Evolution of Idiopathic Pleural Effusion

Evolution of Idiopathic Pleural Effusion* A Prospective, Long-term Follow-up Study Jaume S. Ferrer, MD; Xavier G. Munoz, MD; Ramon M. Orriols, MD; Richard W. Light, MD; and Ferran B. 1J1orell, MD

The management of idiopathic pleural effusion remains controversial. Because the long-term evolution of this entity is not well known, two different approaches, aggressive and conservative, have been proposed. We conducted a 10-year study of the evolution of idiopathic pleural effusion. Methods: Between 1984 and 1994, we prospectively studied 40 consecutive patients (30 men and 10 women; mean [::tSD] age, 53.8::!::19.4 years) with exudative pleural effusion undiagnosed after exhaustive evaluation. The pleural fluid adenosine deaminase level was below 43 lUlL in all; periodic chest radiographs and clinical evaluation were carried out in all patients for a mean of 62 months (range, 36 to 108 months). Further diagnostic procedures were performed whenever the effusion recurred or when indicated by the clinical picture. Results: Effusions resolved in a mean time of 5.6 months (range, 7 days to 48 months). Five patients (12.5%) had one or more relapses of their pleural effusion, and in a further 5 (12.5%), the effusion persisted unchanged for more than 1 month. In 32 cases (80%), no potential cause of the effusion was detected. The diagnoses in the remaining eight cases were asbestos pleural effusion in three, pulmonary adenocarcinoma in one, mesothelioma in one, congestive heart failure in one, liver cirrhosis in one, and rheumatoid arthritis in one. Tuberculosis was not detected in any of the cases, although 19 patients initially had positive tuberculin tests. Conclusions: Most idiopathic pleural effusions follow a benign course. Our results support conservative treatment of patients with idiopathic pleural effusion. Antituberculous treatment does not appear to be warranted, regardless of tuberculin test results, if the pleural fluid adenosine deam(CHEST 1996; 109:1508-13) inase level is not elevated. Key words: adenosine deaminase; pleural effusion; pleurisy; tuberculous pleuritis

Evaluation of exudative pleural effusion usually includes complete clinical examination, appropriate blood tests, radiographs, studies of pleural fluid, and needle biopsy of pleura. However, following this procedure, approximately 20% of patients still have undiagnosed conditions 1·2 and the clinical management of these cases is controversial. vVhen a neoplastic etiology of the effusion is suspected from clinical and radiologic data, more invasive diagnostic techniques such as bronchoscopy, thoracoscopy, or thoracotomy should be performed. 3A However, frequently there are no indications that the effusion is malignant and thus it can be considered idiopathic. Some authors5·6 advocate antituberculous treatment for patients with idiopathic pleural effusion and positive tu*From the Servei de Pneumologia. Hospital General Universitari Vall d'Hebron, Barcelona, Spain (Drs. Ferrer, Munoz, Orriols, and Morell); and the Deeartment of Veterans Affairs, Medical Center, Long Beach, Calif (Dr. Light). Manuscript received October 17, 1995; revision accepted December 20. Reprint mquests: Dr. Ferrer, Sen;ei de Pneumologia, Hospital General Universitari Vall d'Hehmn, Passeig Vall d'Hehron, sin, 08035 Barcelona, Spain; or contact Dr. Light,\!AMC, 5901 East 7th Street, Long Beach, CA 90822

1508

berculin test, since in two previous works, 7·8 between 43% and 6.5% of individuals with an undiagnosed pleural effusion and a positive tuberculin test developed tuberculosis during follow-up. However, to date and to our knowledge, no recent studies have justified this approach. When the tuberculin test is negative, some authors routinely perform thoracoscopy or thoracotomy,9·10 while others perform these more invasive tests only if the pleural effusion continues to worsen on observation. 6 The controversy in establishing guidelines for the management of idiopathic pleural effusion exists because the long-term evolution of this disorder is not well known. In fact, very few reports, all retrospective, have been published on the evolution of idiopathic pleural effusion. 11 -14 These studies give rise to some confusion since most of the malignancies detected during follow-up had initially been suspected 11 -13 or patient selection was based on the nonspecificity of pleural biopsy specimens and not on the lack of diagnosis after exhaustive workup. 14 Previously, we had retrospectively reviewed the outcome of idiopathic pleural effusion diagnosed at our Clinical Investigations

center between 1974 and 1984. We were able to obtain information on 53 of 114 patients, 5 of whom developed malignancy during evolution. In 1984, we initiated a prospective follow-up study and report herein the results in 40 patients included up to 1991 and followed up until March 1994. MATERIALS AND METHODS

