been proposed to aid decision making for drainage. In a pilot study,6 we reported sensitivity (100%), specificity (75%), and AUC (0.89) of the terminal complement complex SC5b-9 in pleural fluid for differentiating nonpurulent CPPE from UPPE when a cutoff point of 1,500 µg/L was used. Likewise, in the study of Alegre et al,11 pleural myeloperoxidase at a cut point limit of 3,000 µg/L was found to be the marker that best discriminated between these two types of PPE (sensitivity, 87.5%; specificity, 85.1%; AUC = 0.91).

TNF-{alpha}, among other cytokines, initiates a wide spectrum of biological activities that help activate the host’s response to infection, and thus may play pathophysiologic roles in PPE. Odeh et al8 measured serum and pleural fluid TNF-{alpha} in 13 patients with UPPE and 8 patients with nonempyemic CPPE. At an optimal cutoff point of 10.7 pg/mL, discriminative properties for pleural TNF-{alpha} were as follows: sensitivity, 87.5%; specificity, 92.3%; and accuracy, 90.5% (AUC was not calculated). These statistics were all 100% for the P/S TNF-{alpha} ratio set at 3.0. By using a cutoff value of 80 pg/mL, we observed that the sensitivity of TNF-{alpha} (78%) to identify CPPE was double that of either pH (41%) or glucose (39%), whereas the specificity of the former (89%) was somewhat lower than those of the latter (94% and 97%, respectively). The combination of TNF-{alpha} and LDH increased the sensitivity to 91%, at the expense of a diminished specificity (77%). Nevertheless, some degree of misclassification cost, ie, to place some unnecessary chest tubes, is acceptable from the clinical standpoint. Finally, unlike other researchers12 who found a significant correlation between levels of pleural TNF-{alpha} and neutrophil counts in both UPPE (r = 0.44, p = 0.04) and CPPE (r = 0.57, p = 0.03), we did not observe such an association. This suggests that factors other than neutrophils account for the TNF-{alpha} concentrations in pleural fluid.

Our study has several limitations. First, neither our study nor any other clinical investigation to determine the accuracy of tests for discriminate UPPE from CPPE can be designed to circumvent the problem of the lack of an errorless "gold standard" for CPPE. Definition of CPPE is ultimately based on the subjective clinician judgment to drain the pleural space, which may be an imperfect reference standard. Thus, physicians may decide to insert chest tubes in patients whose pleural effusions might be resolved with no invasive procedures, resulting in a misclassification of the PPE. A second important problem with this and all studies dealing with tests to better categorize PPE is that sample sizes are too small to allow meaningful comparisons of the relative diagnostic accuracies of the tests.

In summary, elevated levels of pleural TNF-{alpha} could be considered as an indicator of poor prognosis in patients with PPE, and measurement of this cytokine would help refine categorization of PPE in cases of uncertainty. However, until further confirming studies, TNF-{alpha} analysis cannot yet be judged for an eventual future role in the clinical workup of PPE.


   Footnotes
 
Abbreviations: ADA = adenosine deaminase; AUC = area under the receiver operating characteristic curve; CI = confidence interval; CPPE = complicated parapneumonic effusion; LDH = lactate dehydrogenase; LR = likelihood ratio; PPE = parapneumonic effusion; P/S = pleural fluid to serum; TNF = tumor necrosis factor; UPPE = uncomplicated parapneumonic effusion

Received for publication May 6, 2003. Accepted for publication August 21, 2003.


   References
TOP
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Light, RW, Girard, WM, Jenkinson, SG, et al (1980) Parapneumonic effusions. Am J Med 69,507-512[CrossRef][ISI][Medline]
  2. Hernández, L, Romero, S Derrame pleural maligno. Porcel, JM eds. Enfermedades de la pleura 2002,83-94 Edicions de la Universitat de Lleida. Lleida, Spain:
  3. Light, RW A new classification of parapneumonic effusions and empyema. Chest 1995;108,299-301[Free Full Text]
  4. Colice, GL, Curtis, A, Deslauriers, J, et al Medical and surgical treatment of parapneumonic effusions: an evidence-based guideline. Chest 2000;118,1158-1171[Abstract/Free Full Text]
  5. Heffner, JE, Brown, LK, Barbieri, C, et al Pleural fluid chemical analysis in parapneumonic effusions. Am J Respir Crit Care Med 1995;151,1700-1708[Abstract]
  6. Vives, M, Porcel, JM, Gázquez, I, et al Pleural SC5b-9: a test for identifying complicated parapneumonic effusions. Respiration 2000;67,433-438[Medline]
  7. Marie, C, Losser, MR, Fitting, C, et al Cytokines and soluble cytokine receptors in pleural effusions from septic and nonseptic patients. Am J Respir Crit Care Med 1997;156,1515-1522[Abstract/Free Full Text]
  8. Odeh, M, Sabo, E, Oliven, A, et al Role of tumor necrosis factor-{alpha} in the differential diagnosis of parapneumonic effusion. Int J Infect Dis 1999;4,38-41
  9. Heffner, JE Indications for draining a parapneumonic effusion: an evidence-based approach. Semin Respir Infect 1999;14,48-58[Medline]
  10. Porcel JM, Vázquez P, Vives M, et al. Pleural space infections: microbiologic and fluid characteristics in 84 patients. Internet Journal of Pulmonary Medicine 2003; 3(1). Available at: http://www.ispub.com/journals/ijpm/archives
  11. Alegre, J, Jufresa, J, Segura, R, et al Pleural-fluid myeloperoxidase in complicated and noncomplicated parapneumonic pleural effusions. Eur Respir J 2002;19,320-325[Abstract/Free Full Text]
  12. Odeh, M, Sabo, E, Srugo, I, et al Correlation between polymorphonuclear leukocyte counts and levels of tumor necrosis factor-{alpha} in pleural fluid of patients with parapneumonic effusion. Lung 2002;180,265-271[Medline]



This article has been cited by other articles:


Home page
 ChestHome page
G. E. Utine, U. Ozcelik, E. Yalcin, D. Dogru, N. Kiper, A. Aslan, and G. Kanra
Childhood Parapneumonic Effusions: Biochemical and Inflammatory Markers
Chest, September 1, 2005; 128(3): 1436 - 1441.
[Abstract] [Full Text] [PDF]

Home page
 ChestHome page
M. Odeh, A. Oliven, J. M. Porcel, M. Vives, and A. Esquerda
Tumor Necrosis Factor-{alpha} in Parapneumonic Effusion
Chest, May 1, 2005; 127(5): 1868 - 1869.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF) Free
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Article Archive
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (12)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Porcel, J. M.
Right arrow Articles by Esquerda, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Porcel, J. M.
Right arrow Articles by Esquerda, A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS