Evaluation of the Patient with a Pleural Effusion

PCCSU Article | 01.15.08

By Steven A. Sahn, MD, FCCP Dr.Sahn is Professor of Medicine and Director, Division of Pulmonary, Critical Care, Allergy and Sleep Medicine, Medical University of South Carolina, Charleston, SC. Dr.Sahn has disclosed no significant relationships with the companies/organizations whose products or services may be discussed within this chapter. Objectives

1. 2.

To understand the diagnostic value of pleural analysis. To appreciate the value of pleural pH in narrowing the differential diagnosis of the exudate.
Key words: pleural effusion; pleural fluid analysis Pleural effusions are a common occurrence that involve all specialities of medicine and surgery. The annual incidence of new pleural effusions in the United States is estimated to be at least 1.5 million. 1 Thoracentesis is a relatively simple diagnostic procedure that can be performed in the office, at the bedside, in the ICU, or in a dedicated procedure suite. Pleural fluid analysis in isolation may result in only a small percentage of diagnoses with a high clinical likelihood that provides a strong argument for generating a prethoracentesis diagnosis. If the clinician obtains a thorough history, performs a careful physical examination, orders appropriate blood tests, and interprets the chest images thoughtfully prior to thoracentesis, the likelihood of determining a likely clinical, if not definitive, diagnosis is greatly enhanced. In this chapter, I will not only discuss discriminating information from the pleural fluid analysis but also will address salient features of the history, physical examination, and radiographic imaging that should lead to a logical prethoracentesis differential diagnosis.

History
Patients may be symptomatic or asymptomatic on presentation with a pleural effusion. Diagnoses in which patients typically do not present with symptoms are listed in Table 1. For example, about half of 2 the patients with rheumatoid pleural effusion are asymptomatic, and 60 to 70% of patients with benign 3 asbestos pleural effusion (BAPE) have no symptoms. In contrast, as shown in Table 2, patients with bacterial pneumonia, lupus pleuritis,4 postcardiac injury syndrome (PCIS),5 and congestive heart failure virtually always have symptoms in association with their pleural effusions.

Dyspnea and chest pain are the most common symptoms of patients with a pleural effusion. Those with a small pleural effusion and normal underlying lungs may have no perceptible symptoms. In contrast, patients may present with a massive pleural effusion associated with contralateral mediastinal shift, leading to dyspnea at rest. Pleuritic chest pain is the cardinal symptom of pleural inflammation and is

typically accompanied by a pleural effusion. Pleuritic chest pain varies with the intensity of the inflammation. Patients have described pleuritic chest pain as "stabbing," "shooting," or having a "stitch in the side." This type of pain is exacerbated by deep inspiration, cough, or sneezing. Pleuritic chest pain may be focused over the precise location of the inflammation or may be referred. With costal pleural inflammation, the pain tends to be localized directly over the site of pleural involvement and is often associated with tenderness on pressure and cutaneous hypersensitivity. When the lateral, anterior, and portions of the posterior diaphragm are inflamed, pain is perceived diffusely over the lower thorax, back, and abdomen. In contrast, inflammation of the central portion of the diaphragmatic pleura does not result in local pain; pain is referred to the ipsilateral posterior neck, shoulder, and trapezius muscle. Central diaphragmatic pleural inflammation causes referred pain because the sensory fibers of the phrenic nerve enter the spinal cord at the C4 level, which is the usual entry point of sensation from the shoulder.6 Although the number of drugs associated with pleural disease is significantly fewer than those that are presumed to cause parenchymal lung disease, drugs should always be considered as a possible cause of a pleural effusion or pleural fibrosis, especially when the etiology of the effusion is problematic.7 Some of the drugs that have been associated with a pleural effusion, in more than a single case, include bromocriptine, cyclophosphamide, dantrolene, isotretinoin, mesalamine, methotrexate, mitomycin, nitrofurantoin, practolol, procarbazine, and valproic acid.

Physical Examination
Pleural fluid interferes with sound transmission from the lung to the stethoscope, as it separates the lung from the chest wall. Physical signs of a pleural effusion depend on the volume of pleural fluid and the degree of lung compression. The status of the underlying lung and the patency of the bronchial tree will modulate the physical findings. The physical examination results will be essentially normal when the 8 fluid volume is <250 mL. With pleural fluid volume of approximately 500 mL, the typical physical findings include dullness to percussion, decreased fremitus, and normal vesicular breath sounds of decreased intensity compared with the contralateral side.8 When pleural fluid volume exceeds 1000 mL, egophony (E to A change) is heard at the upper level of the effusion.8