Patients

Between January 1984 and March 1991, 394 patients with pleural effusion were admitted to our department. Initial studies included the following: medical history to rule out exposure to asbestos or previous drug ingestion; complete physical examination; blood analysis; chest radiograph; and tuberculin test. Studies on pleural fluid obtained by thoracentesis included the following: determination of the levels of glucose, protein, adenosine deaminase, and lactate dehydrogenase; bacteriologic study with the ZiehlNeelsen stain and culture in the Lowenstein-Jensen medium; and cytologic study. Pleural biopsy with Abram's needle was performed on all exudative effusions of indeterminate cause after thoracentesis if this was possible. Pathologic and mycobacteriologic studies of pleural biopsy samples were always performed. Other examinations such as serologic tests for atypical pneumonia, chest CT, ventilationperfusion scintigraphy, echocardiography, Hberoptic bronchoscopy, thoracoscopy, or thoracotomy were performed at the discretion of the attending physician. Diagnosis was established in 341 patients (86.5%) (Table 1). Pleural tuberculosis was diagnosed if there was a positive Lowenstein culture in pleural fluid or pleural biopsy specimen or if granulomas were detected in pleural tissue. In 56 patients younger than 30 years with lymphocytic pleural fluid, positive tuberculin test, and pleural fluid adenosine deaminase level higher than 43 lUlL, antituberculous treatment had been successfully administered without further studies. The remaining 53 patients (13 ..5%) who had idiopathic pleural effusions were referred to us for follow-up. Our analysis includes the results for 40 patients. The remaining 13 had been excluded initially; in 4 of these, the patient's medical history suggested a cause for the pleural effusion. Two had a history of exposure to asbestos: one of these was diagnosed as having a benign asbestos pleural effusion, and mesothelioma was detected in the other after 3 months. In the other two cases, the effusion was attributed to congestive heart failure and liver cirrhosis, respectively. Nine patients failed to attend follow-up visits, despite attempts to contact them by telephone. Follow-up Studies

Patients were evaluated monthly for the first 3 months and at varying intervals-maximal 6 months-thereafter depending on resolution of the effusion. All patients were questioned on several occasions about previous exposure to asbestos. Clinical status was ascertained and physical examination and chest radiographs were performed at each visit. Additional examinations were obtained depending on the patient's clinical status or radiologic evolution. Statistical Analysis

Results were expressed as the mean, SD, and percentages. Chisquare test was used to analyze the relationship among qualitative variables. RESULTS

Baseline clinical and radiologic data for the 40 patients (30 men and 10 women) at diagnosis are

Table !-Etiology of Pleural Effusions Seen in Our Department Between January 1984 arul March 1991

1-Tuberculosis 2-Malignancy 3---Pneumonia 4--Congestive heart failure 5-Pulrnonary thromboembolism 6--- Miscellaneous 7-Idiopathic Total

No.

%

113 80 73 28 21 26 53 394

28.7 20.3 18.5 7.1 5.3 6.6 13.5

shown in Table 2. A summary of the pleural fluid findings and the procedures used in the diagnostic workup are shown in Table 3. In the 19 patients with positive tuberculin tests, the pleural fluid was lymphocytic in 16 and polymorphonuclear in 3. In the 21 patients with negative tuberculin tests, the pleural fluid was lymphocytic in 19 and polymorphonuclear in 2. All patients had adenosine deaminase pleural fluid levels below 43 lUlL. Pathologic study of the pleura was not available in five patients owing to the scant amount of pleural fluid at diagnosis. The mean follow-up was 62 months (range, 36 to 108 months; median, 60 months). The 19 patients with a positive tuberculin test were followed up for a mean of 69 months (range, 36 to 108 months; median, 72 months), and 16 of them for more than 48 months. Evolution of the pleural effusion until resolution is detailed in Figure l. During the initial hospitalization, one patient had undergone talc pleurodesis following thoracoscopy, and pleural decortication was performed during diagnostic thoracotomy in a second. The mean Table 2-Baseline Clinical arul Radiologic Data for 40 Patients With Idiopathic Pleural Effusion Data

