Q J Med 2000; 93: 819-824
© 2000 Association of Physicians
Tumour necrosis factor alpha in the diagnostic assessment of pleural effusion
From the Departments of Internal Medicine B 1 Pathology, and 2 Microbiology, Bnai Zion Medical Center, and Technion Faculty of Medicine, Israel Institute of Technology, Haifa, Israel
Received 5 May 2000 and in revised form 23 September 2000
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
We investigated the role of tumour necrosis factor-
(TNF) in the evaluation of pleural effusion aetiology. Using a commercially-available ELISA kit, concentrations of TNF were measured in the serum and pleural fluid of patients with malignant effusions (n=19), uncomplicated parapneumonic effusions (n=13), and exudative (n=13) and transudative (n=13) effusions due to congestive heart failure (CHFex and CHFtr, respectively). Serum TNF did not differ significantly between the four groups (p>0.05). In the group with malignancy, pleural fluid TNF was significantly higher than in the other groups (p<0.001), which were not significant different from each other (p>0.05). However, a considerable overlap between all four groups was found. Pleural fluid TNF was significantly higher than serum TNF in the malignant and the uncomplicated parapneumonic groups (p<0.001), and there was a significant positive correlation between serum TNF and pleural fluid TNF in the group with uncomplicated parapneumonic effusion (r=0.7, p<0.005), in the group with CHFex (r=0.54, p<0.01), and in the group with CHFtr (r=0.8, p<0.005), but not in the group with malignancy. Pleural fluid TNF:serum TNF (TNF ratio) was significantly higher in the malignancy group than in the other groups (p<0.001); no significant difference was found between the other three groups (p>0.05). At an optimal cut-off point of 2.0 for TNF ratio, determined by ROC analysis for discrimination between malignant and non-malignant groups, sensitivity was 84%, specificity 90%, and total accuracy 88% (p<0.0001). TNF ratio might be helpful in the diagnostic assessment of exudative pleural effusion. |
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Distinguishing an exudate from a transudate is the initial step in determining the cause of a pleural effusion. Transudative effusions develop when there is a change in systemic factors such as an increase in capillary hydrostatic pressure or a decrease in colloid osmotic pressure with no change in the pleural surface. Transudative effusion is caused by a limited number of diseases, namely, congestive heart failure (CHF), hypoalbuminaemia and urinothorax. If the fluid is transudative, no further diagnostic procedures are necessary, and therapy is directed to the underlying disease, which can be diagnosed by simple means.1 Exudative effusions result from pleural inflammation, infection, injury or lymphatic obstruction, such as occurs in pneumonia with effusion, malignancy, tuberculosis, pulmonary infarction and rheumatoid arthritis.1 Exudative effusion may also present in CHF after treatment with diuretics, due to significant changes in the pleural fluid chemistry.2 An exudative process always requires a more extensive and invasive diagnostic evaluation, which may include pleural biopsy and thoracoscopy.1 Following a complete diagnostic evaluation, including closed pleural biopsy, 20% of effusions remain undiagnosed,3,4 and the clinical management of these cases is controversial.
Tumour necrosis factor- (TNF) is a small polypeptide with pleiotropic effects on biological and immunological processes. It is synthesized by various activated cells, particularly the monocyte/macrophage system, and a wide variety of infectious or inflammatory stimuli are capable of triggering TNF biosynthesis.5–7 High levels of this cytokine in pleural fluid have been observed in several diseases.8–11 However, the diagnostic usefulness of pleural fluid TNF concentration in these diseases is limited because of considerable overlap among different diseases.8,10 In addition, serum levels of TNF have not been useful in the differential diagnosis of the underlying disease causing pleural effusion.10–12 Nevertheless, the utility of the pleural fluid TNF:serum TNF ratio (TNF ratio) has not previously been evaluated in the diagnostic assessment of pleural effusion.
