Equine Restrictive Lung Disease - Part 2: Pleuropneumonia

Introduction

Bacterial pleuropneumonia, frequently referred to as pleuritis, is a common and often severe disorder of horses [1-8]. The condition involves bacterial colonization of the pulmonary parenchyma, development of pneumonia and/or pulmonary abscesses, and subsequent extension to the visceral pleura and pleural space. In humans it is noted that up to 40% of patients with bacterial pneumonia have accompanying pleural effusions [9]. While similar data is not available for the horse, the increased use of thoracic ultrasound has documented that pleural effusion is not uncommon in any horse with pneumonia and is not restricted to those horses with severe pleuropneumonia [10].

Pathogenesis

The first stage of bacterial pleuropneumonia is an exudative stage characterized by rapid outpouring of sterile pleural fluid into the pleural space in response to inflammation of the pleura. The associated pneumonic process is usually contiguous with the visceral pleura and results in increased permeability of the capillaries in the visceral pleura. If appropriate antimicrobial therapy is initiated at this stage the pleural effusion may progress no further.

With progression the bacteria invade the pleural fluid from the contiguous pneumonic process and the second, fibropurulent, stage evolves. This stage is characterized by the accumulation of large amounts of pleural fluid with many neutrophils, bacteria, and cellular debris. Fibrin is deposited in a continuous sheet covering both the visceral and parietal pleural in the involved area. As this stage progresses, the tendency is for loculation and the formation of limiting membranes. These loculations prevent extension of the empyema, but make drainage of the pleural space with chest tubes increasingly difficult.

The last stage is the organization stage in which fibroblasts grow into the exudate from both the visceral and parietal pleura surfaces and produce an inelastic membrane called the pleural peel (Fig. 1). This inelastic pleural peel encases the lung and renders it virtually functionless. At this stage the exudate is thick.

Figure 1. Postmortem findings of thick pleural peel on surface of right lung. Left lung surface is normal.

Although pleuropneumonia can occur spontaneously, it is often associated with a stressful event such as transportation over an extended distance [1] or recent illness from acute viral disease. It is most commonly seen in Thoroughbred and Standardbred racehorses. Aspiration of pharyngeal secretions may play a significant role in the etiology of pleuropneumonia, as suggested by the bacterial populations responsible for pleuropneumonia. Transportation of horses usually involves an elevation in environmental temperature and relative humidity and an increase in the number of bacterial organisms in the air. These changes, combined with the stress of transportation, may predispose the animal to development of lower respiratory disease. The aerobic bacteria most commonly involved in equine pleuropneumonia include β-hemolytic Streptococcus spp., Pasteurella spp., Actinobacillus spp., E. coli and Klebsiella pneumoniae. The majority of the horses have a mixed infection, with both aerobic and anaerobic bacteria. Commonly isolated anaerobes include Bacteroides spp. and Clostridium spp. A wide variety of other anaerobes are commonly found in these horses [2].

Clinical Signs

Clinical signs include fever, anorexia, depression, cough, respiratory distress, stiff gait, weight loss, sternal or limb edema and colic. In the acute stage of pleuritis, pain in the thorax may be elicited by palpation over the thoracic wall. Pain is demonstrated by grunts, intercostal muscle spasm, or even escape maneuvers by the patient. Horses may abduct their elbows and have a "catch" to inspiration. As more fluid accumulates in the pleural space and the disease becomes chronic, pain is less evident. Auscultation of a horse with pleuropneumonia reveals a normal lung sound in the dorsal lung field with no sounds or only bronchial-tracheal sounds heard ventrally. Pleural friction rubs are often not heard because they are present only in the acute stage of the disease. If they are heard, friction rubs are present predominantly at the end of inspiration and the early part of expiration. They disappear as inflammation decreases or as pleural fluids accumulate. Cardiac sounds are often heard over a wider area of the chest than normal, probably as a result of enhanced conduction of sound through the pleural fluid. Thoracic percussion frequently confirms the impression gained from auscultation. Pleural effusion causes a dullness of the ventral aspects of the lung field and is often delineated by a horizontal line (Fig. 2). Absence of adventitial sounds on auscultation (especially if performed in noisy barns) should not be considered evidence of normality. If a clinical suspicion of pneumonia remains despite unremarkable auscultation, a re-breathing examination, sonographic evaluation, or transtracheal aspirate should be performed to definitively rule out pneumonia.

Figure 2. Horse with pleural effusion secondary to pleuropneumonia. Horizontal line marked by tape indicates fluid level in thorax detected by auscultation and percussion.

