Parapneumonic Effusions and Empyema

Pleural effusions complicate up to 57% of bacterial pneumonias. They vary in severity, ranging from uncomplicated effusions to empyema. Some require only antibiotics and observation, while others require chest tube drainage or even surgery. Since the mortality of empyema is as high as 15%, rapid recognition of such patients is essential.

The general treatment of parapneumonic effusions requires knowledge of:

• Pathogenesis of parapneumonic effusions
• Appropriate imaging
• Pleural fluid analysis
• Assessment of the risk for complications
• Choosing appropriate interventions.

Pathogenesis

Pleural fluid progresses through an exudative phase (uncomplicated or simple parapneumonic effusions) to a fibrinopurulent stage (complicated para­pneumonic effusions), eventually resulting in pus formation (empyema). Infected pleural space may result in residual pleural thickening. Pleural thickening after treatment occurs in up to 13% of patients and the risk factors in its development include purulent fluid and delayed (>15 days) resolution of effusions. However, pleural thickening is not always associated with impaired lung function.

Although classified as distinct stages, it is important to realize that these stages actually represent a continuum of disease rather than rigid classes. The importance of understanding these stages lies in the prognostic value that each stage is associated with and thus the treatment implications that go with each stage (Table 12-1).

Enlarge 

Uncomplicated or Simple Parapneumonic Effusion

The first stage is that of uncomplicated parapneumonic effusion. In this stage, an exudative effusion forms during the first 72 hours related to increased permeability of the vascular and pleural membranes and the restorative capacity of the pleural space is exceeded. These effusions are usually predominantly neutrophilic in nature, typically with total neutrophil counts in excess of 10,000/mL and will usually disappear with resolution of the pneumonia. Since they are self-limiting, they do not require chest tube drainage or other invasive measures.

Complicated Parapneumonic Effusions

Some uncomplicated parapneumonic effusions will progress to become complicated effusions. This may occur if there is persistent bacterial contamination of the pleural space. With persistent infection, there is an increase in the number of neutrophils in the pleural space as well as a corresponding pleural fluid acidosis. This acidosis is characterized by a pleural fluid pH in the range of 7.1 to 7.3 and is caused by anaerobic glucose metabolism by both neutrophils and bacteria.

As neutrophils persist and subsequently lyse within the pleural space, the pleural fluid lactate dehydrogenase (LDH) increases, often in excess of 1,000 IU/L. Persistent inflammation leads to deposition of a dense layer of fibrin on both the visceral and parietal pleura that may subsequently lead to loculations and adhesions. Importantly, cultures of the pleural fluid at this point may be sterile, since bacteria can be cleared relatively rapidly while still initiating this fibrinopurulent stage. The fibrinopurulent stage may last anywhere from 3 to 7 days.

Empyema

In the third stage, empyema develops, with the accumulation of pus within the pleural space. Bacteria may be evident on Gram’s stain of the pleural fluid. However, a positive culture or Gram’s stain is not always present since patients may be infected with anaerobic organisms that are difficult to isolate. Similarly, patients may have been on antibiotics prior to pleural fluid collection, significantly lowering the diagnostic yield of subsequent cultures.