No.

Age, yr (mean±SD) Sex, M/F Smoking, yes/no Previous diagnosis, % of patients COPD Systemic hypertension Asthma Diabetes Clinical data at hospital admission, % of patients Dyspnea Chest pain Fever Weight loss Pericardia! effusion Positive tuberculin test, % of patients Radiology of pleural effusion, % of patients Right Left Bilateral Massive Higher than \2 hemithorax

53.8±19.4 30/10 27/13

CHEST I 109 I 6 I JUNE, 1996

27.5 15 5 2.5 52.5 77.5 12.2 14.6 5 47.5 32.5 60 7.5 7.5 22.5

1509

Table 3---Baseline Diagnostic Workup Data for 40 Patients With Idiopathic Pleural Effusion No.

Data Pleural fluid data Serosanguineous, % of patients Exudates,* % of patients ADA, 1 UI, mean±SD LDH,t UI, mean±SD Cytology, % of patients Polymorphonuclear Lymphocytic Pleural histology, % of patients 1 pleural biopsy "'2 pleural biopsies No pleural histology Pleuroscopy, % of patients Thoracotomy, % of patients Chest CT, % of patients Fiberoptic bronchoscopy, % of patients Lung scintigraphy, % of patients Echocardiography, % of patients

5 100 12.7±6.7 314±210 12.5 87.5 35 52.5 12.5 37.5 2.5 57.5 45 40 15

*According to the Light criteria. 1ADA=adenosine dearninase; LDH=lactate dehydrogenase.

time for the resolution of the effusion in the remaining 38 patients was 5.6 months (range, 7 days to 48 months; median, 1.7 months). The effusion improved during the first month after hospital discharge in 33 of these patients; however, it took more than 2 months for

the effusion to completely resolve in 12 patients. In 5 patients (12.5%), the pleural effusion failed to improve in the first month and the subsequent resolution time varied from 4 months to 4 years. Five patients (12.5%) had a relapse of the effusion after an initial resolution; 1 at 1 month, 2 at 2 months, 1 patient had 2 relapses, at 4 and 9 months, and the fifth had had 4 relapses before being referred to our center (Fig 1). Thoracentesis was performed at each relapse in all patients. Further thoracenteses, ranging from one to four depending on the evolution, were performed during follow-up in the five patients in whom improvement was not evident in the first month. Results of pleural fluid biochemical, cytologic, and bacteriologic studies were all normal in these ten patients. Twelve patients (30%) had occasional pain in the affected hemithorax. The pain bore no relation to the type of radiologic resolution (Table 4). A possible explanation for the initial pleural effusion was obtained during the follow-up in eight patients. Three patients (7%) eventually recalled having had previous contact with asbestos. A 37-year-old man with a pleural effusion that resolved slowly in a year stated after 6 months that at the age of 17 years he had had contact with asbestos for 2 years while working on the manufacture of brake linings. Another 47-year-old man with a pleural effusion who had relapses on 3 occasions

FOLLOW-UP

IMPROVEMENT IN THE FIRST MONTH

COMPLETE

RELAPSES

RESOLUTION

~

[

~

G{

-

16

~

40

1 ASBESTOS 1 MESOTHELIOMA

1 ASBES'NS

TIME

CAUSE

TIME

12 MONTHS 4 YEARS

1 MESOTHELIOMA

5 YEARS""

24 MONTHS 1 P. THROMBOEMBOLISM 18 MONTHS* 1 CXlPD 18 MONTHS* 1 INTESTINAL PERFORATION 3 YEARS

--

6 MONTHS

[J I NO

--t.r

NO

-R IDIOPATHIC PLEURAL __. EFFUSION

DIAGNOSIS

DEATHS

1 SUICIDE

I [1]

- - - 4 MONTHS-4 YEARS _.