Radiology
The standard chest radiograph may provide further diagnostic insight prior to thoracentesis. Diverse differential diagnoses should be considered if the chest radiograph shows a solitary pleural effusion or is associated with another abnormality. With a solitary pleural effusion, infectious causes, such as a tuberculous pleural effusion,9viral pleurisy, 10 or a limited bacterial pneumonia, are possibilities. An isolated pleural effusion is more commonly observed with lupus pleuritis4 and rheumatoid pleurisy11 than with other thoracic manifestations of these diseases or other connective tissue diseases. Metastatic cancer to the pleura, non-Hodgkin lymphoma, and leukemia can also present as a solitary pleural effusion. When a massive effusion (opacification of the entire hemithorax) is present and causes 12 contralateral mediastinal shift, the most likely diagnosis is carcinoma, usually a nonlung primary. With a large or massive pleural effusion without contralateral mediastinal shift, lung cancer of the ipsilateral main stem bronchus13 and malignant mesothelioma14 are most likely. Solitary effusions also may be associated with diseases below the diaphragm, including transudates from hepatic hydrothorax, nephrotic syndrome, urinothorax, peritoneal dialysis and exudates from acute and 15 chronic pancreatitis, chylous ascites, and abscesses or infarcts of the liver and spleen.

Bilateral effusions are most commonly transudates, as seen with congestive heart failure, nephrotic syndrome, hypoalbuminemia, and constrictive pericarditis. The cardiac silhouette is virtually always enlarged in congestive heart failure but may be of normal size with nephrotic syndrome and constrictive pericarditis. 16 Bilateral pleural effusions with a normal heart size are most likely related to malignancy from a nonlung primary but can also occur with lupus pleuritis, rheumatoid pleurisy, hepatic hydrothorax, and hypoalbuminemia.16 A chest radiograph with interstitial infiltrates raises the differential diagnosis of congestive heart failure, rheumatoid disease,11 asbestos pleuropulmonary disease,3 lymphangitic carcinomatosis,17lymphangioleiomyomatosis,18 viral and mycoplasma pneumonia,19 sarcoidosis, 20 and Pneumocystis cariniipneumonia.21 Pleural effusions associated with multiple pulmonary nodules suggest cancer (most common), Wegener granulomatosis,22 rheumatoid disease, 11 septic pulmonary emboli,23 sarcoidosis,20 or tularemia.24

Pleural Fluid Analysis

Virtually all patients with a newly discovered pleural effusion should undergo thoracentesis to confirm a diagnosis. Exceptions include the patient who has typical congestive heart failure with a clinical diagnosis that does not raise suspicion for an alternative diagnosis or a pleural effusion that is extremely small, as with viral pleurisy. Observation is warranted in the previously mentioned examples; however, if the clinical situation worsens or is atypical, a thoracentesis should be performed without delay. For example, if a patient with congestive heart failure has pleuritic chest pain, fever, a unilateral effusion, a left effusion greater than the right effusion, a normal cardiac silhouette, or an oxygen tension valve out of 25 proportion to the clinical situation, a thoracentesis should be done immediately. In a prospective study of 129 patients with pleural effusion published 20 years ago, thoracentesis provided a definitive diagnosis in only 18% and a presumptive diagnosis in 55%.26 In the remaining 27% of patients, the pleural fluid findings were not helpful diagnostically because the values were compatible with two or more clinical possibilities. However, in a number of these patients, the findings excluded possible diagnoses, such as empyema. Over the years, health-care professionals have become better educated about pleural fluid analysis; in conjunction with the clinical presentation, this should enable a definitive or confident presumptive diagnosis in close to 95% of patients. Table 3 enumerates the diseases where a diagnosis can be established "definitively" by pleural fluid analysis.27 Positive cytologic findings, pus in the pleural space, or isolation of an organism from the fluid is obviously diagnostic. However, finding a low pH and a high salivary amylase in the pleural fluid establishes the diagnosis of esophageal rupture in the absence of malignancy. Other examples are the characteristic cytologic findings in a patient with rheumatoid pleurisy28 or finding 2-transferrin in the pleural fluid,29 diagnostic for a duropleural fistula.

Observation The initial diagnostic step is to closely examine the pleural fluid as it is aspirated from the pleural space. The color, character, and odor of the fluid may either be diagnostic or helpful in diagnosis (Table 4).27 Clear, straw-colored fluid suggests a transudate, but a paucicellular exudate cannot be excluded. Sanguinous fluid (hematocrit value <1%) is not helpful diagnostically; however, when the fluid is grossly bloody, the differential diagnosis can be narrowed to malignancy, BAPE, postcardiac injury syndrome, or pulmonary infarction in the absence of trauma. A hemothorax (defined as a pleural fluid/serum hematocrit value 50%) is most commonly due to blunt or penetrating chest trauma but may also occur with invasive procedures, anticoagulation with a hemorrhagic pulmonary infarction, and catamenial 30 hemothorax. When a white or milky fluid is aspirated from the pleural space, the diagnosis is either a chylothorax or 32 a cholesterol effusion; an empyema may simulate this appearance. Centrifugation of the fluid will separate a lipid effusion from an empyema; in the lipid effusion, the supernatant will remain white, while the empyema supernatant is clear. A yellowish-green fluid suggests rheumatoid pleurisy,33 and green pleural fluid is virtually diagnostic of a biliopleural fistula. 34 If a central venous catheter has migrated extravascularly into the mediastinum, the pleural fluid will be similar to the infusate (white if lipid is 35 being infused).
31