Finally, many times empyema will be loculated and the pleural fluid aspirate may represent only a sterile inflammatory area adjacent to an infected loculation (Figure 12-1). As the infection persists, additional neutrophils are recruited, resulting in even more severe pleural fluid acidosis as well as a marked decrease in pleural fluid glucose. Eventually, a thick pleural peel develops, encasing and entrapping the lung as it organizes. This organizational phase of empyema usually occurs over 2 to 3 weeks. Importantly, not all organisms are equally likely to cause empyema. The most common organisms causing empyema are shown in Table 12-2. In a United Kingdom study of patients with complicated parapneumonic effusion and empyema (Multicentre Intrapleural Streptokinase trial [MIST1]), the most common causes of community-acquired pleural infection were the Streptococcus anginosus group (including Streptococcus intermedius, Streptococcus constellatus and Streptococcus mitis [28%]), S pneumoniae (14%) and staphylococci (12%). In the same series, parapneumonic effusions caused by hospital-acquired infections (following pneumonia, trauma, or pleural procedures) were most often caused by methicillin-resistant Staphylococcus aureus (MRSA) (27%), staphylococci (22%), enterobacteria (20%), or enterococcus (12%). In another study, pleural fluid from 434 patients with pleural infections underwent standard culture and a screen for bacteria by amplification and sequencing of bacterial 16S ribosomal RNA gene. Approximately 50% of community-acquired infections were streptococcal and 20% included anaerobic bacteria. Approximately 60% of hospital-acquired infections included bacteria frequently resistant to antibiotics (MRSA, 25%; Enterobacteriaceae, 18%; Pseudomonas spp, 5%; enterococci, 12%). The site of acquisition of pleural infection is important in determining the microbial etiology.

Enlarge 

Enlarge 

Imaging of Pleural Effusions

Initial imaging of pleural effusions is usually accomplished with a chest radiograph. Adjunctive measures include computed tomography (CT) of the chest and ultrasonography. These techniques usually assist in identifying loculations or concurrent chest masses or to help guide thoracentesis. The chest radiograph remains the best initial choice and can provide substantial information rapidly.

On a plain upright posterior-anterior and lateral chest film, up to 75 mL of effusion can occupy the subpulmonic space without spillover. As the amount of fluid increases to >75 mL, the lateral chest radiograph demonstrates obliteration of the posterior costophrenic sulcus. Blunting of the costophrenic angle usually occurs in the presence of approximately 200 mL of pleural fluid, although the plain film can be normal with up to 500 mL fluid. Once a minimum of 175 mL is present, the lateral costophrenic angle is obscured as well. At 1,000 mL, the effusion will typically reach approximately the fourth intercostal space anteriorly.

Decubitus radiographs with the patient in both lateral decubitus positions are useful to further evaluate the effusion. Free-flowing effusions that are >10 mm thick on a lateral decubitus film are usually amenable to thoracentesis. For quantification purposes:

• Small effusions usually will be <15 mm thick on decubitus view.
• Moderate effusions will be 15 to 45 mm thick.
• Large effusions will be >45 mm thick.

While a chest radiograph is the primary tool for the initial evaluation of effusions, it does have several important limitations, primarily related to patient positioning, subpulmonic effusions and loculated effusions. Mobile effusions may layer along the posterior chest in supine patients, leading to underestimation of the amount of pleural fluid. As the effusion layers along the posterior thorax, it produces a diffuse veil-like effect throughout the entire chest that may be missed or interpreted as air-space disease. This will appear as a diffuse haziness throughout all lung fields. Distinguishing this from parenchymal lung disease may be difficult. Features that suggest a posterior effusion rather than parenchymal disease include the absence of air bronchograms and visible pulmonary vessels throughout the area of increased opacity created by the effusion.

Subpulmonic effusions may elevate the lung bases while producing minimal costophrenic-angle blunting. If the fluid is free flowing, a lateral decubitus film will serve to readily identify this problem. Features that suggest a subpulmonic effusion include shifting of the apex of the diaphragm laterally, diaphragmatic inversion, making the diaphragm appear concave and separation of the lung from the stomach bubble >2 cm.

Loculated effusions are the result of adhesions. Adhesions may arise from preexisting disease or may be a consequence of the pneumonia itself. Because the fluid is not free flowing, loculated effusions will not necessarily layer on lateral decubitus films and may be mistaken for parenchymal masses or consolidation. Loculation is frequently encountered when hemothorax, empyema, chylothorax and tuberculous infection cause the effusion. Features that suggest a loculated effusion include an obtuse angle between the pleural mass and the chest wall, a smooth border and homogenous density.