5

5

{1

NO

5

__.

HEPATHIC CIRRHOSIS 1 RHEUIIATOID ARTRITIS

- - { 1 CONGESTIVE

7 MONTHS

12 MONTHS*

1 HEPATHIC ENCEPHALOPATY

7 YEARS

1 PULMONARY 2,5 YEARS

1 YEAR

TROMBOEMBOLISM

1 STRORE

7 YEARS

3 YEARS

HEART

FAILURE

-.------...... DECORTICATION 1

r;;l-- { 1 PULMONARY ~

ADENOCARCINOMA

1 PULMONARY 2 YEARS

ADENOCARCINOMA

2, 5 YEARS

FIGURE 1. Evolution of the 40 patients with idiopathic pleural effusion. Asterisk (*)=necropsy was available; T =a talc pleurodesis was carried out on the fifth relapse of his pleural effusion.

1510

Clinical Investigations

during the first year, admitted after 12 months, sporadic exposure to asbestos 25 years earlier while working as a linotypist. Finally, a 70-year-old man with a pleural effusion that resolved in 3 months, commented after 2 years that he had had an asbestos board in the kitchen of a vacation house that he used every summer for 47 years. All3 patients remain asymptomatic after more than 3 years of follow-up. Neoplasms were detected in 2 patients (5%). A 43year-old male smoker with a right -sided effusion occupying less than half the hemithorax, undiagnosed after bronchoscopy and thoracotomy, presented with a pulmonary nodule in the right upper lobe that was diagnosed as adenocarcinoma 2 years after the initial effusion. The patient died 6 months later but necropsy was not available. Another 51-year-old male smoker had presented with a relapsing left pleural effusion on 4 occasions, which remained undiagnosed after 4 needle pleural biopsies. He was referred to our center on the fifth relapse, and results of studies, including thoracoscopy, proved normal. Six months later, he presented with pain and left pleural thickening. Trucut biopsy of the thickened pleura yielded the diagnosis of mesothelioma. The patient died at another center 10 months later, but necropsy was not performed. Three patients developed diseases during the follow-up, which possibly could have been responsible for the initial effusion. A 77-year-old man initially presented with a persistent exudative pleural effusion with no evident liver disease. Two years later, the pleural effusion recurred and he was diagnosed as having cirrhosis of the liver with ascites. Analysis of the pleural effusion showed it to be a transudate at this time. The pleural effusion resolved with treatment of the liver disease. The patient died from hepatic encephalopathy after 8 years of follow-up. A 65-year-old woman with persistent pleural effusion for 7 months was diagnosed as having rheumatoid arthritis 2.5 years later. A 72-year-old woman with a history of arterial hypertension had presented with an exudative right pleural effusion. Subsequently she was diagnosed as having left heart failure after 1 year without recurrence of her pleural effusion. She died from a stroke after 3 years of follow-up. Pericardia! effusion had been detected in the initial pleural effusion study in 2 patients (5%). The pleural effusion improved initially and resolved within a year in both cases. The pericardia! effusion resolved in 1 patient after 6 months of follow-up, while the other progressed to constrictive pericarditis and currently, 3.5 years later, is awaiting pericardiectomy. Nine patients (22%) died during follow-up (Fig 1). The one patient who died from pulmonary thromboembolism had a normal perfusion-ventilation scintigraphy during the initial hospital admission.

Table 4-Relation Between Chest Pain and Residual Radiologic Findings in Patients With Idiopathic Pleural EJJUsion* Chest Radiograph at the End of Follow-up

No.