The aim of this prospective study was to investigate the diagnostic usefulness of TNF in identifying the cause of exudative pleural effusion in cases of diverse aetiology including malignancy, pneumonia with uncomplicated parapneumonic effusion, and CHF.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Patients
The study population consisted of 58 patients with pleural effusion. According to the final diagnosis and the type of the pleural effusion, the patients were divided into four groups: (i) 19 patients, nine males and 10 females, aged 51–90 years, with malignant exudative pleural effusion (4 adenocarcinoma of lung, 3 adenocarcinoma of breast, 2 adenocarcinoma of colon, 6 due to, respectively, adenocarcinoma of stomach, lymphoma, adenocarcinoma of pancreas, squamous cell carcinoma of lung, sarcoma of kidney, and mesothelioma of pleura, and 4 adenocarcinoma of unknown origin); (ii) 13 patients, seven males and six females, aged 56–91 years, with exudative uncomplicated parapneumonic effusion; (iii) 13 patients, seven males and six females, aged 56–91 years, with CHF and exudative pleural effusion (CHFex); (iv) 13 patients, seven males and six females aged 59–94 years, with CHF and transudative pleural effusion (CHFtr).
Effusions were considered malignant if malignant cells were demonstrated at cytological examination or in a biopsy specimen in the absence of other diseases causing pleural effusion. A pleural effusion was considered to be uncomplicated parapneumonic when there was an acute febrile illness with purulent sputum, pulmonary infiltrate and responsiveness to antibiotic treatment, in the absence of malignancy or other diseases causing exudative or transudative pleural effusion, and with no direct or indirect evidence of bacterial invasion of the effusion. In all these patients the pleural effusion completely resolved by antibiotic treatment, and did not recur during the follow-up period. CHF was diagnosed when all of the following criteria were met: (i) cardiomegaly; (ii) evidence of cardiac dysfunction (clinical, echocardiographic and/or by MUGA scan); (iii) radiological evidence of congested lungs and/or peripheral oedema; and (iv) response to treatment of the CHF. In all cases, there was an absence of pulmonary embolism, purulent sputum, malignancy, and pulmonary infiltrates. In all these patients, pleural effusion completely resolved or diminished significantly by treatment of the CHF. In the groups of non-malignant pleural effusion, malignancy was excluded by performing repeat cytological analysis, and closed pleural biopsy if pleural effusion persisted, in addition to the routine diagnostic procedures for evaluating occult malignancy for all patients with exudative pleural effusion. All patients in the non-malignant groups were monitored for at least 12 months to exclude malignancy.
Techniques
Pleural fluid samples were obtained by intercostal needle aspiration. Pleural fluid drainage and blood sampling were performed for the routine diagnostic procedures and/ or for relief of dyspnoea, but not particularly for TNF assay. Samples of pleural fluid and peripheral venous blood were obtained within 15 min of each other for LDH, protein, glucose, pH, WBC, and TNF analysis. All pleural fluids were cultured and stained for the presence of bacteria, and analysed cytologically for the presence of tumour cells. For the laboratory classification of pleural fluids as exudates or transudates, protein and LDH were interpreted according to the criteria of Light et al.13 Exudative pleural effusions met at least one of the following criteria, whereas transudative pleural effusions met none: (i) Pleural fluid protein/serum protein >0.5; (ii) pleural fluid LDH/serum LDH >0.6; and (iii) pleural fluid LDH more than two-thirds normal upper limit for serum. Pleural biopsies were taken under local anesthesia with the Abrams' needle. The study met the criteria of the Ethics Committee of our institution.
Measurement of TNF concentrations
Samples of pleural fluid and peripheral venous blood (after clotting) were centrifuged at 2500 rpm, at 4 °C centrifuge temperature for 10 min. The supernatants were stored at -70 °C until TNF assay, which was performed within 4 months of sample storage. TNF concentrations were measured using commercially available (Quantikine HS, R&D Systems) solid-phase, high-sensitivity enzyme-linked immunosorbent assay (ELISA) according to the manufacturers recommendation. All pleural fluid and serum samples were run in duplicate, and coded so that investigators running the assays were blinded as to their source. Results were expressed in pg/ml. The lower detection limit was 0.18 pg/ml, and the intra-assay and the inter-assay coefficients of variation were 5.8% and 4.6%, respectively.
