Published: August 2010
The pleural cavity contains a relatively small amount of fluid, approximately 10 mL on each side.1 Pleural fluid volume is maintained by a balance between fluid production and removal, and changes in the rates of either can result in the presence of excess fluid, traditionally known as a pleural effusion.
The classic work of Light and colleagues in 1972 demonstrated that 99% of pleural effusions could be classified into two general categories: transudative or exudative (see Light's criteria under “Diagnosis”).2 A basic difference is that transudates, in general, reflect a systemic perturbation, whereas exudates usually signify underlying local (pleuropulmonary) disease.
Pleural disease, specifically pleural effusions, is one of the more common clinical problems encountered by the internist. Estimates of the incidence of pleural effusions vary, with some estimating an annual incidence of up to 1 million in the United States. The more common causes of transudative effusions are congestive heart failure and hypoalbuminemic states (e.g., cirrhosis), and those of exudative effusions are malignancy, infection (e.g., pneumonia), and pulmonary embolism.
The accumulation of pleural fluid can usually be explained by increased pleural fluid formation or decreased pleural fluid absorption, or both. Increased pleural fluid formation can result from elevation of hydrostatic pressure (e.g., congestive heart failure), decreased colloid osmotic pressure (e.g., cirrhosis, nephrotic syndrome), increased capillary permeability (e.g., infection, neoplasm), passage of fluid through openings in the diaphragm (e.g., cirrhosis with ascites), or reduction of pleural space pressures (e.g., atelectasis). Decreased pleural fluid absorption can result from lymphatic obstruction or from elevation of systemic venous pressures resulting in impaired lymphatic drainage (e.g., superior vena cava [SVC] syndrome).
The presence of fluid in the normally negative-pressure environment of the pleural space has a number of consequences for respiratory physiology. Pleural effusions produce a restrictive ventilatory defect and also decrease the total lung capacity, functional residual capacity, and forced vital capacity.3 They can cause ventilation-perfusion mismatches and, when large enough, compromise cardiac output.
The differential diagnosis of pleural effusions is briefly summarized in Boxes 1 and 2.
|Box 1: Select Causes of Transudates
|Congestive heart failure|
|Box 2: Select Causes of Exudates
|Connective tissue disease|
|Drugs (e.g., amiodarone)|
|Infection (bacteria, viruses, fungi, tuberculosis, or parasites)|
Many patients are asymptomatic on the discovery of a pleural effusion. When present, symptoms are usually due to the underlying disease process. Pleuritic chest pain indicates inflammation of the parietal pleura (because the visceral pleura is not innervated and thus not sensitive to pain). Other symptoms include dry, nonproductive cough and dyspnea. Physical examination findings that can reveal the presence of an effusion are reduced tactile fremitus, dull or flat note on percussion, and diminished or absent breath sounds on auscultation. It is also important to note the presence of other clues that can point to the cause of the effusion (e.g., signs of heart failure, breast masses).
The posteroanterior and lateral chest radiographs are still the most important initial tools in diagnosing a pleural effusion (Figure 1). Free pleural fluid gravitates to the more dependent portions of the space; thus, most fluid collects around the inferior surface of the lung posteriorly, spilling out laterally and anteriorly as the amounts increase. About 50 mL of fluid is needed to be visible on the lateral radiograph as a meniscus posteriorly, and when more than 500 mL is present, the meniscus usually obscures the entire hemidiaphragm.4 The lateral decubitus films help in differentiating free fluid from loculated fluid (that which is confined by fibrous pleural adhesions).
Ultrasound is useful both as a diagnostic tool and as an aid in performing thoracentesis. Its major advantage over conventional radiography is its ability to differentiate between solid and liquid components and thus assist in identifying pleural fluid loculations. It is also valuable in detecting subpulmonic or subphrenic pathology.
Cross-sectional computed tomography (CT) (Figure 2) helps distinguish anatomic compartments more clearly (e.g., the pleural space from lung parenchyma). This modality is useful as well in distinguishing empyema (split pleura sign) from lung abscess, in detecting pleural masses, and in outlining loculated fluid collections.5
Ideally, the workup of a pleural effusion begins with a diagnostic thoracentesis followed by classification of the pleural fluid into either a transudate or an exudate. In 1972, Light and coworkers developed the currently accepted benchmark in classifying pleural fluid2:
Pleural fluid is classified as an exudate if it meets any one of the aforementioned criteria. Conversely, if all three characteristics are not met, then the fluid is classified as a transudate. Following these guidelines, the original study of Light and colleagues2 had a diagnostic sensitivity of 99% and specificity of 98% for an exudate. In more recent years, as noted by Tarn and Lapworth,6 a number of studies used modifications to Light's criteria but had poorer diagnostic accuracy.