While specific features may help to identify some subpulmonic and loculated effusions, in many cases these findings may be absent or not specific enough to allow an accurate assessment; additional imaging with either CT and /or ultrasound is usually necessary.

CT has several advantages over routine chest radiographs. Chest CT has increased sensitivity for small effusions when compared with regular chest radiographs. As little as 2 to 10 mL of pleural fluid can be detected on chest CT. In supine patients and in those with subpulmonic effusions, fluid is easily visualized, even in the posterior and subpulmonic locations. In those with loculated effusions or underlying parenchymal processes, CT allows visualization and differentiation of the underlying parenchyma from the pleural effusion. CT also images the pleura more accurately, allowing more precise measurement of pleural thickness.

Finally, CT may help to distinguish empyema from lung abscess. In the fibrinopurulent and organizing stages of empyema, CT will demonstrate strong enhancement of the visceral and parietal pleurae, producing the so-called split pleura sign. In addition, with empyema formation, there is usually concurrent thickening of the pleura during the fibrinopurulent and organizing phases that may only be apparent on CT.

Ultrasonography is the other imaging modality commonly employed to evaluate pleural effusions. The intercostal spaces are used as sonographic windows. A small footprint probe (either linear or phased array) allows the easiest intercostal access and probe selection is a balance between spatial resolution and penetration. Ultrasound easily identifies free or loculated effusions and facilitates differentiation of loculated effusions from any underlying mass. This is especially useful in guiding thoracentesis procedures. With ultrasound guidance, effusions that would normally be too small to undergo thoracentesis can be drained. Loculated effusions are also easier to drain under ultrasound guidance, allowing the clinician to identify and drain the largest collections within any given effusion. Thus ultrasound-guided thoracentesis is a relatively quick and easy technique to collect pleural fluid for diagnostic assessment. However, CT facilitates catheter placement for empyema drainage or biopsy of pleural lesions.

Pleural Fluid Analysis

In the setting of a pleural effusion of unknown cause, pleural fluid analysis is critical to rapidly identifying a specific etiology. Pleural fluid analysis in combination with clinical judgment allows an accurate diagnosis in 75% of patients. Even when nondiagnostic, pleural fluid analysis can be useful in excluding certain diagnoses. Thus, clinical decision making is influenced by pleural fluid analysis in >90% of cases. While observation of a pleural effusion may be warranted in other clinical settings, such as when there is heart failure (HF) in the setting of pneumonia, early thoracentesis is usually warranted to rule out empyema and to obtain cultures to help guide antibiotic therapy. Therefore, early thoracentesis to obtain pleural fluid for analysis should be considered whenever there is >10 mm of fluid on lateral decubitus films.

Initial fluid analysis should include observation of the gross appearance of the fluid, Gram’s stain, culture, pleural fluid LDH, protein, glucose, pH, cell count and differential. Other tests that may be useful include pleural fluid amylase, triglycerides, cholesterol and adenosine deaminase. Table 12-3 summarizes the conditions associated with each of these tests.

Parapneumonic effusions are usually exudative in nature (Table 12-4). If thoracentesis of a pleural effusion indicates a transudative effusion in the setting of clinical pneumonia, this should lead to consideration of differential diagnosis or other concurrent diseases that may cause a transudative effusion, such as HF. In patients with bilateral effusions, one side may rarely demonstrate a transudative pattern while the other side demonstrates an exudative pattern.

Bacterial parapneumonic effusions are typically neutrophilic exudative effusions. Lymphocytic exudative effusions are characteristic of tuberculous pleural effusions as well as malignant effusions. In the setting of pneumonia of unknown cause, the presence of a lymphocytic exudative effusion should prompt consideration of tuberculosis (TB) or concurrent neoplastic disease.

