Pyothorax

"In cases of empyema treated by the cautery or incision, when the matter is pure, white, and not fetid, the patient recovers; but if of a bloody and dirty character, he dies."
Hippocrates (470-410 BC) in his original works on empyema [1].

Pyothorax or thoracic empyema is a disease that affects dogs and cats. This disease has been recognized in humans for over 2000 years, with Hippocrates being one of the first people to document it [1]. Pyothorax is characterized by a septic thoracic effusion and infection of the pleural space. This rare but life-threatening disease has many proposed etiologies and often the cause is not found antemortem.

Etiology

The many proposed mechanisms for pyothorax include: inhaled or ingested plant material (plant awns), migrating foreign bodies, esophageal perforation, parasitic migration, penetrating thoracic wounds, hematogenous or direct extension of infection (systemic sepsis), secondary to bronchopneumonia, parapneumonic spread, mediastinitis, iatrogenic causes, and subphrenic infection [2-4]. The majority of the time the inciting cause of pyothorax is never determined antemortem.

In cats, one of the widely believed mechanisms for pyothorax is penetrating thoracic bite wounds. Although bite wounds account for a small percentage of documented cases of pyothorax, this cause is rarely identified antemortem [2,3,5]. The mechanism of development of pyothorax from a bite wound is contamination from normal oral flora [3,6,7]. This has been popularized because there is an overrepresentation of cats from multi-cat households, but interestingly enough, intact male outdoor cats are not overrepresented in some studies [3,5]. Also no statistical difference exists in the number of indoor cats versus outdoor cats in these studies [3,5]. The most commonly held description for the overrepresentation of cats from multi-cat households is that these bite wounds are small and go unnoticed, and the wounds go on to abscess and rupture intrathoracically [3,5]. The authors of one paper contest this belief and subscribe to the thought that parapneumonic spread of infection after colonization and invasion of lung tissue by oropharyngeal anaerobes is the most frequent cause of feline pyothorax [3]. The overrepresentation of cats from multi-cat households can be attributed to the fact that cats from this population are more predisposed to upper airway infections, thus leading to pneumonia and then parapneumonic spread of bacteria, leading to pyothorax [3]. Because of the controversy surrounding the cause of pyothorax in cats a complete physical exam (including dermatologic examination for bite wounds) and a thorough history (inquiring about recent upper airway infections) should be performed on all cats.

In dogs, the most commonly accepted etiology is the migration of plant awn material via inhalation, ingestion, or transdermal migration from the awn penetrating elsewhere in the body. Common species of plants that have been identified have been grasses of the Poaceae family, Hordeum jubatum (in the majority of the United States), Stipa and Setaria (in the southern United States), and Hordeum murinumin France and Australia [8]. This etiology has gained support because of the predominace of sporting breeds in the reported literature [2,8-13]. The inhaled or ingested plant material is thought to carry oral flora into the lungs with the awn as well as environmental contaminants such as Actinomyces spp. and Nocardia spp. The plant awns that migrate through the skin are thought to carry environmental bacterial species as well as skin contaminants. This idea is supported by the fact that in most cases the pyothorax is composed of polymicrobial populations that contain these oral and environmental microorganisms [2,3,9,10,12,14,15]. Adding further confusion to the problem, some dogs that have pyothorax have not had exposure to plant awn material. With any case of pyothorax, a complete physical and thorough history should be performed, although it is uncommon to find the inciting cause in many cases of pyothorax [2,3,10,11,14].

Microbes

Pleural fluid that is suspicious for pyothorax (exudative fluid) should be submitted to a laboratory for cultures and antimicrobial sensitivity analysis for aerobic, Mycoplasma, and anaerobic microbes. Pyothorax has a high prevalence of polymicrobic infections [2,3,9,10,12,14,15]. Both Actinomyces spp. and Nocardia spp. are common environmental microorganisms that are thought to be introduced into the animal through plant material foreign bodies. Actinomyces spp. can also be found as part of the normal oropharynx flora in both dogs and cats [4]. Actinomyces spp. are opportunistic pathogens that rely on mechanical disruption of normal mucosal barriers, such as what happens with plant awn migration [4]. Infections with Nocardia spp. are thought to be similar in nature because Nocardia spp. are ubiquitous soil saprophytes that degrade organic matter and are found in soil, water, and on plants [4]. The most common isolated microorganisms from dogs and cats are Bacteroides, Clostridium, Peptostretococcus, and Pasteurella multocida [7,15]. These bacteria are considered normal flora of the oropharynx and upper respiratory system of dogs and cats (Table 56-1) [16].