Although the reason is unclear, cholesterol concentration is higher in exudates than in transudates. Various studies have looked at the usefulness of cholesterol measurements alone, as a fluid-to-serum ratio, or in combination with LDH, with cutoffs ranging from 45 to 60 mg/dL. Currently, pleural fluid cholesterol measurements, on their own, probably reduce misclassifications but cannot be used as a substitute to measurements of protein and LDH.
One of the limitations of the Light criteria is that they can misidentify some transudates as exudative effusions (e.g., in patients with heart failure who undergo diuretic treatment). Roth and colleagues used the serum-effusion albumin gradient (serum albumin concentration minus effusion albumin concentration) with a cutoff of 12 g/L (exudates if below that level, transudates if above), and obtained a specificity of 100% as compared with 72% with Light's criteria.7 However, use of this marker alone can result in misclassification of many exudates as well.
Very low glucose levels (<25 mg/100 mL), although not pathognomonic, are seen in a few diseases. Rheumatoid arthritis, tuberculosis, empyema, and tumors or malignancy with extensive involvement of the pleura are most commonly associated with very low glucose levels.
Elevated pleural fluid amylase is seen with pancreatitis and esophageal rupture and in approximately 10% of malignant effusions.
Normal pleural fluid pH has been estimated to be around 7.64. Good and colleagues noted that a pH of less than 7.30 suggests the presence of an inflammatory or infiltrative process.8 These processes can include parapneumonic effusions, empyema, malignancy, connective tissue diseases, tuberculosis, and esophageal rupture. Urinothorax is peculiar in that it is the only cause of a low pH transudative effusion. According to the current American College of Chest Physicians (ACCP) consensus statement on the treatment of parapneumonic effusions,9 pH is the preferred pleural fluid chemistry test (determined using a blood gas analyzer) for classifying the category of a parapneumonic effusion for subsequent management (See “Treatment and Outcomes” and Table 1).
|Pleural Anatomy||Pleural Fluid Bacteriology||Pleural Fluid Chemistry||Need for Drainage|
|Minimal effusion (<10 mm on lateral decubitus view); free-flowing||Cx and GS unknown||pH unknown||No|
|Small to moderate effusion (>10 mm to <one half of hemithorax on lateral decubitus view); free-flowing||Negative Cx and GS||pH > 7.20||No|
|Large effusion (>one half of hemithorax on lateral decubitus view) or loculated fluid or thickened pleura||Positive Cx or GS||pH < 7.20||Yes|
|Any||Pus||pH < 7.0||Yes|
Cx, culture; GS, Gram stain.
Adenosine deaminase levels tend to be higher in tuberculous pleural effusions than in other exudates. A level greater than 70 U/L is highly suggestive of tuberculous pleuritis, whereas a level less than 40 U/L virtually rules out this diagnosis. Other pleural diseases where high adenosine deaminase levels may be seen are rheumatoid pleuritis and empyema.10
The use of an Abrams needle to obtain specimens from the parietal pleura has become less common with the increasing availability of improved serum markers and thoracoscopy. At present, a needle biopsy of the pleura is used mainly to diagnose tuberculous pleuritis when other markers (e.g., adenosine deaminase) are negative.
Invasive techniques for the diagnosis of pleural effusions have gained more popularity with the advent of video-assisted technology. Thoracoscopy offers the advantages of visual evaluation of the pleura, direct tissue sampling, and therapeutic intervention (e.g., dissecting loculations and pleurodesis). Medical thoracoscopy (performed by pulmonologists under conscious sedation) and video-assisted thoracoscopic surgery (VATS), which is performed by surgeons under general anesthesia, are indicated for diagnosing pleural effusions that have remained undiagnosed despite previous, less-invasive tests (e.g., thoracentesis).10
Drainage of a pleural effusion is indicated in complicated parapneumonic effusions or empyema (see Table 1), for symptomatic relief of dyspnea, and to evaluate underlying lung parenchyma. The current guidelines proposed by the ACCP for the treatment of parapneumonic effusions9 categorize the risk of poor outcome as well as the need to drain the effusion based on the pleural space anatomy, pleural fluid bacteriology (culture and Gram stain), and pleural fluid chemistry (pH).