Tuberculous effusions are usually serous, with total protein >5.0 g/dL (77% of cases) and with anywhere from 2000 to 8000 nucleated WBCs. Most of these WBCs are lymphocytes; in 90% of TB cases, >60% will be lymphocytes. The exception to this is that early in the course of disease, polymorphonuclear (PMN) leukocytes may predominate with lymphocytes becoming more common as the disease progresses. Features that make TB less likely include pleural fluid eosinophilia and the presence of >5% mesothelial cells. While the measurement of pleural fluid adenosine deaminase, lysozyme and interferon-γ levels may be suggestive, they are not diagnostic of pleural TB.

Unfortunately, the diagnostic yield of acid-fast bacilli (AFB) stain and culture on thoracentesis is also low and therefore the absence of AFB on pleural fluid smear does not exclude TB. Skin testing was previously used since a purified protein derivative (PPD) is positive in 69% to 100% of tuberculous pleural effusions. However, a negative PPD is possible secondary to circulating mononuclear cells that suppress sensitized T lymphocytes in the peripheral blood and skin but do not suppress the inflammation in the pleural space. Similarly, a positive PPD only establishes exposure to TB but does not confirm active disease. Interferon gamma releasing assays like QuantiFERON-TB Gold have replaced routine skin testing (PPD) but have been less well studied with respect to the diagnosis of pleural tuberculosis. Because of the limited utility of pleural fluid smear and culture, the diagnosis of pleural TB is best made by a combination of pleural tissue and fluid histology and culture. The yield of these individual tests is shown in Table 12-5. Closed pleural biopsy establishes the diagnosis of TB in 90% to 95% of cases.

Enlarge 

Enlarge 

Enlarge 

Assessing the Risk for Complicated Effusions

The importance of pleural fluid analysis in parapneumonic effusions does not lie in the classification of effusions as transudative or exudative. The ultimate goal of pleural fluid analysis is to distinguish those effusions that are most likely to develop into complicated effusions and/or empyema. In addition to pleural fluid analysis, other factors must also be considered in this assessment of risk. These include the virulence of the pathogen as well as host factors. Careful consideration of all of these factors helps determine appropriate interventions.

In the evaluation of most parapneumonic effusions, the most useful diagnostic tests are pH, LDH, protein, glucose, cell count and Gram’s stain. In a recent meta-analysis, pleural fluid pH was found to be the most useful test in determining which parapneumonic effusions should undergo drainage. A pleural pH <7.2 represents the threshold for consideration of chest tube drainage. However, other diseases, including malignancy, TB, rheumatoid arthritis, lupus pleuritis, and urinothorax may be associated with low pleural pH. Depending on other concurrent clinical factors, a pH <7.2 should prompt consideration of a chest tube. A pH of <7.1 strongly supports the need for facilitated chest tube drainage.

While diagnostically useful, pleural fluid LDH, protein, glucose and cell count are not sufficiently specific to help determine the need for chest tube placement. However, any positive Gram’s stain should lead to drainage since this situation is likely to either represent or develop into an empyema. Similarly, if the fluid is grossly purulent, chest tube drainage is warranted.

The virulence of the pathogen must also be considered in assessing the risk for complications. The specific pathogen is frequently unknown at the time of the initial evaluation. If the pathogen is known, however, this may provide additional insight into the likelihood of complications. Organisms such as S aureus, gram-negative bacilli, and anaerobes are more likely to cause empyemas. Similarly, although S pneumoniae is among the most common causes of pneumonia, it rarely causes an empyema. In these cases, a more conservative approach may be warranted.

Host factors and preexisting disease also play a significant role in assessing the risk for complications. In elderly and immunocompromised patients, as well as in those with preexisting lung disease, the spectrum of pathogens causing bacterial pneumonia is significantly different from that in young and otherwise healthy individuals. The former groups of patients are more likely to have gram-negative pneumonias that are associated with a higher mortality. A more aggressive strategy of intervention may be warranted in this population.