History, Signalment, and Physical Examination Findings

Pyothorax has been recognized in dogs with a mean age of 3 to 4 years (range: 1-11 years) with one case report of a neonate boxer puppy [2,9,11-14,17]. Cats range in age from 1 to 11 years with a mean of 3.83 to 5.8 years with one report of a 4-month-old lion cub and a report of 1-month-old kitten [3,5,11,18-22]. Sex predilection for pyothorax is still somewhat disputed. Several studies report that canine males are more predisposed than females: 2:1 in one study and 1.25:1 in another [2,9,11-14]; but other studies suggest that no sex predilection exists [12,13,15]. Some discrepancy in sex predilection seems to exist in cats as well. One study suggested a male to female ratio of 2 to 2.5:1 [5,11] and other studies showed no difference between the sexes [3,15,21]. The majority of affected dogs are medium- to large-breed dogs, with hunting/working dogs being overrepresented in certain geographies [12]. No specific breed predilection is apparent, with most cases occurring during fall and winter months, which corresponds with typical hunting seasons [10].

Dogs and cats often present with clinical signs consistent with pleural effusion but this can vary depending on the severity of the effusion. The signs include: exercise intolerance, poor athletic performance, shortness of breath, tachypnea, lethargy, and weight loss. Other signs are open-mouth breathing with dyspnea on the inspiratory phase, abducted elbows, extended head and neck, labored breathing, muffled heart sounds, decreased lung sounds, dull percussive sounds of the ventral thorax, thoracic and/or abdominal wall swellings, penetrating wounds or other trauma [2,3,9-11,13,14,17,23]. The majority of cats present with a history of partial or complete anorexia as the most common historical finding [3], with other signs similar to that of dogs. Owners often miss the subtle signs in cats with pleural effusion. Less common signs such as lethargy, reluctance to lie down, and anorexia can be initial complaints for dogs [2,3,9-11,13,14,17]. Less common presenting complaints include cough, diarrhea, polyuria, and polydipsia. The most severe cases present with acute respiratory distress, shock, dehydration, pale mucous membranes, and hypothermia [10,19].

Complete Blood Count and Biochemical Analysis

Biochemical analysis is not specific for pyothorax; the most common abnormality is hypoalbuminemia. Other abnormalities include hyponatremia, hypochloremia, hyperglobulinemia, elevations in alanine aminotransferase or alkaline phosphatase, hyperglycemia, hypoglycemia, increase in blood urea, increase in creatinine, and increased bile acids [2,3,5,11,14,19].

It is recommended that all cats with pleural effusion be tested for feline leukemia virus (FeLV), feline immunodeficiency virus (FIV), and heartworm disease [24].

Results of hematology studies vary and are typically indicative of bacterial infections. These abnormalities include neutrophilic leukocytosis with or without a left shift, monocytosis, and eosinophilia. Anemia may or may not be present [2,3,5,11,14,19].

Diagnostic Imaging

Right and/or left lateral and either ventrodorsal or dorsoventral views can be taken in patients that are stable. Findings include blurring of the cardiac silhouette, presence of interlobar fissure lines, rounding of the lung margins at the costophrenic angles, separation of the lung borders away from the thoracic wall, scalloping of the lung margins dorsal to the sternum, lung lobe collapse or suspected pulmonary mass, elevation of the trachea, widened mediastinum, thickening of visceral pleura, enlarged sternal lymph nodes, diffuse and focal alveolar infiltrates, air bronchograms within consolidated lung lobes, bronchointerstitial disease, and diffuse interstitial disease [2,3,10,11,25]. Atelectasis can be documented in some cases. It is not always possible to distinguish among recoil atelectasis, lobar consolidation, and encapsulated pleural fluid [3]. It is advantageous to remove as much of the pleural effusion as possible before attempting radiographs. This will improve visualization of thoracic anatomy and often makes the patient more comfortable while positioning for the radiographs. If the patient is in respiratory distress, horizontal-beam radiographs can be taken [21]. These views are often useful because they help the clinician to differentiate pyopneumothorax from gas-producing bacteria. Radiographs can detect as little as 50 to 100 ml of fluid; the recumbent ventrodorsal position appears to be the most sensitive for detecting the smaller amounts of fluid [26]. It has been reported that animals with radiodense lesions in the mediastinum or pulmonary fields have a poorer prognosis if they are treated medically rather than surgically.2

The majority of affected dogs and cats present with bilateral effusion [2,3,5,11,21]. One study reported 82% (of total animals) presenting with bilateral effusion; 29% of cats had unilateral effusion, and 14% of dogs had unilateral effusion [11]. Another study in cats reported a similar finding of 76% bilateral and 24% unilateral effusion.3 In this study, 40% of cats had radiographic evidence of concurrent pneumonia.

Thoracic ultrasonography can be helpful in the diagnosis of pyothorax and may be used to find small volumes of pleural fluid not detected radiographically, to quantitate or further characterize an effusion, or to guide thoracocentesis [27]. Abnormal findings include free pleural fluid, consolidated or collapsed lung lobes, and/or fibrinous or fibrous tags that can extend between the parietal and visceral pleura [5,11]. Occasionally, effusions have been accompanied by septations and parietal pleural thickening [27]. Cranial mediastinal masses, pulmonary abscesses, pericardial effusion, and restrictive pleuritis can be detected using ultrasonography [11,27]. It is also feasible to conclude that foreign bodies could be detected with ultrasound; however, to our knowledge, this has not been documented in published reports. It is advantageous to have pleural effusion present when performing ultrasound in the thorax as the fluid provides an "acoustic window" that allows the ultrasonographer to visualize more of the thoracic anatomy [27].