Therapeutic thoracentesis may be repeated if indicated; however, more definitive therapy (e.g., pleural sclerosis; see following) is usually needed to treat recurrent symptomatic pleural effusions. At any one time, no more than 1 L to 1.5 L of fluid should be removed (unless pleural space pressure is monitored) to avoid re-expansion pulmonary edema and post-thoracentesis shock. Supplemental oxygen is probably of benefit as well, because post-thoracentesis decreases in arterial oxygenation have also been reported, the magnitude and duration of which roughly correlate with the amount of fluid removed.
The use of a sclerosing agent to produce a chemical serositis and subsequent fibrosis of the pleura is indicated in recurrent symptomatic malignant effusions. Agents such as talc, doxycycline, bleomycin, and quinacrine have been used. All fluid must be drained initially and that full expansion of the underlying lung (usually via a tube thoracostomy) is essential before proceeding with sclerosis. Failure of treatment is usually due to the inability to approximate the pleural surfaces during administration of the agent. With proper technique, however, doxycycline sclerosis has been reported to be 80% to 90% effective.
Randomized, controlled trials have shown that fibrinolytics (urokinase or streptokinase instilled via a tube thoracostomy) improved fluid drainage and chest radiograph findings significantly, and it was an effective treatment for managing parapneumonic effusions.11,12
The inadequacy of conventional drainage strategies has led the ACCP consensus panel to recommend video-assisted thoracoscopic surgery (VATS) and thoracotomy as acceptable approaches to managing patients with complicated pleural effusions. Parietal pleurectomy and decortication of the visceral pleura are definitive procedures with excellent response rates. Morbidity and mortality rates remain high, however, and the patient's general medical condition, expected long-term prognosis, and baseline lung function should be considered before proceeding with surgery.13
The pleura is involved in a majority of patients with systemic lupus erythematosus (SLE) at some time during the course of their disease. These pleural effusions are usually small and bilateral, and the most common symptom is chest pain. Previous studies have shown that the finding of lupus erythematosus cells and high antinuclear antibody titers in pleural fluid have a high specificity but are not particularly sensitive in diagnosing this condition. Therefore, routine use of these tests is not currently recommended. SLE effusions are usually responsive to corticosteroids.
Pleural effusions occur less commonly in patients with rheumatoid arthritis and, in contrast to SLE effusions, they occur more commonly in men. A striking characteristic of rheumatoid effusions is their low glucose level (<25 mg/dL). The measurement of rheumatoid factor in pleural fluid is also not useful, because this can be elevated in other inflammatory states. In contrast to SLE, there is little evidence that corticosteroids are beneficial in treating rheumatoid pleurisy, probably because the natural history of this disease is much more variable.
The pleura is involved in neoplastic disease more commonly through metastasis than through primary tumors. Lung and breast cancers are the leading causes of metastatic disease to the pleura. Other less common causes are hematologic (e.g., lymphoma, leukemia), ovarian, and gastrointestinal tumors. Cytologic examination of the pleural fluid is positive in more than 50% of cases with pleural involvement. Tumor markers (e.g., carcinoembryonic antigen [CEA] are not specific enough to be recommended routinely in establishing the diagnosis. Immunocytometry has been used to establish the diagnosis of lymphoma and has been helpful in cases of idiopathic effusions when conventional techniques were nondiagnostic.14
Leakage of chyle from a disruption of the thoracic duct leads to a chylothorax. Common causes of this condition are listed in Box 3. Although the gross appearance of milky fluid usually indicates the diagnosis, the best way to ascertain this diagnosis is by measuring pleural fluid triglyceride levels (Figure 3). A triglyceride level greater than 110 mg/dL confirms the diagnosis, whereas a level less than 40 mg/dL excludes the diagnosis. The finding of chylomicrons in the effusion (using electrophoresis) also establishes the diagnosis.
|Box 3: Common Causes of Chylothorax
|Subclavian venous thrombosis|
|Trauma (including surgery)|
Treatment of a chylous effusion is aimed at preventing the complications of malnutrition due to the continuous loss of protein, fat, and electrolytes. Conservative measures include shifting to a medium-chain triglyceride diet to minimize the accumulation of fluid and total parenteral nutrition. Definitive treatment modalities include thoracic duct ligation or pleuroperitoneal shunt implantation. Pleurodesis is not very effective due to the anti-inflammatory characteristics of chyle.
Whenever the gross appearance of pleural fluid is bloody, a hematocrit level should be determined. Hemothorax is considered present when the pleural fluid hematocrit is greater than 50% of the peripheral blood hematocrit. Hemothorax most commonly results from chest trauma. Nontraumatic hemothorax, although uncommon, must alert the clinician to the possibility of malignancy or pulmonary embolism. Treatment of this condition requires immediate chest tube thoracostomy and, if bleeding persists (drainage >200 mL/hr), subsequent thoracotomy.