In a prospective study involving 1,269 patients admitted with community-acquired pneumonia (CAP), multivariate logistic regression identified albumin <30 g/L (adjusted odds ratio [AOR] 4.55), sodium <130 mmol/L (AOR 2.70), platelet count >400 × 10⁹/L (AOR 4.09), CRP >100 mg/L (AOR 15.7), and a history of alcohol abuse (AOR 4.28) or intravenous (IV) drug use (AOR 2.82) as independently associated with development of complicated parapneumonic effusion or empyema. A history of chronic obstructive pulmonary disease (COPD) was associated with decreased risk (AOR 0.18).

Interventions

The therapeutic options depend upon the stage of the parapneumonic effusion as well as upon an assessment of the risk of complications developing as described above. In uncomplicated effusions, antibiotics alone are usually sufficient. Serial radiographs are warranted to document improvement of the effusion. Any clinical deterioration or increase in size of the effusion should prompt reevaluation and possibly repeat thoracentesis.

Antibiotic selection in the setting of a parapneumonic effusion should, whenever possible, be based on the cause of the underlying pneumonia. Antibiotic selection should be guided by results of blood and pleural fluid culture and sensitivity results. Consideration should also be given to the possibility of anaerobic infection if there is a particularly large effusion or in the setting of empyema. Since anaerobic organisms are frequently difficult to isolate in clinical laboratories, empiric coverage in this setting may be appropriate. Additional anaerobic coverage can be obtained using clindamycin, metronidazole, β-lactam/β-lactamase combinations, or imipenem. Almost all antibiotics will penetrate into the pleural space. The one important exception is the inactivation of aminoglycosides at low pleural pH. Thus, in the setting of a gram-negative pleural space infection, it is probably warranted to avoid using aminoglycosides as primary therapy. In the case of hospital-acquired empyema, treatment for both gram-positive and gram-negative aerobic organisms, as well as anaerobes, is necessary. Possibilities include carbapenems, antipseudomonal penicillins, moxifloxacin, or third-generation cephalosporins with clindamycin or metronidazole. Vancomycin/linezolid should be added in case of MRSA. Patients with pleural fluid infection are frequently malnourished and, therefore, appropriate nutritional support is also important.

Pleural Fluid Drainage

Complicated parapneumonic effusions are more difficult to manage and tend to have a variable response when treated with antibiotics alone. Although some patients can be managed with antibiotics alone, others will develop complications and require chest tube placement. Although there are no prospective clinical trials, there is a bias that early pleural fluid drainage in these cases may improve recovery and decrease length of stay. Certainly, patients with grossly purulent pleural fluid, a positive Gram’s stain, or pleural fluid pH <7.1 should undergo early drainage. The optimum treatment for patients with exudative effusions with a pH in the intermediate zone of 7.1 to 7.3 is more difficult to determine. Options include early drainage or a trial of antibiotic therapy with a repeat chest radiograph and thoracentesis in 24 to 48 hours to assess response. The choice is often not clear and must include consideration of other factors, including comorbidities and the virulence of the pathogen as described earlier.

Current indications for drainage are the aspiration of frankly purulent pleural fluid, the identification of organisms on pleural fluid Gram’s stain or culture, or a pleural fluid pH <7.2 in the clinical setting of a pneumonic illness.

Once the decision has been made that pleural space drainage is needed, there are a number of options available. If the pleural fluid collection is not loculated, then either a large-bore chest tube or a radiographically guided small-bore tube may be used. Ultrasonography and CT can both be used to guide drain placement, although ultrasound guidance is the most commonly used technique. The complication rate following image-guided tube insertion is low, with pneumothoraces occurring in approximately 3% of cases. Importantly, post-drainage imaging is critical to confirm complete pleural fluid drainage. Tube failure has been seen when it was not followed with adequate imaging and resulted in inadequate drainage. The optimal size of the chest tube and duration of drainage remain controversial. Small bore (12 -14 bore French) drains are easier to insert, more comfortable and adequate for the drainage of infected pleural space in most cases. Drain removal is considered when the output falls to <150 mL daily for 2 days in the setting of clinical and radiologic improvement.

