Empyema Thoracis

Introduction and Classifcation

Parapneumonic effusions can be classi ed as either uncomplicated or complicated. Uncomplicated, or simple, parapneumonic effusions represent effusions devoid of infection, derangements of fuid chemistry, or loculation that typically resolve with appropriate antibiotic care. Complicated parapneumonic effusions, however, are characterized by often identi able infection, derangements in pleural chemistry, and the development of loculations with persistent clinical or biochemical signs of sepsis [1]. Finally, empyema is de ned by the presence of purulence in the pleural space.

While the primary cause of empyema remains parapneumonic effusions, other etiologies include bronchogenic carcinoma, esophageal rupture, infected congenital cysts of the airway and esophagus, mediastinitis, and blunt or penetrating trauma [2]. Post-operative empyema is a distinct entity, with frequency decreasing to 1% after 2000, with a progressively decreasing mortality currently estimated at 11.6% [3]. Primary empyema, de ned as a direct invasion

D. Shore (*) · J. W. Toth

Division of Pulmonary, Allergy, and Critical Care Medicine, Department of Medicine, Penn State Milton S. Hershey Medical Center,

Hershey, PA, USA

e-mail: dshore@pennstatehealth.psu.edu; jtoth@pennstatehealth.psu.edu

of the pleural space with bacteria, often manifests as bacterial translocation from the oropharynx or via hematogenous spread [4]. Similar to the incidence of parapneumonic effusions, the incidence of primary empyema seems to be increasing [5].

In addition to its substantial mortality, upwards of 20–30% of patients with empyema will require surgical drainage of the pleural space [2]. Citing the signi cant burden surgery can represent, Steven Sahn and Richard Light in 1989 popularized the now well-known phrase, “The sun should never set on a parapneumonic effusion.” [6] They argued an aggressive posture with regard to pleural fuid sampling and drainage, preferring, “Too many chest tubes placed than obtain a result requiring thoracotomy, embryectomy, and decortication if drainage had not been done or was delayed.” [6] Recognition of failure antibiotic therapy in patients with pneumonia and persistent septic signs should prompt the investigation for a parapneumonic effusion. Delays in pleural drainage, misdiagnosis, inappropriate antibiotic selection, and chest tube mispositioning are al known risk factors for treatment failure [1]. As such, The British Thoracic Society in 2010 published guidelines codifying the need for early consultation by a chest physician or thoracic surgeon in all patients undergoing evaluation for and drainage of a pleural infection [1].

To aid in the appropriate selection of patients requiring pleural drainage, the American College

of Chest Physicians published a consensus statement in 2000 classifying patients with parapneumonic effusions into categories based off risk for poor outcomes. Referred to as the A, B, C classi-cation, parapneumonic effusions were characterized based on three variables: anatomic characteristics, bacteriology, and chemistry [7]. The anatomy is based on the size of the effusion, whether it is free fowing, and whether the parietal pleural is thickened. Bacteriology is based off positive or negative gram stain and culture, or frank purulence seen on sampling. Finally, chemistry­ relies primarily on pH measurement greater or less than 7.2 (alternatively, a pleural fuid glucose greater or less than 60 mg/dL). Based on these three characteristics, a parapneumonic effusion can be characterized into one of four categories: very low (category 1), low (category 2), moderate (category 3), and high (category 4, with categories 3 and 4 requiring drainage).

Historical Perspectives

From around 3000 BCCE, empyema was rst described in the Edwin Smith Surgical Papyrus, a collection of trauma and medical cases classi ed by a physician who is thought to be Imhotep, the chief vizier of Pharaoh Zoser [1, 8, 9]. Later, Hippocrates described “peripneumonia,” or empyema, in much greater detail. In addition to suggesting an etiology where, “Affuent pus runs from the head to the lung,” [10], he is often quoted describing its morbidity, where if, “Persons who become affected with empyema after pleurisy, if they get clear of it in 14 days from the breaking of it, escape the disease; but if not, it passes into phthisis.” [11] Hippocrates even goes as far as describing methodology for enteral feedings, thoracotomy, and pleural irrigation with instillation of antiseptics, a combination of water and olive oil, for 12 h in pursuit of empyema resolution [10].