Computed tomography (CT) and MRI are commonly used in human medicine to diagnose thoracic disease and with these modalities becoming more and more common in veterinary medicine, it is a resource that can be used depending on whether the patient is stable enough for general anesthesia or heavy sedation. These modalities are especially helpful for diagnosing certain disease processes such as neoplasia [28] and pulmonary abscesses.

Pleural Fluid Analysis

Thoracocentesis is a procedure that is simple, diagnostic, and therapeutic. The complications of thoracocentesis are minimal if proper technique is used. These complications can include laceration of underlying lung tissue, causing pneumothorax, hemothorax, or pulmonary hemorrhage [10]. If the pleural effusion is "pocketed," radiographs or ultrasound can help guide the clinician to draw from these "pockets". Approximately 5 to 10 ml of aspirated fluid should be saved for bacterial culture and antimicrobial sensitivity testing as well as analysis. Ideally these samples would be taken at the time of therapeutic thoracocentesis and before antibiotic therapy is initiated. A portion of the fluid should be placed in an EDTA tube for total nucleated cell count, total protein count, specific gravity, and cytologic examination [10,25,29]. The remainder of the sample fluid should be placed in a glass serum tube (red top without clot activator) for biochemical analysis as well as into culture-transport media for aerobic, Mycoplasma, and anaerobic cultures, and antimicrobial sensitivity analysis [10,29]. Samples that are going to be submitted for culture and sensitivity must be handled properly to maximize identification. Refrigeration of the samples must be avoided, and they must be submitted to the laboratory within 24 hours [10,30]. At the time of fluid collection for analysis 4 to 6 direct smears of fluid onto a microscope slide should be made, especially if the samples are to be sent out to a laboratory [25,31]. If these slides are to be sent to an outside laboratory, several should be unstained. If the clinician is going to examine the slides, they should be Gram stained and/or Diff Quick stained for examination [3,25].

Pleural effusion samples should be analyzed for physical, chemical, and cytologic characteristics. Physical parameters include volume, color, turbidity, viscosity, presence of sulfur granules, and odor [10].,25,29,31,32]. Increased turbidity, high viscosity, bad odor, and varying color (e.g., sanguineous, brown, tan, white/greenish, or "creamy" red) are characteristics of septic effusions [10,25]. Degenerating leukocytes and bacteria can compose flocculent debris in the form of yellow flecks or sulfur granules [10,25]. If possible, this debris should be included in the direct smears and in the samples that are submitted for culture and sensitivity [10].

Normal pleural fluid contains less than 1.5 g/dl of protein and less than 500 cells/μl. Exudative fluids typically contain more than 3 g/dl of protein and more than 7,000 cells/μl with variable cell types and have a specific gravity greater than 1.025 [10,29]. Some chemical properties have been shown to be useful in evaluating pleural fluid in humans and cats. To the authors' knowledge these have not been determined to be accurate in canine fluid analysis [10,32]. These parameters include lactate dehydrogenase (LDH), pH, and glucose levels. LDH levels greater than 200 IU/L, acidic fluid, and glucose measurements less than 30 mg/dl have been shown to be consistent with exudative fluids in cats [10,33]. In humans, increased levels of LDH are consistent with cell damage or inflammation [10,32,34].

When evaluating pleural effusion, the lack of microorganisms and/or degenerative neutrophils seen on cytologic analysis does not rule out an infectious cause [10,29]. Some bacteria may not produce enough toxins to cause toxic changes to neutrophils, and the absence of microorganisms may be a result of antibiotic administration prior to sampling or owing to the presence of microorganisms without cell walls (e.g., Mycoplasma) [10].

Prognosis

Pyothorax carries a fair to good prognosis in both dogs and cats treated with aggressive medical or surgical management. Reports of survival rates range from approximately 60% to 90%, depending on early diagnosis, underlying etiology, and aggressiveness of treatment [2,10]. In one retrospective study, dogs that had confirmed infection with Actinomyces spp. had a significant difference in disease-free intervals between those that were treated medically versus those treated surgically (1.56 to 15.78) [2]. In dogs that had evidence of mediastinal or pulmonary lesions, the difference in disease-free intervals was also significant between medically versus surgically treated animals (2.16 to 22.46) [2]. This evidence suggests that animals with confirmed Actinomyces spp. and/or mediastinal or pulmonary lesions have a better prognosis when they are treated surgically versus medically [2]. Cats with pyothorax were more likely to come from multi-cat households than single-cat households [5]. Hypersalivation and bradycardia were more common in non-surviving cats with pyothorax than in survivors [5].

Conclusion

Pyothorax carries a fair to good prognosis with the appropriate management. Often the inciting cause cannot be identified during antemortem examinations. Successful resolution of pyothorax depends on early detection, appropriate antimicrobial selection, and aggressive management. Aggressive management includes early surgical intervention when surgery is warranted.