The character of the fluid can also suggest a diagnosis. If pus is aspirated, an empyema is established. Pus is determined by the gross appearance of the fluid, which is a thick, viscous, yellow-white, opaque fluid. If the pus has a putrid odor, an anaerobic infection is confirmed. If the pleural fluid appears to contain debris, rheumatoid pleurisy with exfoliation of rheumatoid nodules from the visceral pleural surface into the pleural space is a likely cause.28 If pleural fluid smells like ammonia, the diagnosis is urinothorax, which is caused by obstructive uropathy.36

Exudates vs Transudates The next deductive step in evaluating patients with a pleural effusion is to determine whether the effusion is an exudate or transudate. Patients with transudative effusions have normal pleurae and limited diagnostic possibilities, and the effusion is formed because hydrostatic pressures increase or oncotic pressure decreases, or a combination of the two (Table 5).1,25 Rare causes of transudates develop from an extravascular origin (pleural effusions of extravascular origin) and include urinothorax, duropleural fistula, peritoneal dialysis, and extravascular migration of a central venous catheter with saline infusion. 35 In contrast, exudative effusions have a more extensive differential diagnosis, as these effusions are caused by inflammation, infection, malignancy, and lymphatic abnormalities (Table 6).25,27

It is important to distinguish between a transudate and an exudate accurately. The detection of an exudative pleural effusion warrants additional diagnostic testing in most patients to determine the underlying cause. Conversely, the patients clinical presentation is usually sufficient to determine the cause of a transudative effusion without further testing. An exudative effusion is defined by the presence of a high concentration of large, molecular weight proteins compared with transudates. Although several tests have been suggested to separate transudates from exudates, the two tests that appear to have the highest specificity and sensitivity are the pleural fluid-to-serum total protein ratio and the pleural fluid LDH (lactate dehydrogenase) compared 37 with the upper limits of the normal serum LDH. The total protein ratio can be used because the pleural fluid and serum values are related; however, because there is no correlation between pleural fluid and serum LDH, the previously mentioned ratio should be used instead.38 If the pleural fluid-to-serum total protein ratio is >0.50 or the pleural fluid LDH is >0.67 of the upper limit of normal serum LDH, the fluid is most likely an exudate. If both total protein and LDH ratios are 0.50 and 0.67, respectively, the fluid 1 is most likely a transudate. Subsequent validation studies using pooled data from several primary investigations reported that the above criteria have a sensitivity of 98% and a specificity of 74% in identifying an exudative pleural effusion.1 However, the closer the values are to the cut-point, the fluid is likely to be either a transudate or an exudate, while the further the value is from the cut-point, the more likely the fluid is to be a transudate or an exudate. It should be recognized that treatment of the patient may affect pleural fluid values; for example, in patients with congestive heart failure treated with diuretics, either the protein or LDH ratio may be increased from the transudative range prior to diuresis to the exudative range following diuresis. A more recent study by Joseph and colleagues,39 using receiver operating characteristic curve analysis, found that the test with the highest sensitivity and specificity for separating transudates and exudates was a pleural fluid LDH compared with the upper limit of the normal serum LDH ratio of 0.82 (AUC = 0.89); this higher ratio decreases the incidence of false exudates.

Total Protein and LDH The absolute concentrations of total protein and LDH may be of some diagnostic value. For example, a tuberculous pleural effusion rarely has a total protein concentration <4.0 g/dL, while total protein concentrations in malignancy and parapneumonic effusions have a wide range.18 A total protein concentration of 7.0 g/dL suggests Waldenstrm macroglobulinemia, 40 multiple myeloma, 41 or a cholesterol effusion.32 A pleural fluid LDH concentration greater than three times the upper limit of normal (>1,000 IU/L) for the serum LDH is typically seen only in complicated parapneumonic effusions 42,43 44 45 or empyema, rheumatoid pleurisy, or pleural paragonimiasis.

Nucleated Cells 1,25,46 The total nucleated cell count is rarely diagnostic but may provide useful information. Most exudates have >1,000 nucleated cells/ L, while transudates have a few hundred cells per microliter. Pleural fluid nucleated cell counts >10,000/ L are seen most commonly with parapneumonic effusions, acute pancreatitis, subdiaphragmatic abscesses, liver, hepatic and splenic abscesses, and splenic infarction. Nucleated cell counts >10,000/ L at times can occur with pulmonary infarction, PCIS, and lupus pleuritis. When the nucleated cell count is >50,000/ L, the differential is limited to a complicated parapneumonic effusion and empyema and rarely with acute pancreatitis and pulmonary infarction. Chronic exudates typically have nucleated cell counts <5,000/ L, as seen with a tuberculous pleural 9 47 effusion and malignancy. When pus (empyema fluid) is aspirated from the pleural space, the nucleated cell count may be as low as a few hundred cells because the neutrophils have undergone autolysis from the harsh pleural environment of acidosis and low oxygen tension.