Prior to 1918, the treatment for empyema was remarkably similar to what was rst described by Hippocrates thousands of years prior. Updates in management focused on early surgical treatment

via thoracotomy, extensive decortication, and marsupialization of the pleural space to allow continued external drainage (Eloesser fap) [12]. Mortality for this approach was unacceptably high, ranging from 30.2% to 84%, and after facing an epidemic Group A streptococcus pneumonia and resultant empyema in 1917, The United States Army formed The Empyema Commission [12, 13]. Under Major Evans Graham, the Commission standardized several interventions, including early drainage via tube thoracostomy, repeated sterilizing lavage of the pleural space, and a heavy emphasis on nutrition and physical therapy, dropping mortality at Camp Lee from 40% to 4.3% [12, 13]. These interventions, rst widely implemented during the Spanish Infuenza of 1918, remain the framework of modern treatment of empyema.

Incidence

In 2010, the British Thoracic Society described a combined incidence of parapneumonic effusion and empyema of 80,000 in the United States and United Kingdom [1]. This represents a progressive increase in the incidence of empyema over the past 20 years. In 2008, Carlos Grijalva et al. used national hospitalization data to compare incidence of parapneumonic effusion and empyema admissions between 1996 and 2008. He found a start increase, with parapneumonic empyema incidence increasing to 5.98 per 100,000 peoples in 2008 compared to 3.04 in 1996, nearly a twofold increase [14]. The greatest burden was seen in adults greater than 65 years of age, with an annual incidence of 9.94 per 100,000. This increase in parapneumonic empyema hospitalizations was seen in across all age groups, with a 1.9-fold increase seen in children <18 years of age, 1.8-fold increase in the age group 18–39, and 1.7-fold increase in adults ≥ than 65 years of age. While the in-hospital case fatality ratio for empyema hospitalization did decrease from 8.0% to 7.2%, the overall rate of fatal empyema-­ associated hospitalizations increased by 1.8-fold from 0.24 per 100,000 to 0.43 per 100,000.

Epidemiology

Empyema has been described as a disease with a bimodal age distribution, with peaks in both childhood and in older adults [4]. Per Grijalva, however the greatest burden of disease was seen in ages 40–64 years of age, followed by those ≥ than 65 years of age [14]. Men are affected twice as often as women [14–16]. Risk factors associated with empyema include cardiac disease, chronic lung disease, diabetes, immunosuppression including corticosteroid use, gastro­ -­esophageal refux, excessive alcohol intake, and intravenous (IV) drug use [1, 16].

Parapneumonic effusion and empyema increase the mortality compared to that just pneumonia. Compared to patients presenting to the emergency department with pneumonia, patients with pneumonia and concomitant simple parapneumonic effusion were more likely to be admitted, have twice the length of stay, and a higher 30-day mortality with an odds ratio of 2.6. Furthermore, typical pneumonia severity guidelines tend to underestimate mortality seen in patients with pneumonia and simple effusions, with eCURB65 in one study predicting 7% mortality compared to 14% actual mortality. Morbidity in patients with empyema is also high, with the average length of stay ranging from 15 to 30 days [14, 15].

Overall mortality associated with empyema is high, with Rahman et al. in MIST-2 nding a 3-month mortality of 8.3%, and a 12-month mortality of 12.4% [15]. Mortality can be infuenced by a variety of factors, including infective etiology, age, and nutritional status [14, 17]. Grijalva et al. found patients with staphylococcal-related empyema had both longer hospitalization and the highest in-hospital case fatality ratio [14]. In a secondary analysis of First Multicenter Intrapleural Sepsis Trial (MIST-1), Maskell et al. con rmed this nding, further elucidating a signi cant difference in mortality in empyema associated with community acquired pneumonia (CAP) versus hospital acquired pneumonia (HAP), with a 1-year mortality of 17% in CAP and 47% in HAP (relative risk 4.24) [18]. This was largely infuenced, just as Grijalva described,

by bacteriology; HAP-related empyema was associated with S. aureus and gram-negative bacteria (largely Enterobacteriaceae), with a 44% and 45% 1-year mortality, respectively [18]. Maskell et al. further elucidated that patients with streptococcal infection generally have the best prognosis, with 83% 1-year survival, while those with staphylococcal, enterobacterial, and mixed aerobic infections had the worst, with a 1-year survival of only 45% [18].

Age is also strongly associated with mortality. While identifying risk factors for poor outcome, Rahman et al. identi ed that patients older than 70, compared to those <50 years of age, had an odds ratio of 25.63 for mortality [17]. When compounded with like staphylococcal empyema for example, being older than 65 cause a precipitous increase in in-hospital case fatality from an average 7.2 across all age groups and causes of infection of to 21.8% [14].

