| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
1
Department of Pathology, Princess Margaret Hospital, Hong Kong, China
2
Department of Chemical Pathology, The Chinese University of Hong Kong, Prince of Wales Hospital, Hong Kong, China
a Author for correspondence. Fax 852-2636-5090; e-mail ching-wanlam{at}cuhk.edu.hk
To the Editor:
The cause of a pleural effusion is not always easily determined. Invasive procedures such as pleural biopsy are indicated only in patients with exudative pleural effusions. Therefore, a frequent early step in the evaluation of pleural effusions is to classify them as transudates or exudates.
The diagnostic criteria developed by Light et al. (1) characterize pleural exudates as having at least one of the following: pleural fluid/serum total protein ratio >0.5; pleural fluid/serum lactate dehydrogenase (LDH) ratio >0.6, and pleural fluid LDH more than two-thirds of the upper reference limit of serum LDH. However, the results produced by Light et al. (1) are not always reproducible (2), and the low specificity of the criteria of Light et al. may lead to unwarranted invasive intervention in up to 20–30% of patients with transudates (3).
One report (4) described the use of protein
electrophoresis (by Tiselius U-tube) to study patterns of protein in
pleural fluid in disease but did not address the role of protein zone
electrophoresis (PZE) in the differentiation of exudate from
transudate. We postulate that low-molecular weight molecules such as
albumin (Mr 66 400),
1-antitrypsin
(Mr 54 000), and transferrin
(Mr 76 500) pass through the pleura to
enter the pleural spaces in transudative pleural effusion, whereas
high-molecular weight molecules such as
2-macroglobulin
(Mr 725 000), haptoglobin
(Mr 400 000), immunoglobulin (IgG,
Mr 150 000; IgA, Mr
160 000; IgM, Mr 950 000), and
ß-lipoprotein (Mr 250 000) do so
only when capillary permeability increases in exudative pleural
effusion formation.
We tested our hypothesis in patients who presented with pleural effusion in the Princess Margaret Hospital during 1997–1998. No selection criteria were set with respect to the type of disease, age, or sex. Routinely, pleural fluid was aspirated for cytology, bacterial culture, mycobacterial culture, total protein, and LDH measurements. Pleural biopsy was performed if relevant pathology was suspected. Diagnosis was done by the respiratory physicians.
Pleural fluid and blood samples were from specimens sent to the laboratory. Specimens were received on the day of collection. Whenever possible, paired samples of pleural fluid and blood collected on the same date were used. If same-date specimens were unavailable, blood samples taken within 24 h before or after the pleural fluid collection were accepted (3). Blood and pleural fluid samples were centrifuged, and the supernatant was separated into aliquots and kept at -80 °C until analysis.
The diagnosis of tuberculous pleuritis required either the identification of Mycobacterium tuberculosis by culture or biopsy or the presence of caseous granuloma. Malignant pleural effusion was diagnosed when malignant tissue in the pleural cavity was shown by pleural biopsy or cytopathology. Effusion was considered parapneumonic when there was an acute febrile illness associated with pneumonia, lung abscess, or bronchiectasis in the absence of malignancy. Empyema was defined as the presence of purulent pleural fluid and positive bacterial culture associated with parapneumonic effusion. Congestive heart failure was diagnosed when all of the following criteria were satisfied: cardiomegaly, radiological evidence of congested lungs, peripheral edema, and response to treatment for congestive heart failure. Renal failure was diagnosed when urea and creatinine were increased in the presence of clinical evidence of fluid overload and an absence of purulent sputum, malignancy, or pulmonary infiltrates. Nephrotic syndrome was classified when the patient had proteinuria >3.5 g/24 h, edema, and hypoalbuminemia. Liver cirrhosis was diagnosed by clinical and laboratory evidence of hepatic damage with portal hypertension or hypoalbuminemia. Other causes of hypoalbuminemia were determined when serum albumin was <30 g/L in the absence of proteinuria and histologically confirmed liver cirrhosis.
Pleural fluid and serum were analyzed in paired samples by PZE (Beckman
Coulter Paragon® SPE kit). Serum was diluted 1:5
(one volume of serum plus four volumes of barbital buffer, as
recommended by the manufacturer) and pleural fluid 1:3 (one volume of
serum plus two volumes of barbital buffer). When any one of the
-2, ß-lipoprotein, and 
bands are present in the
electrophoretogram, the fluid is classified as an exudate. When only
albumin, 1-antitrypsin, and transferrin are
present in the electrophoretogram, the fluid is classified as a
transudate.
According to our criteria, 51 cases had definitive clinical diagnoses.