The predominant cell population is determined by the type of pleural injury and the timing of thoracentesis in relation to the acute pleural injury. The acute response to any pleural injury, whether infectious, immunologic, or malignant, is the attraction of neutrophils to the pleural space, initiated by the chemotaxin interleukin-8.48,49 Within 72 h following the cessation of acute pleural injury, mononuclear cells enter the pleural space from the peripheral blood and become the predominant cells. 50 This macrophage predominance is subsequently replaced by lymphocytes in effusions that persist for >2 weeks. Therefore, a neutrophil-predominant exudate is the rule when the patient presents shortly after the onset of symptoms, ie, acute bacterial pneumonia, acute pulmonary embolism with infarction, and acute pancreatitis. In contrast, with the insidious onset of disease, as with malignancy and tuberculosis, a lymphocyte-predominant exudate is found. Transudative effusions are never neutrophilpredominant (usually <5% neutrophils); and when the neutrophil percentage is >10 to 15%, a second diagnosis is likely. Transudative effusions are mononuclear-cell-predominant, a combination of lymphocytes, macrophages, and mesothelial cells. When the lymphocyte population is >80% of the total nucleated cells, the differential diagnosis of the exudate is narrowed to those entities shown in Table 7.25,27 All the diagnoses listed in the table typically have lymphocyte populations 80%; the lymphocyte population can occasionally be less but virtually never <50%. In contrast to lymphoma, only ~60% of patients with metastatic carcinoma to the pleura 51 have a majority of lymphocytes, usually 50 to 75% of the total nucleated cell count. A lymphocytepredominant exudate is the most appropriate indication for percutaneous pleural biopsy, which is the most sensitive (~75 to 80%) diagnostic test for a tuberculous pleural effusion,9 with the exception of thoracoscopy.

Pleural fluid eosinophilia (PFE) is defined as a pleural fluid eosinophil count >10% of the total nucleated cell count. It appears that bone marrow eosinophils are attracted to the pleural space primarily by 52 27 interleukin-5. Causes of PFE are shown in Table 8. The most common causes are pneumothorax and hemothorax. Eosinophilicpleuritis is a common, early finding in patients requiring thoractomy or thoracoscopy for treatment of spontaneous pneumothorax.53 In contrast to the rapid movement of eosinophils into the pleura and pleural fluid in pneumothorax, eosinophils do not appear in the pleural space for 1 to 2 weeks following hemothorax.54Furthermore, PFE is associated with peripheral blood eosinophilia following trauma that does not clear until the pleural fluid resolves.55 About 30% of patients with BAPE have PFE, which, at times, totals 50% of the nucleated cells.56,57 It was previously thought

that PFE virtually excluded the diagnosis of malignancy; however, recent studies have shown that the prevalence of malignancy is similar in both eosinophilic and noneosinophilic pleural effusions.58 Pleural fluid macrophages, which originate from the blood monocyte, are of no diagnostic value. 50 Mesothelial cells are exfoliated into normal pleural fluid in small numbers. Although common in transudative effusions and some exudates, mesothelial cells are rarely found in tuberculous pleural effusion, because of the extensive pleural involvement from the hypersensitivity reaction to tuberculin 54,59 proteins that inhibits mesothelial shedding. The paucity of mesothelial cells is also the typical finding in other inflammatory processes, such as empyema, chemical pleurodesis, rheumatoid pleuritis, and chronic malignant effusions.54 A large number of plasma cells in pleural fluid suggests pleural involvement with multiple myeloma, while a small number of plasma cells isnondiagnostic and has been observed in several nonmalignant conditions. 54 A few basophils are occasionally found in pleural fluid and are of no clinical significance. However, when basophils represent >10% of the nucleated cells, leukemic involvement of the pleura is 54 likely.

Pleural Fluid pH and Glucose A limited number of diagnoses are associated with pleural fluid acidosis, which is defined as a pleural fluid pH <7.30 measured with a blood gas analyzer (Table 9).60,61 Pleural fluid pH decreases as a result of high metabolic activity of intrapleural cellular constituents (empyema) or because the pleural membrane is abnormal and impairs the efflux of hydrogen ions and carbon dioxide from the pleural space into the circulation (rheumatoid pleurisy).61 The presence of a low pleural fluid pH is not only helpful diagnostically but has prognostic implications, as well. In the only study of acid-base characteristics of normal human pleural fluid, the pleural fluid pH was measured at 7.64, 0.23 pH units 62 greater than the simultaneously measured blood pH. We, and others, have measured an alkaline pH of 63,64 ~7.60 in normal animals. It appears that a bicarbonate gradient between pleural fluid and blood explains the alkaline pH of normal pleural fluid. Human transudative effusions have a pleural fluid pH

ranging from 7.45 to 7.55. The vast majority of exudative effusions have pH values that range from 7.45 to 7.30, while a small number of exudates are associated with pleural fluid acidosis (pH <7.30).