Statistics
All values were expressed as means±SD, and 95% CIs were calculated. Statistical comparison between multiple continuous variables was performed using one-way analysis of variance, followed by the Bonferroni post hoc test. Correlation between pleural fluid TNF levels and serum TNF levels in each group was evaluated by a simple linear regression method, and Pearson's correlation coefficient (r) was calculated. Receiver-operating characteristic (ROC) analysis was used to detect the optimal cut-off points (i.e. those with the highest total accuracy) for separating malignant from non-malignant effusions. Associations between cut-off point values and categorical outcome variables were evaluated using the 
2 test. A multivariate logistic regression analysis was used to detect significant independent discriminators between malignant and nonmalignant groups. Two-tailed p values of 0.05 or less were considered significant.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The mean concentration of serum TNF for each of the groups is presented in Table 1
. The groups did not show significant differences (p>0.05) (Figure 1). The mean concentration of pleural fluid TNF in each of the groups is also presented in Table 1. Pleural fluid TNF levels were significantly higher in the group with malignancy than in the other three groups (p<0.001); no significant difference was found between these three groups (p>0.05) (Figure 1). Pleural fluid TNF levels were significantly higher than serum TNF levels in the groups with malignancy and uncomplicated parapneumonic effusion (p<0.001), but not in both groups with CHF (p>0.05). A significant positive correlation was found between serum TNF levels and pleural fluid TNF levels in the group with uncomplicated parapneumonic effusion (r=0.7, p<0.005), in the group with CHFex (r=0.54, p<0.01) and in the group with CHFtr (r=0.8, p<0.005), but not in the group with malignancy. The TNF ratio for each of the groups is presented in Table 1. Levels of this ratio were significantly higher in the group with malignancy than in the other three groups (p<0.001); no significant difference for this ratio was found between the three groups (p>0.05) (Figure 2).
|
|
|
Since both pleural fluid TNF and TNF ratio were significantly higher in the malignant group than in the other three groups, we used ROC analysis to detect the optimal cut-off point of both markers for better discrimination between malignant and non-malignant groups. At an optimal cut-off point of 7.9 pg/ml for pleural fluid TNF, 16 patients in the malignant group were above this cut-off, and three were below it. In the non-malignant groups, 11 patients were above this cut-off, and 28 were below it. The sensitivity of this cut-off point for pleural fluid TNF was 84%, the specificity 72%, and total accuracy 76%. The area under the curve was 0.82 (Figure 3
) (p<0.001). At an optimal cut-off point of 2.0 for TNF ratio, 16 patients in the malignant group were above the cut-off, and only three below it. In the non-malignant groups, only four patients were above it (Figure 2
). The sensitivity of this cutoff point was 84%, specificity 90%, total accuracy 88%, and the area under the curve was 0.91 (Figure 4) (p<0.0001). The relationship between the independent variables, pleural fluid TNF and TNF ratio, and the dependent variables, the causes of the pleural effusion (malignant or non-malignant) was tested using the multivariate logistic regression analysis. TNF ratio was found to be the only significant independent discriminator between malignant and non-malignant groups (p<0.0001).
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
The main new observation in this study is that determination of TNF ratio may be helpful in the diagnostic assessment of exudative pleural effusions. Using an optimal cut-off point of 2.0 for this marker, which has not previously been evaluated in the diagnostic assessment of pleural effusion, only 3/19 malignant patients fell below this cutoff point, and only 4/39 non-malignant patients rose above it, with relatively high sensitivity, specificity and total accuracy (Figure 2
).