Diagnostic Procedures

Thoracic Ultrasonography

Thoracic ultrasonography [10-14] is currently regarded as the preferred method to diagnose pleuropneumonia in the horse. While the value of the art of thoracic auscultation and percussion should not be undermined, clinicians managing horses with thoracic disease recognize the limitations of these tools. With the widespread use of thoracic ultrasound, the equine practitioner currently has the ability to determine not only the presence of pleuropneumonia, but also the location and the extent of the disease. Although sector scanners are superior (preferably 3.5 - 5.0 MHz transducers), linear probes can also be used to evaluate the thorax in practice.

Thoracic ultrasonography in horses with pleuropneumonia allows the clinician to characterize the pleural fluid and to evaluate the severity of the underlying pulmonary disease [14]. The appearance of the pleural fluid may range from anechoic to hypoechoic, depending on the relative cellularity (Fig. 3). This fluid is usually found in the most ventral portion of the thorax and causes compression of normal healthy lung parenchyma with retraction of the lung toward the pulmonary hilus. The larger the effusion, the greater the amount of compression atelectasis and lung retraction that occurs.

Figure 3. Sonographic appearance of a large volume of anechoic pleural fluid.

The presence of adhesions, pleural thickening, pulmonary necrosis and compression atelectasis can also be detected [14]. Fibrin has a filmy to filamentous or frond-like appearance and is usually hypoechoic (Fig. 4).

Figure 4. Sonographic appearance of fibrin on visceral and parietal pleura in horse with severe pleuropneumonia.

Fibrin is deposited in layers or in web-like filamentous strands on surfaces of the lung, diaphragm, pericardium, and inner thoracic wall limit pleural fluid drainage [14]. Dimpling of the normally smooth pleural surface results in the appearance of "comet-tail" artifacts, created by small accumulations of exudate, blood, mucus, or edema fluid. Pulmonary consolidation varies from dimpling of the pleural surface to large, wedge-shaped areas of sonolucent lung. Atelectatic lung is sonolucent and appears as a wedge of tissue floating in the pleural fluid. Necrotic lung appears gelatinous and lacks architectural integrity. Peripheral lung abscesses are identified ultrasonographically by their cavitated appearance and the absence of any normal pulmonary structures (vessels or bronchi) detected within. While detection of a pneumothorax may be easy for the experienced ultrasonographer it is not as easy for the less experienced. The gas-fluid interface can be imaged moving simultaneously in a dorsal to ventral direction with respiration, the "curtain sign" reproducing the movements of the diaphragm (Fig. 5).

Figure 5. Sonographic appearance of pneumothorax.

The dorsal air echo moves ventrally during inspiration, similar to the lowering of a curtain, gradually masking the underlying structures. A pneumothorax without pleural effusion is even more difficult to detect ultrasonographically. While free bright gas echoes within the pleural fluid can occur following thoracocentesis, they are more often seen with anaerobic infections or when sufficient necrosis has occurred in a segment of parenchyma to erode into an airway and form a bronchopleural fistula (Fig. 6). The absence of gas echoes in pleural fluid does not rule out the possibility that anaerobic infection may be present.

Figure 6. Sonographic appearances of free gas echoes within the pleural fluid of horse with severe pleuropneumonia.

Ultrasonography is a valuable diagnostic aid in the evaluation of the pleura, lung, and mediastinum of horses with pleuropneumonia. The detection and further characterization of the above abnormalities improve the clinician's ability to form a more accurate prognosis. Adhesions can be detected which ultimately may affect the horse's return to his previous performance level. Horses with compression atelectasis and a non-fibrinous pleuritis have an excellent prognosis for survival and return to performance. The detection of areas of consolidation, pulmonary necrosis, or abscesses all increase the probable treatment and recovery time and the prognosis for survival decreases as they become more extensive. Ultrasonography can be used as a guide to sample or drain the area with a large fluid accumulation or the least loculation. These patients often benefit from progressive scanning to assess response to treatment and the need for drainage.

Thoracic radiography is often limited by the availability of facilities. Radiography has particular utility in demonstrating parenchymal disease which is not apparent sonographically if it is not contiguous with the lateral pleural surfaces or the lung.