Chest Pain Posteffusion

6 18

3 4

16

5 12

Normal Blunting of the costophrenic angle Residual pachypleuritis Total

40

*x2 test showed no significant differences. We attempted to contact the nine patients who were not included in the follow-up because they did not keep their follow-up appointments. We successfully contacted five of these nine patients by telephone. Four had remained asympatomatic, while the fifth patient, a 62-year-old man, died from pneumonia 1 year after his initial pleural effusion. In four cases no information was available. DISCUSSION

The results of this study show that an idiopathic pleural effusion with no clinical or radiologic evidence of malignancy resolves spontaneously, albeit on occasions (12%) after a prolonged period of time or several relapses. A major finding in this study was that none of the patients developed tuberculosis during the follow-up period, despite the fact that 19 patients had a positive tuberculin test. The current recommendation for antituberculous treatment for patients with idiopathic pleural effusion and positive tuberculin test15 is based on the studies by Patiala7 in 1954 and Roper and WaringB in 1955. Those authors studied the pleural effusions without pleural biopsy, and microbiologic study of pleural fluid was not performed in all cases. 8 Bearing in mind that the diagnosis of tuberculosis is now established in 90 to 95% of patients with tuberculous pleuritis with the studies that we performed,5 most of the patients of Roper and WaringB and Patiala7 who later developed tuberculosis would now have their conditions diagnosed at the onset. Consequently, we believe that the evolution described in those studies 7·8 and the treatment recommended8 are applicable to tuberculous, but not to idiopathic pleural effusion. A review of retrospective studies on idiopathic pleural effusion in the Anglo-Saxon population yielded 25 patients with positive tuberculin test. Although it was not stated that they had received empiric tuberculous treatment, none later developed tuberculosis.4.12.16 Moreover, Arrington et al 17 in 1966 did not find tuberculosis on follow-up in any of their 26 untreated positive tuberculin reactors who had nonspecific open pleural biopsies. Consequently, we do not believe that antituberculous treatment should be CHEST/109/6/JUNE, 1996

1511

administered to patients with idiopathic pleural effusion and low pleural fluid adenosine deaminase levels even though they have a positive tuberculin test. However, we do recommend antituberculous therapy for patients with a lymphocytic predominant pleural effusion and positive tuberculin test with a high pleural fluid adenosine deaminase level. The reason for this approach is based on the fact that the pleural biopsy specimen in at least 10% of tuberculous effusions reveals nonspecific histologic and microbiologic findings, 18 and that almost all patients with tuberculous pleuritis have a pleural fluid adenosine deaminase level above 43 IU/L. 19-21 The incidence of malignancy was very low during the follow-up period in the present series. Only 2 patients (5%) developed malignancy during follow-up, whereas in previous recent studies, this percentage has been between 20% and 30%_ll-1.3 This discrepancy may be due to the fact that pleural effusion evolution was reviewed retrospectively in earlier studies, and the baseline features of the patients were not sufficiently defined. In one study, ll the authors excluded patients empirically treated for tuberculosis if their conditions remained undiagnosed during follow-up, a fact that could give rise to bias. Furthermore, this author and others, 12-14 admitted that neoplasia had initially been suspected in the majority of cases in which it was detected during follow-up, which makes one question the initial idiopathic label in these cases. In our two patients who developed mesothelioma and pulmonary adenocarcinoma, the time elapsed from effusion to diagnosis was 4 and 2 years, respectively. In both cases, initial aggressive techniques failed to detect the neoplasia, a fact referred to in other studies, particularly with respect to the mesothelioma.12,22-24 With respect to pulmonary adenocarcinoma, a previous work showed that in 53% of patients with lung cancer and malignant pleural effusion, the primary malignancy was discovered at the same time or later than the effusion. 25 However, in our patient, it seems unlikely that the effusion was initially malignant, since median survival of patients vvith malignant effusion is less than l year. 26 It would be more reasonable to think that the pleural effusion was unrelated to the neoplasia. 27 A history of asbestos exposure was discovered during follow-up in three cases. Asbestos is a cause of pleural effusion that may easily be overlooked. 28 particularly when the exposure is slight, as in our patients, and thus not initially remembered by the patient. None of the three patients developed further disease in the 3 follow-up years, and thus their diagnosis may have been a benign asbestos-induced effusion. 29 The possible coexistence of idiopathic pleural and pericardia! effusions is known. 30 Two of our patients 1512