No significant difference was found between all groups studied regarding serum TNF (Table 1, Figure 1). This is in agreement with the results of previous studies.10,12 Pleural fluid TNF levels were significantly higher in the malignant group than in the non-malignant groups; no significant difference was found between the non-malignant groups. Since no significant difference for serum TNF was found between malignant and non-malignant groups, and also no significant correlation between pleural fluid TNF and serum TNF in the malignant group, the observation that pleural fluid TNF levels were significantly higher in the malignant group than those of the non-malignant groups may be explained, at least partly, by increased local production of TNF in the pleural cavity of the malignant patients. Indeed, macrophages, which are the main source of TNF production, predominate in malignant pleural effusion.14 In these effusions these cells are functionally active,15,16 and can secrete TNF spontaneously and by interaction with malignant cells.17–19 Furthermore, these cells, which have been recently shown to express the HLA-DR antigen, stimulate allogeneic T-lymphocyte proliferation,19 which may also participate in the local production of TNF. In addition, mesothelial cells may release TNF upon exposure to inflammatory processes.20 Since pleural carcinomatosis generally induces local inflammation, pleural mesothelial cells may also participate in the local production of TNF. No excess of macrophages is present in uncomplicated parapneumonic effusions and in effusions due to CHF. In these cases, pleural fluid TNF concentrations are mainly influenced by serum TNF concentrations. This may, at least partly, explain the significant positive correlation between pleural fluid TNF and serum TNF levels in the groups with uncomplicated parapneumonic effusion, CHFex and CHFtr.
Although pleural fluid TNF levels were significantly higher in the malignant group than in the non-malignant groups, this marker was not a good discriminator between the malignant and the non-malignant effusions because of the considerable overlap between all groups (Figure 1, Table 1). Studying pleural fluid TNF and TNF ratio together by multivariate logistic regression analysis, only TNF ratio was found to be a significant independent discriminator between malignant and non-malignant groups (p<0.0001).
Exudative pleural effusion is most commonly parapneumonic or malignant. It also may be, not uncommonly, due to CHF, where in some studies more than 30% of pleural effusions due to CHF were exudative.2,21 Determining the cause of an exudative pleural effusion is not always easy. While false-positive results of pleural fluid cytological studies almost never occur, even repeat negative results do not rule out malignancy as the cause of the effusion. Following a complete diagnostic evaluation including repeat cytological studies and closed pleural biopsy, about 20% of exudative effusions remain undiagnosed.3,4 In these patients, whose clinical management is controversial, and in whom invasive diagnostic techniques, which are not completely free of morbidity, are frequently required, TNF ratio may help in choosing a suitable diagnostic approach.
Patients with tuberculous pleuritis and collagen diseases were not included in our study because we had no such patients during the course of our study. Indeed, in our country, as in many other countries, pleural tuberculosis is rare, and exudative pleural effusions are usually due to pneumonia or malignancy. High levels of adenosine deaminase are present in pleural fluid due to tuberculous pleuritis, but not in pleural fluid due to malignancy, pneumonia, and CHF.1 This parameter may help in distinguishing tuberculous from non-tuberculous pleural effusion. Pleural effusion due to rheumatoid arthritis is not common, and is distinctive as regards the biochemical parameters of the fluid. It usually develops in patients with established rheumatoid arthritis and subcutaneous nodules. Furthermore, only patients with rheumatoid pleuritis have a titre of rheumatoid factor in the pleural fluid at or above 1:320, and equal to or greater than the serum titre.22 The diagnosis of lupus pleuritis is best made by demonstrating an elevated antinuclear antibody (ANA) level in the pleural fluid. Patients with lupus pleuritis have an ANA titre at or above 1:160 or a pleural fluid-to-serum ANA ratio at or above 1. Patients with pleural effusions of other aetiologies do not meet either of these criteria.23
In conclusion, TNF ratio might be helpful in the diagnostic assessment of exudative pleural effusion, particularly in cases where the aetiology of the effusion remains undiagnosed after an extensive evaluation. In these cases, TNF ratio may help in choosing a suitable further diagnostic approach. However, further prospective studies are needed in a larger population to confirm these findings.
|
Notes |
|---|
Address correspondence to Dr M. Odeh, PO Box 6477, Haifa 31063, Israel.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. Light RW. Diagnostic principles in pleural disease. Eur Respir J1997; 10:476–81.[Abstract]
2.
Chakko SC, Caldwell SH, Sforza PS. Treatment of congestive heart failure: Its effect on pleural fluid chemistry. Chest1989; 95:798–802.