If pleural effusion is identified, thoracocentesis should be considered for both diagnostic as well as therapeutic reasons. In the acute stages of pleuropneumonia with small volumes of pleural effusion, thoracocentesis is not necessary if the horse is improving or is not showing signs of respiratory distress. Moderate amounts of pleural effusion may be resorbed quite readily. However, if fluid accumulates rapidly, if the horse is in respiratory distress, or if its condition deteriorates, thoracocentesis should be performed. The preferred site is generally identified with the aid of ultrasonography, but is usually the sixth or seventh intercostal space just dorsal to the palpable costochondral junction. Choosing a site farther caudal may provide a sample but does not allow adequate drainage of the chest. When attempting to aspirate pleural fluid from a horse with a minimum amount of effusion, one should choose a space no farther back than the sixth or seventh intercostal space. If the procedure has caused some trauma, the first fluid obtained may be blood tinged, but this clears as more fluid is withdrawn. If the pleural fluid is blood tinged because of the underlying disease process, the red coloration persists throughout the entire procedure. Prior to draining the contralateral side, it is advisable to re-evaluate the fluid level sonographically. In some cases, mediastinal communication is adequate to allow both hemi-thoraces to be drained using a thoracostomy tube. An aliquot of pleural fluid is transferred from the syringe into tubes containing anticoagulant solution (EDTA) so that appropriate laboratory evaluation may be performed. Part of the fluid should be saved in sterile containers with transport media for subsequent Gram stain and culture. Fluid should be removed as long as it flows freely.

The color, turbidity, viscosity and odor of the pleural fluid should be noted (Fig. 7). Normal pleural fluid is clear and yellow; cloudiness reflects an increased number of white blood cells (WBC). Putrid-smelling pleural fluid is a hallmark of anaerobic infection; however, the absence of odor does not exclude anaerobic infection. In addition to the odor of the pleural fluid, the odor of the horse's breath should be noted, particularly after coughing. The majority of horses with anaerobic infections have a putrid odor associated with the pleural fluid or breath. These horses have a low survival rate.

Figure 7. Abnormal appearing pleural fluid from horse with pleuropneumonia.

The WBC count of normal pleural fluid is generally less than 10,000/ul. The WBC count of pleural fluid in pleuropneumonia can range from 1600 to 300,000 cells/ul, varying in the same pleural fluid sample between the beginning and the end of the thoracocentesis. There is no association between the WBC count in pleural fluid and survival. Pleural fluid protein is greater than 3 g/dl in horses with pleuropneumonia, but this is also not a prognostic indicator.

Pleural fluid should be Gram stained and cultured for bacteria. The Gram stain may provide tentative identification until culture results are obtained. Both aerobic and anaerobic cultures should be performed. Anaerobes occur in 46% of horses with pleuropneumonia. The pleural fluid used for anaerobic cultures should be transferred to the laboratory immediately after collection in a manner that prevents or minimizes exposure to air. Anaerobic transport media is commercially available and should be routinely used. Specimens submitted for isolation of anaerobes should not be refrigerated, since many anaerobes are intolerant to cold. Isolation of anaerobic bacteria from either the pleural fluid or tracheobronchial aspirate provides a poor prognosis.

While pleuropneumonia is the most common cause of pleural effusion in the horse, the second most common cause is neoplasia. Differentiating between the two conditions is a challenge for the equine clinician as there are similarities in the clinical signs and physical examination findings. A handheld glucometer can be used to compare glucose concentrations of pleural fluid to serum; typically septic effusions will have glucose concentrations significantly lower than blood levels due to bacterial consumption.

Pleuropneumonia effusions are more likely to have abnormal nucleated cell count greater than 10,000/ul (usually greater than 20,000) with greater than 70% neutrophils. Bacteria are frequently seen both intra and extracellularly. A putrid odor may be present. Neoplastic effusions have variable nucleated cell count. If caused by lymphosarcoma, there may be a predominance of abnormal lymphocytes. However, neoplastic cells are often not readily apparent and a definitive diagnosis may be difficult. Rarely do neoplastic effusions have a putrid odor nor have bacteria seen cytologically.

Tracheobronchial aspirate provides an excellent specimen for Gram stain and bacterial culture. This can be achieved percutaneously with minimal equipment, or using a triple-guarded catheter trans-endoscopically. Using the former method, care must be taken that the catheter is not retroflexed into the pharynx or normal oral flora will contaminate the sample. This can be minimized by using an antitussive such as butorphanol (5 - 20 mg IV) given at least 10 minutes prior to sampling.

Pleuroscopy, a procedure in which a flexible or rigid endoscope is introduced into the pleural space, has been described as a useful adjunct to the diagnosis as well as treatment of bacterial pleuropneumonia, aiding in assessment and identification of disease, drainage of abscesses or loculated fluid and adhesion transection [15].

