Printer Friendly

Quantitative analysis of pleural fluid cell-free DNA as a tool for the classification of pleural effusions.

Pleural effusions represent a common diagnostic challenge to clinicians. Etiologies include fluid retention states, pulmonary infection, malignant neoplasms, and other less common causes such as chylothorax and fistula formation. Effusions can be classified into exudative and transudative according to their pathophysiologic mechanisms and pleural fluid characteristics. Theoretically, transudative pleural effusions have a limited inflammatory or cellular element because the pathophysiologic mechanism is purely of abnormal fluid and osmotic dynamics, e.g., with fluid retention or hypoproteinemia. By contrast, infective and malignant causes will induce variable but substantial inflammatory and cellular responses in the pleural cavity and produce exudative effusions.

Light et al. (1) proposed as diagnostic criteria for exudative effusion that the pleural-fluid-to-serum total protein ratio be >0.5, the pleural-fluid-to-serum lactate dehydrogenase (LDH) ratio be >0.6, and the pleural fluid LDH activity be >200 U/L. In their study of 150 patients, the criteria produced a diagnostic sensitivity of 99% and specificity of 98% for exudative effusions. Other prospective studies, however, reported much lower diagnostic specificities, ranging from 65% to 86% (2-4). The high sensitivity and specificity reported in the original study by Light et al. (1) could be the result of their stringent diagnostic requirements, which led to the exclusion of 33 of the total 183 cases (5). The most commonly used modification to the criteria of Light et al. (1) is the use of pleural fluid LDH more than two-thirds of the upper limit of the reference interval for serum as a criterion to replace an absolute pleural fluid LDH >200 U/L. This modification of the criteria defined by Light et al. (1) will be used in the rest of this report for method comparison. All other modifications have involved the use of multiple markers, but the diagnostic accuracy is no better than the accuracy achieved with the original criteria (6-8). Thus, for routine practice in many hospitals, clinicians continue to rely on the 30-year-old criteria of Light et al. (1).

Quantitative measurements of plasma cell-free DNA are being proposed for use in prenatal diagnosis, cancer testing, acute trauma, and monitoring of transplantation (9-13). Investigation on cell-free DNA has also been carried out in biological fluids other than plasma, e.g., urine (14). With knowledge of the marked increase and near absence of inflammatory and cellular responses in exudative and transudative effusions, respectively, we hypothesize that the cell-free DNA concentration in pleural fluid would be increased in exudative pleural effusions and that measurement of the DNA concentration in pleural fluid may aid in the classification of pleural effusions. Thus, cell-free DNA in pleural fluid might come from the plasma ultrafiltrate in case of transudative effusion. For exudative effusions, on the other hand, much higher concentrations may come from the degradation of apoptotic cells that are present in the effusion. Thus, it was the primary objective of this study to explore the use of quantitative pleural fluid cell-free DNA in the investigation of the etiology and diagnosis of pleural effusions.

Materials and Methods


Ethical approval was obtained from the Clinical Research Ethics Committee, Faculty of Medicine, The Chinese University of Hong Kong (Reference No. CRE-2001.291). We included consecutive patients presenting with pleural effusions to the Departments of Medicine & Therapeutics and Clinical Oncology, Prince of Wales Hospital, Hong Kong, between August 1, 2001, and March 31, 2002, who required therapeutic or diagnostic aspiration to alleviate or investigate the etiology of the effusions and who consented to the study. Informed consent was documented in a signed form. All patients except those who died during hospitalization were followed up at the outpatient clinics of the two specialties for at least 6 months after discharge, with monitoring of the response to treatment.


We collected 20 mL of pleural fluid, 3 mL of EDTA blood, and 4 mL of clotted blood in one setting from each patient at the time of diagnostic or therapeutic tapping. The pleural fluid, EDTA-blood, and clotted blood samples were centrifuged at 1600g for 10 min. A portion of the supernatants from the pleural fluid and clotted blood samples was used to measure both LDH activity and total protein concentration in serum and pleural fluid; results were interpreted according to the modified Light criteria. The remaining supernatants from the pleural fluid samples and the EDTA plasma (with care not to disturb the buffy coat layer) were transferred into 1.5-mL polypropylene tubes and were subjected to microcentrifugation for 10 min at 13 000g (Eppendorf Centrifuge 5415D). These recentrifuged pleural fluid and EDTA-plasma samples were then used for DNA extraction followed by PCR analysis. All samples were processed within 2 h of sample collection, transferred into polypropylene tubes, and stored at -20[degrees]C until further use.