Pleural fluid pH provides prognostic information and helps guide therapy for parapneumonic and malignant effusions. A metaanalysis examined the prognostic value of pleural fluid pH in patients with

parapneumonic effusions. Using receiver operating characteristic curve analysis, it was observed that the pleural fluid pH was lower in patients who had a complicated course and required pleural space drainage. The primary studies recommended various cut-points for a complicated effusion that varied from 7.10 to 7.30. While no single pH value can be used as a definitive cut-point for classifying patients as having complicated or uncomplicated parapneumonic effusions, the pleural fluid pH serves as adjunctive information that should be combined with the clinicians judgment in determining the need for pleural space drainage. A low pleural fluid pH (<7.30) is seen in approximately one-third of patients presenting with a malignant pleural effusion. 66 A low pleural fluid pH in this setting suggests that the patients will have positive cytologic findings on initial testing, a shorter survival time, and a poorer response to chemical pleurodesis than those with a pH >7.30.66 The explanation for these findings is related to the advanced stage of malignant pleural metastasis that inhibits the end products of glycolysis from exiting the pleural 61 space. Finding a low pH is not an absolute contraindication to pleurodesis but should be considered in the decision when contemplating treatment. In the normal physiologic state, pleural fluid and blood glucose concentrations are equivalent, as glucose is of low molecular weight and moves from blood to pleural fluid by simple diffusion across the endothelial and mesothelial membranes. The same diseases associated with low pleural fluid pH also have a low pleural fluid glucose concentration, which is defined as <60 mg/dL or a pleural fluid/serum 67 glucose ratio of <0.5. As noted in Table 10, the only diagnoses associated with a glucose of 0 mg/dL are chronic rheumatoid pleurisy and empyema.

65

Amylase An increased pleural fluid amylase level, defined as either a value greater than the upper limits of normal serum or a pleural fluid/serum amylase ratio >1.0, is found with pancreatic disease,6871 esophageal rupture,72-74 and malignancy.70,75,76 Both acute pancreatitis and a pancreaticopleural fistula 75 can cause an amylase-rich pleural effusion, the latter often having amylase levels >100,000 IU/L. An increased pleural fluid amylase concentration occurs in 10 to 14% of patients with a malignant pleural effusion. On isoenzyme analysis, the amylase in these malignant effusions is virtually all salivarytype. 68,70,76 Adenocarcinoma of the lung is the most common malignancy associated with a salivary amylase-rich pleural effusion followed by adenocarcinoma of the ovary.70,76 It has been demonstrated that adenocarcinoma cells secrete a salivary-like isoamylase.75 Esophageal rupture is also characterized by the presence of pleural fluid salivary isoamylase. Triglycerides and Cholesterol There are two types of lipid pleural effusions, chylothorax and a cholesterol effusion, also called a chyliform effusion or pseudochylothorax. Table 11 provides detailed information that clearly demonstrates the differences between these two lipid effusions. A chylothorax represents leakage of chyle into the pleural space from the thoracic duct or one of its major tributaries.77,78 The most common cause of a chylothorax is lymphoma, most commonly non-Hodgkin lymphoma. A cholesterol effusion is a chronic form of lung entrapment that is most commonly associated with rheumatoid pleurisy and tuberculosis. 32,79 The diagnosis of chylothorax is highly likely when the pleural fluid triglyceride concentration is >110 mg/dL and is highly unlikely if the triglyceride concentration is <50 mg/dL.77 With intermediate triglyceride values (50 mg/dL to 110 mg/dL), the presence of chylomicrons should be determined; if chylomicrons are detected, the diagnosis of chylothorax is established.

Cytologic Examination Pleural fluid cytology has a widely variable diagnostic yield, ranging from 40 to 90% of patients with a known malignancy.80,81 The reasons for this variability include the following: (1) the effusion may be paramalignant, which is defined as a pleural effusion associated with a known malignancy; however,

malignant cells cannot be detected in pleural fluid or pleural tissue and the effusion results from another mechanism (ie, atelectasis)82; (2) the type of tumor (high adenocarcinoma positivity and low in Hodgkin lymphoma)83,84 ; (3) the number of specimens submitted (yields tend to increase with additional specimens, owing to exfoliation of fresher cells) 84; (4) the stage of pleural involvement (the more advanced stage, the higher the diagnostic yield); and (5) the interest and expertise of the cytopathologist. Flow Cytometry Flow cytometry can be helpful in the diagnosis of lymphoma of the pleura. Flow cytometry can specifically define lymphocyte surface markers.85 Therefore, it can define clonality of a population of lymphocytes to determine whether the cells are from T- or B-cell lineage. Therefore, flow cytometery is most helpful in patients with a lymphocyte-predominant pleural effusion when lymphoma is in the differential diagnosis.