Hematologic findings in horses with pleuritis are usually nonspecific and do not predict the outcome of the case. A low hematocrit (less than 30%) usually reflects an anemia of chronic disease, whereas elevated total plasma proteins (greater than 8 g/dl) are probably caused by hypergammaglobulinemia. Both these findings suggest that the pleuropneumonia is chronic. White blood cells counts can be misleading, since not all affected horses have leukocytosis. Plasma fibrinogen appears to be a more sensitive indicator of inflammation because it is elevated in almost all cases of pleuropneumonia. Recent descriptions of the use of serum amyloid A and plasma surfactant D levels as a sensitive and more specific indicator of pneumonia have been described, but are not routinely used [16].

Management

The primary goals of managing a horse with pleuropneumonia are to stop the underlying bacterial infection, remove excess inflammatory exudate from the pleural cavity, and provide supportive care. Ideally an etiologic agent is identified from either the tracheobronchial aspirate or pleural fluid, and antimicrobial sensitivity determined. It is not unusual for the tracheal aspirate and pleural effusion to contain different populations of bacteria. Ideally, both samples should be cultured if available, either individually or batched. Without bacterial culture results, broad-spectrum antibiotics should be used because many horses have mixed infections of both gram-positive and gram-negative and aerobic and anaerobic organisms. Commonly used therapy is penicillin combined with an aminoglycoside such as gentamicin, or broad-spectrum agents such as trimethoprim and sulfamethoxazole, ceftiofur, or chloramphenicol. Enrofloxacin can be given parenterally or orally, but due to its lack of coverage against Streptococcus species, it is not recommended as a single agent unless Streptococcal disease has been ruled out [17]. Rifampin may be a useful addition to antimicrobial regimens if there is extensive abscessation or tissue necrosis as it has excellent penetration; however, it is not suitable as a single agent therapy as resistance forms very rapidly against it in the absence of additional antimicrobials. Because of the need for long-term therapy, initial intravenous or intramuscular antimicrobials may need to be followed by oral antimicrobials. In addition to systemic therapy, the use of gentamicin via nebulization has been shown to provide therapeutic levels within bronchial fluid while resulting in minimal serum concentration [18]. This may be a useful alternative in horses where systemic aminoglycoside therapy is contraindicated due to its nephrotoxic properties.

Treatment of anaerobic pleuropneumonia is usually empiric, since antimicrobial susceptibility testing of anaerobes is difficult because of their fastidious nutritive and atmospheric requirements. Thus familiarity with antimicrobial susceptibility patterns is helpful in formulating the treatment regimen when an anaerobe is suspected. The majority of anaerobic isolates are sensitive to relatively low concentrations of aqueous penicillin (22,000 IU/kg bwt, IV four times per day). Bacteroides fragilis is the only frequently encountered anaerobe that is routinely resistant to penicillin, although other members of the Bacteroides family are known to produce β lactamases and are potentially penicillin-resistant. Chloramphenicol (50 mg/kg, orally four times per day) is effective against most aerobes and anaerobes that cause equine pleuropneumonia. Metronidazole has in vitro activity against a variety of obligate anaerobes including B. fragilis. Pharmacokinetic studies indicate a dose of 15 mg/kg intravenously or orally, four times a day, is necessary to maintain adequate serum levels. Oral administration rapidly results in adequate serum levels and thus is an acceptable route of administration for horses with pleuropneumonia. Metronidazole is not effective against aerobes and therefore should always be used in combination therapy with the above mentioned drugs for aerobes. It is important to recognize the side effects of metronidazole, including loss of appetite and lethargy, and to stop the use of the drug when these signs are observed. Aminoglycosides and enrofloxacin should not be considered for the treatment of pleuropneumonia caused by an anaerobe unless it is used in combination therapy (i.e., with penicillin).

Other Therapies

Antimicrobial therapy should be continued until the fibrinogen has normalized, and radiographic or sonographic evidence of resolution has been achieved. Anti-inflammatory agents help reduce pain and may decrease the production of pleural fluid. This in turn may encourage the horse to eat and maintain body weight. Flunixin meglumine (500 mg once or twice daily) or phenylbutazone (1 to 2 g twice a day) is commonly used for this purpose. If significant pleurodynia persists or NSAIDs are contraindicated for renal on gastrointestinal reasons, the use of fentanyl patches is a costly but often effective addition. Corticosteroids are not typically used in bacterial pleuropneumonia in veterinary patients due to concern of immunosuppression in the face of bacterial infection. Human medicine is beginning to provide some data that low dose and short duration corticosteroid use may reduce mortality in severe pneumonia [19], but there is still not sufficient information to make a recommendation for its use in equine medicine. Rest and the provision of an adequate diet are important components of the treatment of pleuropneumonia. Because the disease course and period of treatment are usually prolonged, attempts should be made to encourage eating. However, often patients with pleuropneumonia lose significant body condition score prior to close attention being made to their caloric intake. It is worthwhile evaluating feed consumption and considering a calorie-dense diet within the first days of treatment. Intravenous fluids may be indicated in the acute stages of the disease to treat dehydration from anorexia and third-space losses into the thorax. In patients with severe gram-negative disease, evidence of endotoxemia and laminitis should be watched for closely and preventative measures taken at the first signs of disease.

