General Evaluation of the Patient With Respiratory Disease

^aPharmacokinetics data are available for horses, but in most cases safety studies have not been performed in the equine species.

^bTwo doses 4 days apart will provide 10 days of coverage. If a longer treatment period is necessary, weekly (q7days) administration is sufficient after the initial two doses.

^cShould not be used in young growing horses because of the risk of arthropathy.

^dDilute and give by slow IV infusion.

^eAdminister orally only. Intravenous doxycycline has resulted in severe cardiovascular effects, including collapse and death in some horses.

IM, Intramuscular; IV, intravenous; PO, oral.

■ TABLE 31.3

In Vitro Antimicrobial Susceptibility of Aerobic Bacterial Isolates Commonly Isolated From Horses With Bronchopneumonia or Pleuropneumonia

Antimicrobials^a

^aPercentage of susceptible isolates.

^bApproximate number of isolates (some isolates were not tested against every antimicrobial agent).

^cCeftiofur is rapidly metabolized to desfuroylceftiofur in vivo. Desfuroylceftiofur is as active as ceftiofur against most bacterial pathogens, but most coagulase­positive Staphylococcus spp. are resistant. Therefore despite in vitro susceptibility, ceftiofur is not the ideal choice for the treatment of staphylococcal infections. —, Not tested or testing not warranted; AMI, amikacin; AMP, ampicillin; CHL, chloramphenicol; E, erythromycin; ENR, enrofloxacin; GM, gentamicin; P, penicillin; RIF, rifampin; TE, tetracycline; TMS, trimethoprim-sulfonamide; XNL, ceftiofur.

Adapted from Giguere S, Afonso T: Antimicrobial drug use in horses. In: Giguere S, Prescott JF, Dowling PM, editors: Antimicrobial therapy in veterinary medicine, 5th ed. Ames, IA Wiley-Blackwell; 2013, pp. 457-472. Isolates were obtained from multiple equine clinical specimens including but not restricted to tracheobronchial aspirates and pleural fluid.

Therapy of horses with mild lower respiratory tract infec­tions should be continued for a minimum of 10 days or until clinical signs resolve. In some cases the duration of therapy needed may preclude continuous intramuscular therapy because of muscle soreness. Trimethoprim-sulfamethoxazole (TMS) combinations offer the advantage of oral administration. The usefulness of TMS combinations for the treatment of severe bacterial respiratory tract infections in horses may be limited by their apparent limited in vivo activity against S. Zooepidemicus. In contrast to penicillin, TMS was ineffective in eradicating S. Zooepidemicus in a tissue chamber model of infection in horses.^50,51 This failure of in vivo response was observed despite in vitro susceptibility of the isolate and high concentrations of TMS in the tissue chamber fluid.^50,51 However, in a recent study oral TMS was effective in the treatment of mild to moderate lower respiratory tract infections caused by S. Zooepidemicus.⁵²

In more severe cases of bronchopneumonia and in all cases of pleuropneumonia, antimicrobial therapy ultimately should be selected based on results of culture and in vitro susceptibil­ity testing. Before obtaining these results, selection should be based on knowledge of the prevalence and susceptibility pattern of bacteria commonly isolated from affected horses (see Table 31.3). Polymicrobial and mixed aerobic/anaerobic infec­tions are common; thus broad-spectrum antimicrobial therapy is initially required. A combination of gentamicin for gram­negative coverage and penicillin for gram-positive and anaerobic coverage is commonly used as initial broad-spectrum therapy for moderate to severe bronchopneumonia. Enrofloxacin can be used as a substitute to gentamicin for gram-negative coverage in adult horses. Advantages of enrofloxacin over gentamicin include greater activity against Enterobacteriaceae, better penetration in phagocytic cells and tissues, and better activity in purulent material. However, enrofloxacin should never be used as standalone initial therapy in horses with bronchopneumonia because of its lack of activity against anaerobes and variable activity against streptococci such as S. Zooepidemicus. Ampicillin or cefazolin can replace penicillin for gram-positive coverage and offer the supplementary advantage of providing additional gram-negative coverage.

Treatment of anaerobic pleuropneumonia is usually empiri­cal, as 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 susceptible to relatively low concentrations of penicillin. However, B. fragilis, a frequently encountered anaerobe in horses with pleuropneumonia, is routinely resistant to penicillin. Other members of the Bacteroides family are known to produce β-lactamases and are potentially penicillin resistant. Metroni­dazole has excellent in vitro activity against a variety of obligate anaerobes, including B. fragilis. Oral administration rapidly results in adequate serum levels and thus is an acceptable route of administration for horses with pleuropneumonia. Therefore if anaerobic infection is suspected, oral metronidazole is usually added to the combinations mentioned previously. Malodorous nasal discharge, halitosis, or the presence of gas echoes in pleural effusion seen during ultrasonography suggests the presence of an anaerobic infection. However, the absence of these findings does not rule out anaerobic bacterial infection. Horses with extensive lung consolidation and horses with pleuropneumonia may benefit from having metronidazole added to the treatment regimen. Metronidazole is not effective against aerobes and therefore should always be used in combination therapy. Chloramphenicol is active against most aerobes and anaerobes cultured from horses with bronchopneumonia or pleuropneumonia. However, because of human health concerns, its use should be limited to the treatment of horses with severe anaerobic bacterial infections refractory to metronidazole therapy in countries where its use is allowed. Rifampin is bactericidal and active against streptococci and some species of anaerobes. It penetrates well into abscesses and may be helpful in the treatment of infections with walled-off abscesses. Rifampin should always be used in combination with another antimicrobial agent to decrease the likelihood of emergence of resistant mutants.