Intrapleural Fibrinolytics

If the effusion is loculated, treatment options include chest tube placement with or without intrapleural fibrinolytics and/or thoracoscopy. Intrapleural fibrinolytic therapy consists of regular chest tube placement followed by instillation oftPA (tissue plasminogen activator) and DNase at a dose of 10 mg and 5 mg respectively, usually administered twice daily for 3 days. These agents are utilized because fibrin strands bridge pleural membranes and transform free-flowing pleural fluid in the early exudative stage of empyema into loculations and fibrous peels in the late organized stage. Patients who progress through the intermediate fibrinopurulent stage commonly do not benefit from chest tube drainage and require surgical intervention to drain infected pleural fluid. Intrapleural fibrinolytic (tPA/DNase) agents are thought to disrupt the fibrin loculations, enhance chest tube drainage and obviate the need for surgical drainage.

One early study showed a clear lack of benefit for streptokinase in terms of mortality at 3 and 12 months. Thus, the existing data favor early use of VATS for patients with fibrinopurulent empyemas that cannot be managed by chest tube drainage, and fibrinolytic therapy.

Surgery

Thoracoscopy represents an alternative treatment for loculated effusions. The procedure is somewhat limited by the fact that it can become prolonged if there are extensive adhesions and it is thus best suited for patients with early disease. In those patients who are unable to have adequate drainage or lung reexpansion with thoracoscopy, ventilator-associated tracheobronchitis (VATS) decortication may be required. One retrospective series found that up to 30% of empyemas managed with thoracoscopy ultimately required open decortication. Unfortunately, large prospective comparisons are still lacking and thus specific recommendations have not yet been issued regarding the indications for thoracoscopy.

Case series seem to suggest that thoracoscopy may be superior to intrapleural fibrinolytic therapy. Early thoracoscopy in these series resulted in improved rates of complete pleural drainage, a reduction in chest tube duration and decreased length of hospital stay. Importantly, there is a paucity of data directly comparing these procedures, making specific recommendations of limited value. In addition, identification of patients who may benefit from early surgical intervention represents a further difficulty and presently there are no robust clinical predictors of outcome in empyema.

Despite adequate drainage procedures, an empyema cavity may fail to resolve. In these cases, the empyema cavity often remains because the underlying lung is unable to reexpand because of visceral pleural fibrosis. The thickness of the visceral pleura, as well as the size of the cavity, determines how likely this is to occur. Those cases that demonstrate failure of pleural apposition after chest tube placement with thickened visceral pleura are candidates for surgical drainage and or decortication. VATS, done robotically, is often employed.

Decortication represents the most invasive treatment for organized empyema cavities. Typically, this procedure is reserved for those who not only demonstrate pleural thickening but who also remain clinically ill. After allowing the pleural space to organize into a pleural peel over a period of several weeks, open resection of the pleural space is undertaken under general anesthesia. All fibrous tissue and purulent material is resected at this time. This procedure is associated with significant perioperative morbidity and is often too difficult for debilitated patients to withstand. In contrast to open thoracostomy drainage, however, open decortication allows a more rapid recovery, with a decreased length of hospital stay.

Summary

Early evaluation of parapneumonic effusions is critical to correctly managing pneumonia. Early thoracentesis should be considered in all patients with significant parapneumonic effusions. The goal of the diagnostic assessment is to identify those effusions that are most likely to develop into complicated effusions or empyemas. A low pH, positive Gram’s stain, or purulent-appearing pleural fluid are all indicators of empyema or complicated effusions and should prompt tube drainage.

In patients with complicated parapneumonic effusions that are borderline, a trial of antibiotics with repeat thoracentesis may be warranted. Serial radiographs are warranted in all cases to confirm a proper response to drainage. In those patients with effusions that fail to drain or those with loculated effusions, VATS may be needed. Finally, in those patients in whom thoracoscopy has failed, open thoracostomy or decortication may be an option. An algorithm summarizing this diagnostic approach is shown in Figure 12-2.

Enlarge