Another portion of the aspirated pleural fluid was sent for cytologic analyses, including a check for malignant cells, cell counts, Gram stain, routine microbiological culture, acid-fast stain, and culture for Mycobacterium tuberculosis, for the clinical management of the patients. Pleural biopsies were performed in selected cases of pleural effusions if the initial clinical suspicion for pulmonary tuberculosis was high. All of the above clinical evaluations were performed by K.M.C., C.B.L., and A.T.C.C., all of whom had obtained the relevant specialist qualification (i.e., Membership in the Royal College of Physicians) and were blinded to the pleural fluid DNA results. The clinical evaluations also formed the reference standard for the diagnosis and were not revealed to the laboratory researchers (M.H.M.C., L.Y.S.C., K.C.K.C., C.W.L., and Y.M.D.L.) until the end of the analyses for pleural fluid cell-free DNA concentration.


DNA was extracted from the above-processed pleural fluid aliquots with use of a QIAamp Blood Kit (Qiagen) according to the blood and body fluid protocol, as recommended by the manufacturer. The volume of pleural fluid used for DNA extraction was 600-800 [micro]L per column.


All of the pleural fluid aliquots were subjected to real-time quantitative PCR amplification for the ([beta]-globin gene as described previously (15). The volume of extracted pleural fluid DNA used for amplification

was 5 [micro]L. Real-time quantitative PCR was performed in a 7700 Sequence Detector (Applied Biosystems). The theoretical and practical aspects of real-time quantitative PCR have been described in detail elsewhere (16). Duplicate analyses were performed for each sample, and the mean result was used for further analysis. A calibration curve was constructed in parallel with each assay. Doubly distilled water was used as the negative control for quantitative real-time PCR. The results were expressed as genome-equivalents by use of the conversion factor of 6.6 pg of DNA per cell (17). Amplification data were analyzed and stored by the Sequence Detection System Software, Ver. 1.6.3 (Applied Biosystems). The pleural fluid DNA concentrations, expressed in genome-equivalents/mL, were calculated as described previously (15).


Data analysis for correlation coefficient (r), linear regression, nonparametric Kruskal--Wallis test statistics, and ROC curve analysis was performed with the MedCalc 7.0 program ( The best cutoff concentration for pleural fluid DNA was chosen automatically by the MedCalc 7.0 program as the concentration with the highest diagnostic accuracy, i.e., the sum of the false-negative and false-positive rates was minimized. With the cutoff concentration determined, indicators of the diagnostic accuracy of pleural fluid DNA and the modified Light criteria could be calculated using the discharge, microbiological, or histologic diagnoses and the progress at follow-up sessions at outpatient clinics as the reference standard. The 95% confidence intervals (CIs) for sensitivity, specificity, and the positive and negative likelihood ratios were calculated according to Simel et al. (18).


We recruited 50 patients after informed consent between August 1, 2001, and March 31, 2002, of whom 38 were from the Department of Medicine & Therapeutics and 12 were from the Department of Clinical Oncology. There were 32 males and 18 females with an age range of 6-99 years (median, 69 years; interquartile range, 54-77 years). No patients experienced adverse events as a result of the diagnostic or therapeutic aspiration. The refusal rates in the Departments of Medicine & Therapeutics and Clinical Oncology were at 10% and 20%, respectively. The patient demographics with their respective discharge, microbiological, or histologic diagnoses are presented in Table 1.

The interassay analytical CV of the real-time PCR system for pleural fluid cell-free DNA, based on 20 replicate analyses of pleural fluid samples, was 26% for samples from patients with low pleural fluid cell-free DNA concentrations [mean (SD), 19 (5) genome-equivalents/mL], 13% for samples from patients with medium cell-free DNA concentrations [358 (48) genome-equivalents/mL], and 14% for samples from patients with high cell-free DNA concentrations [5040 (710) genome-equivalents/mL].

There were significant differences in the pleural fluid [beta]-globin DNA concentrations between the malignant and transudative (P <0.0001), malignant and infective (P = 0.048), and infective and transudative (P <0.0001) groups.