Among them, 43 were exudative conditions and 8 were transudative
conditions (Table 1
).
|
In most transudates, only low-molecular weight molecules were present,
and no high-molecular weight molecules were present (Fig. 1
, lane P-1), whereas typical exudates (lanes P-2 and P-3)
contained -2, ß-lipoprotein, and bands.
|
The criteria of Light et al. (1) correctly classified 41 exudates (95%) and 3 transudates (38%), whereas PZE correctly classified 43 exudates (100%) and 4 transudates (50%).
The criteria of Light et al. (1) failed to identify an
exudate attributable to pneumonia. PZE clearly shows an -2 band and
a band. The patient had a pleural/serum total protein ratio of
0.33, a pleural fluid LDH of 99 U/L, and a pleural/serum LDH ratio of
0.33. In another case diagnosed as parapneumonia, PZE showed an -2
band and a band. However, the patient had a pleural/serum total
protein ratio of 0.38, a pleural fluid LDH of 79 U/L, and a
pleural/serum LDH ratio of 0.31.
In four transudates diagnosed by clinical criteria, protein zone
electrophoretograms had characteristics of exudates. Three of these
four transudates were classified as exudates by both the criteria of
Light et al. (1) and PZE. They included one patient
suffering from acute pulmonary edema and two patients suffering from
end-stage renal failure. In the remaining case of nephrotic syndrome,
the criteria of Light et al. classified it as a transudate
(pleural/serum total protein ratio, 0.33; pleural/serum LDH ratio,
0.18; and pleural fluid LDH, 47 U/L), but the PZE classified it as an
exudate (-2 and bands present). The patient also suffered from
systemic lupus erythematosus, which probably complicated the influence
of nephrotic syndrome in pleural fluid formation. It is suspected that
this patient also suffered from pleuritis. Apparently, our stringent
clinical criteria for transudative pleural effusion still
underestimated the degree of exudative activity in the pleura.
To put qualitative analysis into quantitative assessment, we used
a Beckman Coulter Appraise® Densitometer System
to scan the electrophoretic gels. The relative area (as a percentage)
of each protein fraction was obtained, and the ratio of the -2
protein fraction to the albumin fraction was calculated. The
data were divided into two separate groups according to the
qualitative separation by PZE into exudates and transudates. There was
a statistically significant difference between the -2/albumin ratios
of PZE-classified transudates and exudates (P = 0.0134,
Mann–Whitney U-test). By ROC curve analysis
(MedCalc, Ver. 4.20), the optimal -2/albumin cutoff ratio for
exudate identification was 0.28. The area under the ROC curve was 0.84
(95% confidence interval, 0.70–0.93). At the optimal cutoff point,
the diagnostic sensitivity and specificity were 85% (95% confidence
interval, 70–94%) and 80% (95% confidence interval, 29–97%),
respectively. At a cutoff value of 0.35, the diagnostic sensitivity and
specificity were 70% (95% confidence interval, 54–83%) and 100%
(95% confidence interval, 100–100%) respectively.
The accuracy and precision of densitometric scanning are affected by many factors: the densitometer, the wavelength used, the agarose slides, and the duration of electrophoresis, fixation, staining, and destaining. In addition, the protein concentration is not always proportional to its staining intensity because of the variations in staining specificity of different protein fractions. Another major shortcoming of densitometric scanning lies in the subjectiveness of manual adjustment of boundaries of protein zones. In this series, there is a negative deviation of albumin concentration obtained by densitometric scanning from that measured by bromocresol green method (Trace® Scientific) of up to 56%. Sources of imprecision in protein fraction quantification by densitometry have been documented elsewhere (5)(6). An even more precise measure could be based on capillary electrophoresis or chromatography with protein detected by ultraviolet absorption (7).
In conclusion, there is a good agreement between the results obtained with the PZE and the criteria of Light et al. (1). In some cases, PZE also provides additional information for the diagnostic separation of exudates from transudates.
Acknowledgments
Special thanks to Albert Y.W. Chan for invaluable comments and suggestions. We also thank W.C. Yu for providing samples and making clinical diagnoses for this study. We thank the reviewers for helpful comments.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  R. Claeys, C. Groven, and F. K. Gorus Capillary Zone Electrophoresis of Proteins in Body Fluids: Comparison of Capillary and Agarose Gel Electrophoresis Clin. Chem., May 1, 2001; 47(5): 967 - 970. [Full Text] [PDF]
|
|
|
|
|
|
J. M. Porcel, M. Vives, A. Esquerda, M. C. Rivas, C.-W. Lam, and M.-L. Chen Pleural Protein Capillary Electrophoresis for the Separation of Transudates and Exudates Drs. Lam and Chen respond: Clin. Chem., May 1, 2001; 47(5): 975 - 976. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