Pathogenesis

Parapneumonic effusions and empyema are exudative with a predominance of neutrophils [19]. Thus, initial sampling, usually by thoracentesis, will determine the exudative or transudative nature of the effusion. In the 1970s, Light established differential criteria based on levels of lactate dehydrogenase (LDH) and protein found in pleural fuid compared with the patient’s serum. Light’s criteria de ne an exudate as having any one of the following characteristics: (1) pleural fuid protein to serum protein ratio greater than 0.5; (2) pleural fuid LDH to serum LDH ratio greater than 0.6; (3) pleural fuid LDH greater than two thirds of the upper limit of normal serum LDH [20].

Parapneumonic effusions progress through three stages within a continuous spectrum. Therst, or exudative, stage is characterized by fuid with a relatively low LDH, normal glucose, and normal pH. The pleural fuid at this stage is sterile, and microbiologic studies are negative. This accumulation of sterile fuid is caused by a rapid outpouring of fuid into the pleural space due to both increased pulmonary interstitial fuid and

increased permeability of the capillaries in the pleural space [1]. Proinfammatory cytokines including tumor necrosis alpha (TNF-a) and interleukin 8 (IL-8) drive this increased capillary vascular permeability [1]. This stage is ideally treated by antibiotics, drainage, and observation since the prognosis is favorable at this point. When untreated or inappropriately treated, the second, or brinopurulent, stage occurs. Characterized by bacterial invasion, formation of loculations, and derangements of pleural chemistry because of an infammatory milieu, positive bacterial studies, glucose level below 60 mg/dL, pH less than 7.2, and high LDH characterize the fuid in this stage. There is progressive loculation due to increased brin deposition, cellular debris, and white blood cells with the ultimate formation of limiting brin membranes. At this stage, drainage is indicated, and will become more dif cult as more loculations form. If the effusion is allowed to persist, the third, or organization, stage occurs as broblasts grow into the exudative brin sheet coating the visceral and parietal pleura. This process results in the formation of an inelastic membrane or pleural “peel” which encases the lung and renders it functionless. At this point, surgery with decortication is indicated. While the rate of progression through these stages can be variable, it has been estimated that it can take 2–3 weeks to progress from the brinopurulent to the organizing stage [21]. While most parapneumonic effusions progress through all three stages, primary empyema can proceed straight to the brinopurulent stage.

Clinical Presentation

The clinical features of parapneumonic effusion and empyema are poorly distinct from that of pneumonia.As initially described by Hippocrates, patients typically present as, “The fever was acute, and there were pains on either side, or in both, and if cough was present, and the sputa expectorated had a blood or vivid colour, or likewise was thin, frothy, and forid.” [10] As such, pneumonia both with or without a parapneumonic effusion present as a constellation of fever,

dyspnea, pleuritic pain, productive cough, and leukocytosis. Persistent clinical and biochemical signs of sepsis despite several days of antibiotics, however, should prompt the investigation of a parapneumonic effusion. The median duration of symptoms prior to presentation is 2 weeks; however, anaerobic infections can present more insidiously [2]. Symptoms can be acute or chronic, with both elderly patients and those with anaerobic infections presenting with anemia, fatigue, and failure to thrive [2, 22].

Radiologic Evaluation

The initial radiologic assessment should include the standard posterior-anterior and lateral projections of a chest radiograph. When free-fowing, pleural fuid collects in the most dependent area of the involved hemithorax: initially the costophrenic angle, then laterally, anteriorly, andnally apically. As much as 100 mL of pleural fuid may be undetectable via plain chest radiography. Between 175 to 500 mL is needed to blunt the lateral costophrenic angle. The lateral decubitus views can better de ne pleural effusions. The well-performed decubitus view can detect small volumes of fuid, reveal sub-pulmonic collections or pseudotumors, and identify loculations. The ubiquitous posterior-anterior and lateral chest radiograph can separate the broad air-fuid level of an empyema from the more spherical fuid collection surrounded by lung parenchyma characteristic of lung abscess.

Ultrasonography is widely available and frequently employed after a suspicious chest radiograph. Ultrasound is rapid and portable, but is operator dependent. This technique can localize small volume of fuid and de ne loculations; identify and characterize pleural peels; and de ne solid lesions such as pleural or parenchymal tumors. Using a 3.5 MHz or 5.0 MHz transducer and an intercostal acoustic window, an empyema is characterized as having acoustic homogeneity. Complex or advanced empyema has debris and foating fronds. An organized empyema has an echogenic pleural peel, and the lung appears immobile or entrapped. Diagnostic thoracentesis,