The ROC curve for pleural fluid [beta]-globin DNA concentration is plotted in Fig. 1. The area under the curve was 0.95 (95% CI, 0.84-0.99). The best cutoff concentration (see Materials and Methods) for pleural fluid [beta]-globin DNA was 509 genome-equivalents/mL, calculated automatically by the MedCalc 7.0 program. At this cutoff concentration, 46 of 50 [sensitivity, 91% (95% CI, 76-98%); specificity, 88% (95% CI, 64-98%)] pleural effusions were correctly classified into the exudative and transudative groups compared with the reference standard. The positive and negative likelihood ratios were 7.7 (95% CI, 3.1-19.5) and 0.10 (95% CI, 0.04-0.27), respectively, at this cutoff concentration. When we used the modified Light criteria, 43 of 50 [sensitivity, 97% (95% CI, 91-100%); specificity, 67% (95% CI, 45-89%)] pleural effusions were correctly classified into the exudative and transudative groups compared with the reference standard. The positive and negative likelihood ratios for the modified Light criteria were 3.9 (95% CI, 1.8-4.6) and 0.05 (95% CI, 0.01-0.66), respectively. The positive predictive values for pleural fluid [beta]-globin DNA and the modified Light criteria were 89% and 84%, respectively. The negative predictive values for pleural fluid [beta]-globin DNA and the modified Light criteria were 100% and 92%, respectively.

There were significant correlations between pleural fluid [beta]-globin DNA concentration and pleural fluid LDH activity [r = 0.76 (95% CI, 0.61-0.86); P<0.0001] as well as pleural fluid [beta]-globin DNA and pleural fluid total protein concentration [r = 0.67 (95% CI, 0.48-0.80); P <0.0001] as shown in Figs. 2 and 3, respectively. Pleural fluid [beta]-globin DNA was not significantly correlated with plasma [beta]-globin DNA [r = 0.12 (95% CI, -0.17 to 0.39); P = 0.43].



To the best of our knowledge, this is the first prospective study to demonstrate the potential diagnostic use of cell-free DNA in pleural fluid for the classification of pleural effusions, as evidenced by the marked difference in pleural fluid [beta]-globin DNA concentrations among malignant, infective, or transudative effusions. We chose the [beta]-globin gene as the representative marker for pleural fluid cell-free DNA because each diploid nucleated cell contains two copies of this gene.



We hypothesize that the possible origins of pleural fluid DNA could be ultrafiltration from the plasma or local production from dying or apoptotic cells. A higher inflammatory element is expected to be involved in the process of infective effusion than in malignant processes. This inflammatory process is brought on by the infiltration of polymorphs followed by lymphocytes. Most of the polymorphs will eventually die in the late phase of inflammation. The tissue response induced by lymphocytic cytokines would also promote the ongoing process of cell death and apoptosis (19). These dying cells will release their DNA contents into the pleural space, allowing detection and quantification by the real-time quantitative PCR technique. By the same token, there could be continuous turnover or death of malignant cells in the pleural effusions. Our data also demonstrate that the pleural fluid [beta]-globin DNA concentration is higher in infective effusions than in malignant effusions. For transudative effusions, the degree of inflammatory and cellular response is virtually nonexistent; therefore, the small amounts of cell-free DNA detected could be the result of a plasma ultrafiltration process rather than a local production.

The diagnostic accuracy of pleural fluid cell-free DNA, in terms of its high specificity (88%; 95% CI, 64-98%), is at least as good as that of the modified Light criteria (67%; 95% CI, 45-89%). The sensitivities of these two approaches are similar, 91% (95% CI, 76-98%) and 97% (95% CI, 91-100%), respectively. Moreover, the modified Light criteria utilize two biochemical markers from both the pleural fluid and serum with a calculation of their respective ratios. Such derived markers are inferior to the use of a single marker from the pleural fluid because the "OR" rule in the modified Light criteria, i.e., one of the three criteria is met, is suboptimal in the sense that a high sensitivity is obtained at the sacrifice of its specificity (20). On the basis of the area under the ROC curve (0.95; 95% CI, 0.84-0.99), pleural fluid DNA is able to achieve a high diagnostic accuracy that is comparable to the accuracies reported in the literature for different single markers (8).

Pleural fluid cell-free DNA can provide some pathophysiologic information to clinicians about the degree of cell death or apoptotic activity within the pleural cavity. The significant correlations of pleural fluid [beta]-globin DNA concentration with the pleural fluid LDH activity and total protein concentration imply that cell-free DNA, LDH, and total protein arise by similar mechanisms through cell death and the subsequent release of the intracellular contents into the pleural cavity. The finding that pleural fluid [beta]-globin DNA was not significantly correlated with plasma [beta]-globin DNA [r = 0.12 (95% CI, -0.17 to 0.39); P = 0.43] further supports that the appearance of cell-free DNA in pleural fluid is attributable to local release in the pleural cavity.