Following the selection of an appropriate antimicrobial agent, the next decision to be made is whether to drain the pleural space [20]. Ideally the decision is based on an examination of the pleural fluid. If the pleural fluid is thick pus, drainage using a chest tube should be initiated. If the pleural fluid is not thick pus, but the Gram stain is positive and WBC counts are elevated, pleural drainage is recommended. Another indication for therapeutic thoracocentesis is the relief of respiratory distress secondary to a pleural effusion. There are many options for thoracic drainage including the following: intermittent chest drainage, indwelling chest tube (Fig. 8), pleural lavage, pleuroscopy and debridement, open chest drainage/debridement - no rib resection (standing), open chest drainage/debridement - rib resection (standing), open chest drainage and debridement (general anesthesia), and lung resection (general anesthesia).

Figure 8. Chest tube drainage of pleural fluid in a horse with severe pleuropneumonia.

Drainage of a pleural effusion can be accomplished by (1) using a cannula, (2) indwelling chest tubes or (3) thoracostomy. Thoracostomy [15,21] is generally reserved for severe abscessation of the pleural space or treatment of extensive adhesions. Thoracocentesis is easily accomplished in the field and may not need to be repeated unless considerable pleural effusion re-accumulates. Indwelling chest tubes are indicated when continued pleural fluid accumulation makes intermittent thoracocentesis impractical. If properly placed and managed, they provide a method for frequent fluid removal and do not exacerbate the underlying pleuropneumonia or increase the production of pleural effusion. The chest entry site and end of the drainage tube must be maintained aseptically. A one-way flutter (Heimlich) valve may be attached to allow for continuous drainage without leakage of air into the thorax. If a thoracostomy tube is placed aseptically and managed correctly, it can be maintained for several weeks. It should be removed as soon as it is no longer functional. Heparinization of tubing after drainage helps maintain patency. Local cellulitis may occur at the site of entry into the chest, but is considered a minor complication. Bilateral pleural fluid accumulation requires bilateral drainage in most horses.

Open drainage or thoracostomy may be considered when tube drainage is inadequate. It is important not to begin open drainage too early in the disease. An incision is made in the intercostal space with or without rib resection exposing the pleural cavity and causing a pneumothorax, unless the visceral and parietal pleura adjacent to the drainage site have not been fused by the inflammatory process. The wound is kept open for several weeks while the pleural space is flushed and treated as an open draining abscess. Some clinicians advocate placing an open thoracostomy tube (no valve) the day prior to a standing thoracotomy to ensure that the horse can tolerate the unilateral pneumothorax prior to surgery.

Pleural lavage may be helpful to dilute fluid and remove fibrin, debris and necrotic tissue. Lavage appears to be most effective in sub acute stages before loculae develop; however, pleural lavage may help break down fibrous adhesions and establish communication between loculae. Care must be exercised that infused fluid is communicating with the drainage tube. Lavage can be performed by infusing fluid through a dorsally positioned tube and draining it through a ventrally positioned tube. Ten liters of sterile, warm lactated Ringer's solution is infused into each affected hemithorax by gravity flow. After infusion, the ventrally placed chest tube is opened and the lavage fluid is allowed to drain. Pleural lavage is probably contraindicated in horses with bronchopleural communications because it may result in spread of septic debris up the airways. Coughing and drainage of lavage fluid from the nares during infusion suggests the presence of a bronchopleural communication. Therapeutic bronchoalveolar lavage (BAL) has been reported to be successful in reducing mortality and morbidity in horses with transport-associated pneumonia [22], but this technique has not yet gained widespread acceptance.

Prognosis

Prognosis for life and return to performance is largely dependent on severity of disease and type of organism cultured. In studies, overall survival is 43 - 88% [23,24], but generally considerably lower for animals with anaerobic disease [2]. Horses with compression atelectasis and a non-fibrinous pleuritis have an excellent prognosis for survival and return to performance. In addition, one study found that horses with pulmonary abscessation had a 90% survival rate [1]. Data regarding post-pneumonia performance is variable, but generally approximately half of thoroughbreds and a majority of Standardbreds were successfully returned to racing, with the effect on performance being variable but not devastating [1,25].