Conclusion
For pleural fluid analysis to be most valuable, the clinician must have a solid prethoracentesis diagnosis based on a patients history, physical examination, laboratory findings, and radiographic imaging. With a presumptive clinical diagnosis in concert with a good working knowledge of pleural fluid analysis, a definitive or confident clinical diagnosis can be determined in 95% of patients. Without this approach, the clinician may be left with an unacceptable number of problematic or undiagnosed pleural effusions.

Poststudy Questions

References 1. 2. 3. 4. 5. 6. 7. 8. 9. Light RW. Pleural diseases. 4th ed. Baltimore, MD: Lippincott, Williams and Wilkins, 2001 Carr DT, Mayne JG. Pleurisy with effusion in rheumatoid arthritis, with reference to the low concentration of glucose in pleural fluid. Am Rev Respir Dis 1962; 85:345-350 Epler GR, McLoud TC, Gaensler EA. Prevalence and incidence of benign asbestos pleural effusion in a working population. JAMA 1982; 247:617-622 Good JT Jr, King TE, Antony VB, et al. Lupus pleuritis: clinical features and pleural fluid characteristics with special reference to pleural fluid antinuclear antibody titers. Chest 1983; 84:714-718 Stelzner TJ, King TE, Antony VB, et al. The pleuropulmonary manifestations of the postcardiac injury syndrome. Chest 1983; 84:383-387 Sahn SA, Heffner JE. Approach to the patient with pleurisy. In: Kelley WN, ed. Textbook of internal medicine. 2nd ed. Philadelphia, PA: JB Lippincott Company, 1991; 1887-1890 Sahn SA. Drug-induced pleural disease. In: Camus P, Rosenow E, eds. Drug- induced iatrogenic lung disease. London, UK: Hodder Arnold. In press Hopkins HU. Principles and methods of physical diagnosis. 3rd ed. Philadelphia, PA: WB Saunders, 1965; 202230 Berger HW, Mejia E. Tuberculous pleurisy. Chest 1973; 63:88-92

10. Alptekin F. An epidemic of pleurisy with effusion in Bitlis, Turkey: a study of 559 cases. US Armed Forces Med J 1958; 9:1-11 11. Walker WC, Wright V. Pulmonary lesions in rheumatoid arthritis. Medicine 1968; 47:501-519 12. Maher GG, Berger HW. Massive pleural effusions: malignant and non-malignant causes in 46 patients. Am Rev Respir Dis 1972; 105:458-460 13. Lieberson M. Diagnostic significance of the mediastinal profile in massive unilateral pleural effusions. Am Rev Respir Dis 1963; 88:176-180