When we used the abbreviated Light criteria adopted by Heffner et al. (8) (pleural-fluid-to-serum total protein ratio and pleural fluid LDH more than two-thirds of the serum reference interval), 43 of the 50 cases in our study were correctly classified into exudates and transudates with the same diagnostic accuracy [sensitivity, 97% (95% CI, 91-100%); specificity, 67% (95% CI, 45-89%)] as for the modified Light criteria. This finding could easily be explained by the high degree of correlation between pleural fluid LDH and pleural fluid to serum LDH ratio. Thus, as suggested by Heffner et al. (8), the two markers should not be used together in the "OR" rule test set.

Sources of potential bias must be considered in the present study. Although all clinical collaborators involved in this study were blinded to the pleural fluid DNA concentration results until the end of the study, they were not blinded to the results of the modified Light criteria because of their clinical need to treat the patients. Moreover, patients recruited from the Departments of Medicine & Therapeutics and Clinical Oncology could represent only a proportion, albeit the majority (~90%) of cases of undiagnosed pleural effusion in our hospital. The only exclusion criterion was patient refusal, which was slightly more frequent in the Department of Clinical Oncology. Thus, the present study might have a small selection bias toward patients from the Department of Medicine & Therapeutics, where more cases of pulmonary infection and fluid overload were encountered. In addition, the exudative effusion category consisted of only infective and malignant causes. The diagnostic performance of our proposed strategy should also be further evaluated and confirmed in a larger sample size with a broader spectrum of diseases. The criteria of Light et al. (1) were known before the study, whereas the cutpoint for pleural fluid cell-free DNA was determined after analysis of the data set. This difference in sequence may lead to another area of potential bias (21).

Another concern is the incremental value of the cell-free DNA analysis in pleural fluid. At present, multichannel analyzers are widely available to measure total protein concentration and LDH activity in both serum and pleural fluid. The running cost is low, and the turnaround time is fast enough to meet clinical needs. For pleural fluid cell-free DNA to be applicable clinically, the main determination factors are cost-effectiveness and the availability of automated methods. With the development of miniaturized technologies, automated DNA extraction followed by PCR and detection in a hand-held cartridge may be practical in the future, whereas the analytical cost will be expected to drop.

In conclusion, the pleural fluid cell-free DNA concentration has the potential to differentiate between exudative and transudative effusions. This conclusion requires confirmation in a larger study in which the cutpoint is not determined after the data are collected. The incremental value of the test also requires investigation.

Received December 3, 2002; accepted February 26, 2003.


(1.) Light RW, McGregor MI, Luchsinger PC, Ball WC. Pleural effusions: the diagnostic separation of transudates and exudates. Ann Intern Med 1972;77:507-13.

(2.) Hirsch A, Ruffle P, Nebut M, Bignon J, Chretien J. Pleural effusion: laboratory tests in 300 cases. Thorax 1979;34:106-12.

(3.) Peterman TA, Speicher CE. Evaluating pleural effusions. A two-stage laboratory approach. JAMA 1984;252:1051-3.

(4.) Roth BJ, O'Meara TF, Cragun WH. The serum-effusion albumin gradient in the evaluation of pleural effusions. Chest 1990;98: 546-9.

(5.) Tarn AC, Lapworth R. Biochemical analysis of pleural fluid: what should we measure? Ann Clin Biochem 2001;38:311-22.

(6.) Romero S, Candela A, Martin C, Hernandez L, Trigo C, Gil J. Evaluation of different criteria for the separation of pleural transudates from exudates. Chest 1993;104:399-404.

(7.) Vives M, Porcel JM, Vicente de Vera M, Ribelles E, Rubio M. A study of Light's criteria and possible modifications for distinguishing exudative from transudative pleural effusions. Chest 1996; 109:1503-7.

(8.) Heffner JE, Brown LK, Barbieri CA, for the Primary Study Investigators. Diagnostic value of tests that discriminate between exudative and transudative pleural effusions. Chest 1997;111:970-80.

(9.) Lo YMD, Hjelm NM, Fidler C, Sargent IL, Murphy MF, Chamberlain PF, et al. Prenatal diagnosis of fetal RhD status by molecular analysis of maternal plasma. N Engl J Med 1998;339:1734-8.