presented with idiopathic pleural effusion associated with pericardia! effusion, and one of them developed constrictive pericarditis. The association between constrictive pericarditis and pleural effusion has been previously described. 31 ·32 The pleural effusion resolved after resolution of the pericardia! effusion in one patient, and resolved without treatment of the constrictive pericarditis in the other. Liver cirrhosis was diagnosed in one patient at 2 years. Since the pleural fluid was initially an exudate, hepatic hydrothorax without clinical ascites is very unusual,33 and the subsequent effusion was a transudate, we believe that it is unlikely that the initial effusion was due to liver disease. Congestive heart failure and rheumatoid arthritis, respectively, were diagnosed during follow-up in two cases. Both entities may cause pleural effusion, 6·34 but in our patients, the absence of initial clinical symptoms, the evolution of the effusion, and the time elapsed until diagnosis make it unlikely that the effusions were due to the subsequently diagnosed diseases. In most cases (80%), the cause of the effusion was not established even after prolonged follow-up. It is currently thought that viral or mycoplasma infection, pulmonary embolization, asbestos exposure, or connective-tissue diseases could be responsible in these cases. 35 We think that the long-term follow-up in this study makes it improbable that a connective-tissue disease was the cause of the effusion. Pulmonary embolization was already ruled out in some patients in whom it was suspected. We believe that viral or mycoplasma infection could have been responsible for at least some of the effusions, but it is difficult to establish these diagnoses definitely. 36 In respect to asbestos, a previous exposure was not remembered by 32 patients. However, we believe that asbestos exposure that was not remembered by the patient could have been responsible in some of tl1e cases. In conclusion, the results of this work confirm the impression of some authors 6 that the approach to an idiopathic pleural effusion should be conservative. A patient with an idiopathic pleural effusion and a positive tuberculin test need not be treated for tuberculosis if the pleural fluid adenosine deaminase level is not higher than 43 IU/L. In our experience, persistence or relapse of the effusion does not imply a poor prognosis or justify a change in the approach. However, we recommend follow-up of patients with idiopathic pleural effusion to ensure its resolution and to detect the etiology in some cases. ACKNOWLEDGMENT: The writers thank Cristina O'Hara for help with translation. REFERENCES

Storey DD, Dines DE, Coles DT. Pleural effusion: a diagnostic dilemma. JAMA 1976; 236:2183-86 Clinical lnvestigatons

2 Hirsch A, Ruffle P, Nebut M, eta!. Pleural effusion: laboratory test in 300 ca..'ies. Thorax 1979; 34:106-12 3 Black LF. Pleural effusions. Mayo Clin Proc 1981; 56:201-02 4 Williams T, Thomas P. The diagnosis of pleural effusions by fiberoptic bronchoscopy and pleuroscopy. Chest 1981; 80:56fi-69 5 Salm SA. The pleura. Am Rev Respir Dis 1988; 138: 184-234 6 Light RW. Pleural diseases. 2nd ed. Philadelphia: Lea & Fehiger, 1990; 82 7 Patiala J. Initial tuberculous pleuritis in the Finnish Armed Forces in 1939-1945 with special reference to eventual post pleuritic tuberculosis. Acta Tuberc Scand 1954; 36(suppl):1-.57 8 Roper WH, Waring JJ. Prima1y serofibrinous pleural effi.1sion in military personnel. Am Rev Tuberc 19.5.5; 71:616-34 9 Boutin C, Astoul PH, Seitz B. The role of thoracoscopy in the evaluation and management of pleural effi.1sion. Lung 1990; suppl:1113-21 lO Canto A, Rivas J, Saumench J, et a!. Points to consider when choosing a biopsy method in cases of pleurisy of unknown origin. Chest 1983; 84:l7fi-79 11 Gunnels JJ. Perplexing pleural effusion. Chest 1978; 74:390-93 12 Ryan CJ, Rodgers RF, Unni KK, et a.l. The outcome of patients with pleural effusion of indeterminate cause at thoracotomy. Mavo Clin Proc 1981; 56:145-49 13 Po~ RH, Israel RH, Utell MJ, et a.l. Sensitivity, specificity, and predictive values of closed pleural biopsy. Arch Intern Med 1984; 144:325-28 14 Leslie WK, Kinasewitz CT. Clinical characteristics of the patients with nonspecific pleuritis. Chest 19S8; 94:603-08 15 Light RW. Disorders of the pleura, mediastinum and diaphragm. In: Harrison's principles of internal medicine. New York: McGraw-Hill, 1994 16 Feinsilver SH, Barrows AA, Braman SS. Fiheroptic bronchoscopy and pleural effusion of unknown origin. Chest 1986; 90:51fi-19 17 Anington CW, Hawkins JA, Richert JH, et a!. Management of undiagnosed pleural effi.1sions in positive tuberculin reactors. Am Rev Respir Dis 19fi6; 93:.587-92 18 Levine H, Metzger W, Kay L. Diagnosis of tiJberculous pleurisy by culture of pleural biopsy specimen. Arch Intern Med 1970; 126:269-71 19 Ocana I, Martinez-Vazquez JM, Segnra RM. et al. Adenosine