14. Heller RM, Janower ML, Weber AL. The radiological manifestations of malignant pleural mesothelioma. Am J Roentengol 1971; 108:53-59 15. Lorch DG, Sahn SA. Pleural effusions due to diseases below the diaphragm. SemRespir Med 1987; 9:75-85 16. Rabin CB, Blackman NS. Bilateral pleural effusions: its significance in association with a heart of normal size. J Mt Sinai Hosp 1957; 24:45-53 17. Janower ML, Blennerhassett JB. Lymphangitic spread of metastatic cancer to the lung: a radiologic-pathologic classification. Radiology 1971; 101:267-273 18. Taylor JR, Ryu J, Colby TV, et al. Lymphangioleiomyomatosis: clinical course in 32 patients. N Engl J Med 1990; 323:1254-1260 19. Fine NL, Smith LR, Sheedy PF. Frequency of pleural effusions in mycoplasma and viral pneumonias. N Engl J Med 1970; 283:790-793 20. Soskel NT, Sharma OP. Pleural involvement in sarcoidosis: case presentation and detailed review of the literature. SeminRespir Med 1992; 13:492-514 21. Horowitz ML, Schiff M, Samuels J, et al. Pneumocystis carinii pleural effusion: pathogenesis and pleural fluid analysis. Am Rev Respir Dis 1993; 148:232-234 22. Maguire R, Fauchi AS, Doppman JL, et al. Unusual radiographic features of Wegeners granulomatosis. Am J Roentengol 1978; 130:233-238 23. Gumbs RV, McCauley DI. Hilar and mediastinaladenopathy in septic pulmonary embolic disease. Radiology 1982; 142:313-315 24. Rubin SA. Radiographic spectrum of pleuropulmonarytularemia. Am J Roentengol 1978; 131:277-281 25. Sahn SA. The pleura. Am Rev Respir Dis 1988; 138:184-234 26. Collins TR, Sahn SA. Thoracentesis: complications, patient experience and diagnostic value. Chest 1987; 91:817-822 27. Sahn SA. The diagnostic value of pleural fluid analysis. SeminRespirCrit Care Med 1995; 16:269-278 28. Nosanchuk JS, Naylor B. A unique cytologic picture in pleural fluid from patients with rheumatoid arthritis. Am J ClinPathol 1968; 50:330-335 29. Skedros D, Cass S, Hirsch B, et al. Beta-2 transferrin assay in clinical management of cerebral spinal fluid and perilymphatic fluid leaks. J Otolaryngol 1993; 22:341-344 30. Joseph J, Sahn SA. Thoracic endometriosis syndrome: new observations from an analysis of 110 cases. Am J Med 1996; 100:164-170 31. Doerr C, Allen M, Nichols FR, et al. Etiology of chylothorax in 203 patients. Mayo ClinProc 2005; 80:867-870 32. Coe JE, Aikawa JK. Cholesterol pleural effusion. Arch Intern Med 1961; 108:763-774 33. Lillington GA, Carr DT, Mayne JG. Rheumatoid pleurisy with effusion. Arch Intern Med 1971; 128:764-768 34. Strange C, Allen ML, Freedland P, et al. Biliopleural fistula as a complication of percutaneous biliary drainage: experimental evidence for pleural inflammation. Am Rev Respir Dis 1988; 137:959-96l 35. Sahn SA. Pleural effusions of extravascular origin (PEEVO). Clin Chest Med 2006; 27:285-308 36. Stark DD, Shanes JG, Baron RL, et al. Biochemical features of urinothorax. Arch Intern Med 1982; 42:15031511 37. Light RW, MacGregor I, Lusinger PC, et al. Pleural effusion: the diagnostic separation of transudates and exudates. Ann Intern Med 1972; 77:507-513 38. Heffner JE, Brown LK, Barbieri C. Diagnostic value of tests that discriminate between exudative and transudative pleural effusions. Chest 1997; 111:970-979 39. Joseph J, Badrinath P, Basran G, et al. Is the pleural fluid transudate or exudate? A revisit of the diagnostic criteria. Thorax 2001; 56:867-870 40. Winterbauer RH, Riggins RC, Griesman FA, et al. Pleuropulmonary manifestations of Waldentrmsmacroglobulinemia. Chest 1974; 66:368-375 41. Rodriguez JN, Pereira A, Martinez JC, et al. Pleural effusion and multiple myeloma. Chest 1994; 105:622-624 42. Light RW, Girard WM, Jenkinson SG, et al. Parapneumonic effusions. Am J Med 1980; 69:507-511 43. Potts DE, Levin DC, Sahn SA. Pleural fluid pH and parapneumonic effusions. Chest 1976; 70:328-331 44. Pettersson T, Klockers M, Helmstrom PE. Chemical and immunological features of pleural effusions: comparison between rheumatoid arthritis and other diseases. Thorax 1982; 37:354-361