(10.) Lo YMD, Chan LYS, Lo KW, Leung SF, Zhang J, Chan ATC, et al. Quantitative analysis of cell-free Epstein-Barr virus DNA in plasma of patients with nasopharyngeal carcinoma. Cancer Res 1999;59: 1188-91.

(11.) Lo YMD, Rainer TH, Chan LYS, Hjelm NM, Cocks RA. Plasma DNA as a prognostic marker in trauma patients. Clin Chem 2000;46: 319-23.

(12.) Lo YMD, Tein MSC, Pang CCP, Yeung CK, Tong KL, Hjelm NM. Presence of donor-specific DNA in plasma of kidney and liver-transplant recipients. Lancet 1998;351:1329-30.

(13.) Lui YYN, Chik KW, Chiu RWK, Ho CY, Lam CWK, Lo YMD. Predominant hematopoietic origin of cell-free DNA in plasma and serum after sex-mismatched bone marrow transplantation. Clin Chem 2002;48:421-7.

(14.) Botezatu I, Serdyuk 0, Potapova G, Shelepov V, Alechina R, Molyaka Y, et al. Genetic analysis of DNA excreted in urine: a new approach for detecting specific genomic DNA sequences from cells dying in an organism. Clin Chem 2000;46:1078-84.

(15.) Lo YMD, Tein MSC, Lau TK, Haines CJ, Leung TN, Poon PMK, et al. Quantitative analysis of fetal DNA in maternal plasma and serum:W implications for non-invasive prenatal diagnosis. Am J Hum Genet 1998;62:768-75.

(16.) Heid CA, Stevens J, Livak KJ, Williams PM. Real time quantitative PCR. Genome Res 1996;6:986-94.

(17.) Lo YMD, Zhang J, Leung TN, Lau TK, Chang AMZ, Hjelm NM. Rapid clearance of fetal DNA from maternal plasma. Am J Hum Genet 1999;64:218-24.

(18.) Simel DL, Samsa GP, Matchar DB. Likelihood ratios with confidence: sample size estimation for diagnostic test studies. J Clin Epidemiol 1991;44:763-70.

(19.) Schuster N, Krieglstein K. Mechanisms of TGF-R-mediated apoptosis. Cell Tissue Res 2002;307:1-14.

(20.) Porcel JM, Vives M, Esquerda A, Rivas MC. Pleural protein capillary electrophoresis for the separation of transudates and exudates. Clin Chem 2001;47:975-6.

(21.) Greenhalgh T. How to read a paper: papers that report diagnostic or screening tests. BMJ 1997;315:540-3.


Departments of [1] Chemical Pathology, [2] Medicine & Therapeutics, and [3] Clinical Oncology, The Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong SAR.

* Author for correspondence. Fax 852-2194-6171; e-mail
Table 1. Patient demographics and their respective discharge,
microbiological, or histologic diagnoses.

 Type of effusion

 Malignant Infective
 (n = 20) (n = 11)

M/F 14/6 7/4
Age, years
 Range 39-86 6-93
 Median 70 42
 Interquartile range 62.5-74.5 27.3-72.5
Diagnosis (n) Lung carcinoma (11) Pulmonary
 Non-small cell (7) tuberculosis (9)
 Small cell (3) Empyema (1)
 Adenocarcinoma (1) Pneumonia (1)
 Breast carcinoma (1)
 Colon carcinoma (1)
 carcinoma (1)
 carcinoma (1)
 T-cell lymphoma (1)
 Unknown primary
 cancer (4)

 Type of effusion

 Transudative Others
 (n = 18) (n = 1)

M/F 11/7 0/1
Age, years
 Range 41-99 82
 Median 68
 Interquartile range 63-78
Diagnosis (n) Congestive heart Chylothorax (1)
 failure (6)
 End-stage renal
 failure (8)
 Cirrhosis (4)
COPYRIGHT 2003 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2003 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Molecular Diagnostics and Genetics
Author:Chan, Michael H.M.; Chow, Kai Ming; Chan, Anthony T.C.; Leung, Chi Bon; Chan, Lisa Y.S.; Chow, Kathe
Publication:Clinical Chemistry
Date:May 1, 2003
Previous Article:Evaluation of electronic microarrays for genotyping factor V, factor II, and MTHFR.
Next Article:Purification of human kallikrein 6 from biological fluids and identification of its complex with [[alpha].sub.1]-antichymotrypsin.

Terms of use | Copyright © 2018 Farlex, Inc. | Feedback | For webmasters