3. Storey DD, Dines DE, Coles DT. Pleural effusion: a diagnostic dilemma. JAMA1976; 236:2183–6.[Abstract]
4. Hirsch A, Ruffie P, Nebut M, et al. Pleural effusion: laboratory test in 300 cases. Thorax 1979; 34:106–12.
5. Cerami A, Tracey KJ, Lowry SF, et al. Cachectin: A pluripotent hormone released during the host response to invasion. Recent Prog Horm Res1987; 43:99–112.
6. Beutler B, Cerami A. Cachetin: More than a tumor necrosis factor. N Engl J Med1987; 316:479–85.
7. Tracey KJ, Vlassara H, Cerami A. Cachectin: Tumour necrosis factor. Lancet1989; ii: 1122–6.
8.
Soderblom T, Nyberg P, Teppo A-M, et al. Pleural fluid interferon-
and tumour necrosis factor- in tuberculous and rheumatoid pleurisy. Eur Respir J1996; 9:1652–5.[Abstract]
9.
Barnes PF, Fong S-J, Brennan PJ, et al. Local production of tumor necrosis factor and IFN- in tuberculous pleuritis. J Immunol1990; 145:149–54.[Abstract]
10. Xirouchaki N, Bouros D, Alexandrakis M, et al. The role of cytokines in pleural effusion differentiation. Eur Respir J1994; 7 (Suppl. 18):471s.
11. Gursel G, Gokcora N, Elbeg S, et al. Tumor necrosis factor-alpha (TNF-alpha) in pleural fluids. Tuber Lung Dis1995; 76:370–1.[ISI][Medline]
12. Orphanidou D, Gaga M, Rasidakis A, et al. Tumour necrosis factor, interleukin-1 and adenosine deaminase in tuberculous pleural effusion. Respir Med1996; 90:95–8.[ISI][Medline]
13. Light RW, MacGregor MI, Luchsinger PC, et al. Pleural effusions: The diagnostic separation of transudates and exudates. Ann Intern Med1973; 77:507–13.
14. Lin C-C, Liu C-C, Lin C-Y. Changes in cell population and tumor necrosis factor, interleukin-6, and interleukin-8 in malignant pleural effusions after treatment with intrapleural tetracycline. Am Rev Respir Dis1993; 147:1503–6.[ISI][Medline]
15. Doers J, Baughman RP, Buchsbaum J, et al. Tumor necrosis factor production by pleural macrophages. Am Rev Respir Dis1989; 139:A360.
16. Baughman RP, Doers J, Lower EE, et al. Isolation of functionally active pleural macrophages. J Clin Immunol1989; 30:147–51.
17. Ishii Y, Uchiyama Y, Hasegawa S, et al. Detection of tumour necrosis factor/cachectin in pleural effusion of patients with lung cancer. Clin Exp Immunol1990; 80:350–3.[Medline]
18. Baughman RP, Lower EE. An inhibitor of tumor necrosis factor found in pleural effusions. J Lab Clin Med1991; 118:326–31.[Medline]
19.
Gjomarkaj M, Pace E, Melis M, et al. Mononuclear cells in exudative malignant pleural effusions. Characterization of pleural phagocytic cells. Chest1994; 106:1042–9.
20. Antony VB, Rothfuss KJ, Godbey SW, et al. Mechanisms of tetracycline-hydrochloride-induced pleurodesis. Tetracycline hydrochloride-stimulated mesothelial cells produce a growth-factor like activity for fibroblasts. Am Rev Respir Dis1992; 146:1009–13.[ISI][Medline]
21. Peterman TA, Speicher CA. Evaluating pleural effusion: a two-stage laboratory approach. JAMA1984; 252:1051–3.[ISI][Medline]
22. Halla JT, Schrohenlogher RE, Volanakis JE. Immune complexes and other laboratory features of pleural effusions. Ann Intern Med1980; 92:748–52.
23.
Good JT Jr, King TE, Antony VB, Sahn SA. Lupus pleuritis: Clinical features and pleural fluid characteristics with special reference to pleural fluid antinuclear antibodies. Chest1983; 84:714–18.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||