deaminase in pleural fluids: test for diagnosis of tuberculous pleural effusion. Chest 1983; 84:51-.3 20 Fontes Baganha M, Pego A, Lima MA, et al. Serum and pleural adenosine deaminase: correlation with lymphocytic populations. Chest 1990; 97:60.5-IO 21 Valdes L, San Jose E, Alvarez D, et a.l. Diagnosis of tuberculous pleurisy using the biologic parameters adenosine deaminase, lysozyme, and interferon gamma. Chest 1993; 103:458-6.5 22 Boutin C, Viallat JR, Cargnino P, et a.l. Thoracoscopy in malignant pleural effusions. Am Rev Respir Dis 1981; 124:88-92 23 Page RD, Jeffrey RR, Donnelly RJ. Thoracoscopy: a review of 121 consecutive surgical procedures. Ann Thorac Surg 1989; 4S:66-8 24 Menzies R, Charbonneau M. Thoracoscopy for the diagnosis of pleural disease. Ann Intern Med 1991; 114:271-76 2.5 van de Molengraft FJJM, Vooijs GP. The interval between the diagnosis of malignancy and the development of effusions, with reference to the role of cytologic diagnosis. Acta Cytol 1987; 32:1S3-87 26 Chernow B, Sahn SA. Carcinomatous involvement of the pleura: an analysis of 96 patients. Am J Med 1977; 63:69.5-702 27 Salm SA. Pleural effusion in lung cancer. Clin Chest Med 1993; 14:189-200 28 Mattson SB. Monosymptomatic exudative pleurisy in persons exposed to asbestos dust. Scand J Respir Dis 1975; 56:263-72 29 Epler GR, McLoud TC, Gaensler EA. Prevalence and incidence of benign asbestos pleural effusion in a working population. JAMA 1982; 247:617-22 30 Weiss JM, Spodick DH. Association ofleft pleural effusion with pericardia! disease. N Eng! J Med 19S3; 308:696-97 .31 Hirschmann JV. Pericardia.! constriction. Am Heart J 1978; 96:110-22 32 Tomaselli G, Gamsu G, Stulharg MS. Constrictive pericarditis presenting as a pleural effusion of unknown origin. Arch Intern Med 1989; 149:201-03 33 Johnston RF, Loo RV. Hepatic hydrothorax: studies to determine the source of the fluid and report of 13 cases. Ann Intern Med 1964; 61:38.5-401 34 Walker WC, Wright V. Rheumatoid pleuritis. Ann Rheum Dis 1967; 26:467-74 3.5 Gaensler EA. Mysterious pleural effusions. Lancet 1982; 1:1226 36 Gaensler EA. 'Idiopathic' pleural effusion. N Eng! J Med 1970; 283:816-17

CHEST /109/6/ JUNE, 1996

1513