45. Johnson JR, Falk A, Iber C, et al. Paragonomyacis in the United States: a report of 9 cases in Hmong immigrants. Chest 1982; 82:168-171 46. Light RW, Erozan YS, Ball WC. Cells in pleural fluid: their value in differential diagnosis. Arch Intern Med 1973; 132:854-860 47. Pettersson T, Riska H. Diagnostic value of total and differential leukocyte counts in pleural effusions. Acta Med Scand 1981; 210:129-135 48. Antony VB, Repine JE, Sahn SA. Experimental models of inflammation in the plerual space. In: Chretien J, Bignon J, Hirsch A, eds. The pleura in health and disease. New York, NY: Marcel Dekker, 1985; 253-266 49. Antony VB, Godbey SW, Kunkel SL, et al. Recruitment of inflammatory cells to the pleural space chemotatic cytokines, IL8, and monocyte chemotatic peptide 1 in human pleural fluid. J Immunol 1993; 151:7216 -7223 50. Antony VB, Sahn SA, Antony AC, et al. Bacillus calmette-guerin stimulated neutrophils release chemotaxins for monocytes in rabbit pleural spaces and in vitro. J Clin Invest 1985; 76:1514-1521 51. Yam LT. Diagnostic significance of lymphocytes in pleural effusions. Ann Intern Med 1967; 66:972-982 52. Nakamura Y, Ozaki T, Kamei T, et al. Factors that stimulate proliferation and survival of eosinophils in eosinophilic pleural effusion: relationship to granulocytes/macrophage colony stimulating factor, interleuken -5, and interleuken-5. Am Rev Cell MolBiol 1993; 8:605-611 53. Askin FB, McCann BG, Kuhn C. Reactive eosinophilicpleuritis. Arch Pathol Lab Med 1997; 101:187 -191 54. Spriggs AI, Boddington MM. The cytology of effusions. 2nd ed. New York, NY: Grunn and Stratton, 1968 55. Maltais F, Laberge F, Cormier Y. Blood hyper-eosinophilia in the course of posttraumatic pleural effusion. Chest 1990; 98:348-351 56. Mattson SB. Monosymptomatic exudative pleurisy in persons exposed to asbestos dust. Scand J Respir Dis 1997; 56:263-272 57. Hillerdal G, Ozesmi M. Benign asbestos pleural effusion: 73 exudates in 60 patients. Eur J Respir Dis 1987; 71:113-121 58. Rubins JB, Rubins HB. Etiology and prognostic significance of eosinophilic pleural effusions: a prospective study. Chest 1996; 110:1271-1274 59. Hurwitz S, Leiman G, Shapiro C. Mesothelial cells in pleural fluid: TB or not TB? S Afr Med J 1980; 57:938-939 60. Good JT, Taryle DA, Maulitz RM, et al. The diagnostic value of pleural fluid pH. Chest 1980; 78:55-59 61. Sahn SA. Pleural fluid pH in the normal state and in diseases affecting the pleural space. In: Chretien J, Bignon J, Hirsch A, eds. The pleura in health and disease. New York, NY: Marcel Dekker, 1985; 253-266 62. Yamada S. Uberdieseroseflussigkeit in der plerachohle der gesunden menschen. Z GesExp Med 1933; 90:343 348 63. Sahn SA, Willcox ML, Good JT Jr, et al. Characteristics of normal rabbit pleural fluid: physiologic and biochemical implications. Lung 1979; 156:63-69 64. Rolf LL, Travis DM. Pleural fluid-plasma bicarbonate gradients in oxygen-toxic and normal rats. J Physiol 1973; 224:857-861 65. Heffner JE, Brown LK, Barbieri C. Pleural fluid chemical analysis in parapneumonic effusions: a meta-analysis. Am J RespirCrit Care Med 1995; 51:1700-1708 66. Sahn SA, Good JT Jr. Pleural fluid pH in malignant effusions: diagnostic, prognostic, and therapeutic implications. Ann Intern Med 1988; 108:345-349 67. Sahn SA. Pathogenesis and clinical features of diseases associated with a low pleural fluid glucose. In: Chretien J, Bignon J, Hirsch A, eds. The pleural in health and disease. New York, NY: Marcel Dekker, 1985; 267-285 68. Light RW, Ball WC Jr. Glucose and amylase in pleural effusions. JAMA 1973; 225:257-260 69. Kaye MD. Pleuropulmonary complications of pancreatitis. Thorax 1968; 23:297-306 70. Joseph J, Viney S, Beck P, et al. A prospective study of amylase-rich pleural effusion with special reference to amylase isoenzyme analysis. Chest 1992; 102:1455-1459 71. Rockey DC, Cello JP. Pancreaticopleural fistula: a report of 7 cases and review of the literature. Medicine 1990; 69:332-334 72. Maulitz RM, Good JT Jr, Kaplan RL, et al. The pleuropulmonary consequences of esophageal rupture in an experimental model. Am Rev Respir Dis 1979; 120:363-367 73. Abbott OA, Mansour KA, Logan WD Jr, et al. A traumatic so-called "spontaneous" rupture of the esophagus: a review of 47 personal cases with comments on the new method of surgical therapy. J ThoracCardiov Sur 1970; 59:67-83

74. Sherr HP, Light RW, Morson MH, et al. Origin of pleural fluid amylase in esophageal rupture. Ann Intern Med 1972; 76:75-76 75. Ende N. Studies of amylase activity in pleural effusions and ascites. Cancer 1960; 13:283-287 76. Kramer MR, Sepero RJ, Pitchenik AE. High amylase in neoplasm-related pleural effusions. Ann Intern Med 1989; 110:567-569 77. Staats BA, Ellefson RD, Budahn LL, et al. The lipoprotein profile of chylous and non-chylous pleural effusions. Mayo ClinProc 1980; 55:700-704 78. Teba L, Dedhia AT, Bowen R, et al. Chylothorax review. Crit Care Med 1985; 13:49-52 79. Hamm H, Pfalzer B, Fabell H. Lipoprotein analysis in a chyloform pleural effusion: implications for pathogenesis and diagnosis. Respiration 1991; 58:294-300 80. Grunze H. The comparitive diagnostic accuracy, efficiency, and specificity of cytologic techniques used in the diagnosis of malignant neoplasm and serous effusions of the pleuroperitoneal cavities. ActaCytol 1964; 8:150164 81. Jarvi OH, Kunnas RJ, Laitio MT, et al. The accuracy and significance of cytologic cancer diagnosis of pleural effusions. ActaCytol 1972; 16:152-157 82. Sahn SA. Malignant pleural effusions. SeminRespir Med 1987; 9:43-53 83. Naylor B, Schmidt RW. The case for exfoliator cytology of serous effusions. Lancet 1964; 1:711-712 84. Melamed MR. The cytologic presentation of malignant lymphomas in related diseases and effusions. Cancer 1963; 16:413-431 85. Moriarty AT, Wiersema L, Snyder W, et al. Immunophenotyping of cytologic specimens by flow cytometry. DiagnCytopathol 1993; 9:252