Approximately one half of patients who undergo coronary artery bypass grafting develop pleural effusions. The precise pathophysiology of this postoperative occurrence is unclear, but it is probably related to pleural trauma during surgery or bleeding into the pleural space. Light and coworkers15 divided these large effusions into two categories: those that occur within 30 days of surgery and those that occur after. Within 30 days of surgery, the fluid is bloody, eosinophilic, and easily resolvable with drainage (thoracentesis). After 30 days, the fluid is clear yellow and predominantly lymphocytic, but these effusions are difficult to manage because they often recur. In either case, it is easy to distinguish these effusions from those caused by congestive heart failure, because the former are usually exudative.
Air between the right lung and chest wall (in the pleural space) is termed pneumothorax (Figure 4). Box 4 lists the classification of pneumothoraces. Common causes of pneumothorax include trauma, iatrogenic factors (e.g., thoracentesis, mechanical ventilation), chronic obstructive pulmonary disease, infection, and malignancy.
|Box 4: Classification of Pneumothorax
|Traumatic or iatrogenic|
|Spontaneous (without antecedent cause):
The incidence of primary spontaneous pneumothorax is higher in men younger than 40 years, and the relative risk rises with heavy smoking. Secondary spontaneous pneumothorax is a more serious condition, because it further compromises an already abnormal lung function. Most secondary spontaneous pneumothoraces are related to chronic obstructive pulmonary disease or infection (e.g., Pneumocystis jiroveci). Trauma-related pneumothorax can result either in an open (to the atmosphere) pneumothorax or a closed (tension) pneumothorax, in which intrapleural pressures commonly exceed atmospheric pressures.
Box 5 summarizes the currently adopted ACCP guidelines for the treatment of spontaneous pneumothorax.16 Traumatic pneumothorax usually requires placement of a thoracostomy tube until the air leak resolves. The ACCP consensus statement also recommends surgical intervention (thoracoscopy with bullectomy and a procedure to produce pleural symphysis) in preventing the recurrence of secondary pneumothoraces.16
|Box 5: Management of Spontaneous Pneumothorax|
|Primary Spontaneous Pneumothorax
Stable Patients with Small Pneumothoraces
|Stable Patients with Large Pneumothoraces|
|Unstable Patients with Large Pneumothoraces|
|Secondary Spontaneous Pneumothorax|
© 2002 The Cleveland Clinic Foundation.
The spectrum of pleural diseases with asbestos exposure ranges from the classic pleural plaques to effusions and malignancy. Pleural plaques are fibrous lesions found mostly on the parietal pleura after more than 20 years of exposure. They are considered markers of clinically relevant asbestos exposure and can occur without any evidence of asbestos-related lung disease17; furthermore, they are not considered to be premalignant lesions. Small, benign effusions are common and are often the earliest manifestations (within the first 20 years) of exposure, and the pathologic findings are nonspecific. On examination of the pleural fluid, however, the presence of mesothelial cells with atypical features makes it difficult to distinguish these benign effusions from effusions due to mesothelioma. Therefore, benign asbestos pleural effusions are exudates that can represent a diagnostic problem when other signs of asbestos exposure have not yet appeared.
Wagner and associates recognized the association of mesothelioma and asbestos in 1960.18 Most patients are middle-aged and have a significant history of asbestos exposure. The diagnosis is often suggested by the history of cough and pleuritic chest pain as well as chest CT results and findings of elevated hyaluronic acid levels in pleural fluid. The diagnosis is confirmed by tissue biopsy through thoracoscopy or thoracotomy. The prognosis of patients with mesothelioma is generally poor (<1 year survival after diagnosis), and the management involves multimodality therapy.
The immunologic impairment in AIDS leads to a variety of infectious and neoplastic processes. Infectious complications include the development of bacterial parapneumonic effusions and empyema. In developing countries, tuberculous involvement of the pleura is common. P. jiroveci, although a common cause of pneumonia in patients with AIDS, is rarely a cause of pleural effusions. However, P. jiroveci has been associated with pneumothorax in this patient population, so much so that the development of an unexplained spontaneous pneumothorax in a person infected with human immunodeficiency virus should prompt a search for P. jiroveci infection.19 Pleural effusions can also occur with Kaposi's sarcoma and non-Hodgkin's lymphoma, and responses to treatment for these disease entities have been poor.