In cases of severe bronchopneumonia, lung abscesses or pleuropneumonia long-term antimicrobial therapy ranging from 3 weeks to several months may be required. In the initial stages of therapy, IV antimicrobials are preferred to achieve higher serum concentrations. Oral antimicrobial agents can be used later in the course of the disease if appropriate based on in vitro susceptibility testing. Clinical signs, lung auscultation, fibrinogen concentrations, and repeated ultrasonographic and/ or radiographic examination are useful in assessing response to therapy and deciding when to discontinue antimicrobial therapy. Stall rest must be enforced during therapy of pneu­monia, and return to exercise should be gradual and permitted only after the horse is clinically normal and antibiotic therapy has been completed.

AEROSOLIZED ANTIMICROBIAL AGENTS. Aerosolized anti­microbial agents might be a useful adjunct to oral or systemic antimicrobial agents, particularly in horses with chronic septic inflammatory airway disease and no or minimal involvement of the lung parenchyma. The rate and extent of penetration of a drug into most sites outside the vascular space, such as lung tissue, are determined by the drug's concentration in plasma, molecular charge and size, extent of plasma protein binding, and blood flow.⁵³ In other tissues such as the central nervous system and the eye, a lipid membrane provides a barrier to drug diffusion.⁵³ There is a similar barrier between blood and the bronchial epithelium, restricting penetration of some drugs into bronchial secretions and epithelial lining fluid of the lower airways.⁵⁴ Aerosol administration of antimicrobial agents can result in high drug concentrations in the respiratory tract while minimizing systemic concentrations and their resulting toxicity.

Antimicrobial delivery by inhalation is greatly influenced by the product formulation and type of nebulizer. Aerosol use of IV formulations can lead to exposure to potentially irritant or toxic additives and inappropriate pH or osmolality ranges. In one study, the particle size distribution and particle density of gentamicin sulfate and ceftiofur sodium aerosols were affected by the antimicrobial concentration of the solution.⁵⁵ Gentamicin concentrations of 50 mg/mL or ceftiofur concentrations of 25 mg/mL produced the optimal combinations of particle size and aerosol density when a medical ultrasonic nebulizer was used.⁵⁵ Development of a silent battery-operated nebulizer for horses (Flexineb [Nortev Ltd., Galway, Ireland]) has greatly facilitated delivery of antimicrobial drugs by inhalation in horses.

In healthy horses, aerosolization of 20 mL of the com­mercially available IV gentamicin sulfate solution (diluted to 50 mg/mL) using an ultrasonic nebulizer resulted in bronchial lavage fluid concentrations approximately 12 times higher than concentrations achieved by IV administration at a dose of 6.6 mg/kg.⁵⁶ In the same study, serum concentrations following aerosol administration were below 1 pg/mL at all times.⁵⁶ Once-daily aerosol administration of gentamicin to healthy horses for 7 consecutive days did not result in pulmonary inflammation or drug accumulation in the respiratory tract.⁵⁷ The major limitation to the use of aerosolized gentamicin in horses is its lack of activity against S. zooepidemicus, the most common bacterial pathogen of the equine respiratory tract. Nebulization of ceftiofur sodium (diluted to 50 mg/mL in sterile water) at a dose of 2.2 mg/kg with the Flexineb mask was well tolerated and resulted in drug concentrations in pulmonary epithelial lining fluid above the minimum inhibitory concentration of the drug required to inhibit the growth of 90% of S. zooepidemicus, Pasteurella spp., and Actinocbacillus spp. for approximately 24 hours after administration.⁵⁸ Additional studies are required to assess the efficacy of aerosolized antimicrobial agents for the treatment of bacterial respiratory tract infections in horses.

ANCILLARY TREATMENTS. The need for ancillary treatments depends on the severity of the disease and is most often neces­sary in horses with pleuropneumonia. Nonsteroidal antiinflam­matory agents (NSAIDs) may be beneficial to minimize inflammation, provide analgesia, and control high fevers. Additional analgesia may be necessary in horses with severe pleurodynia. Adequate hydration should be maintained in patients receiving these agents for extended periods, especially if aminoglycosides are used concurrently. Intravenous fluid therapy may be necessary to correct hypovolemia in acute stages, but it is rarely required for chronic cases. Intrapharyngeal insufflation of oxygen is indicated in horses that remain severely hypoxemic despite adequate drainage of the pleural cavity. Adequate parenteral or preferably enteral nutritional support via a nasogastric tube is beneficial in horses that remain anorectic for several days. In horses with severe systemic illness, distal limb cryotherapy is indicated in an attempt to prevent develop­ment of laminitis.

PLEURAL DRAINAGE. Small amounts of pleural effusion may resorb quite readily with appropriate antimicrobial therapy. Therefore although a small sample of pleural fluid is useful diagnostically, pleural drainage is not necessarily indicated in all cases of pleuropneumonia. Drainage of pleural effu­sion results in removal of exudate and debris and allows for reexpansion of the lungs. Indications for drainage include a poor response to conservative therapy or the presence of pleural fluid with at least one of the following characteristics: (1) sufficient volume to cause respiratory distress, (2) fetid odor, or (3) cytologic or biochemical evidence of sepsis. Pleurocen- tesis and chest tube placement are described earlier in this chapter.

PLEURAL LAVAGE. Pleural lavage may be helpful to dilute thick, viscous pleural fluid and remove fibrin, debris, and necrotic tissue. Pleural lavage is most helpful in subacute stages before loculated pockets of pleural fluid develop. However, pleural lavage may help break down fibrous adhesions and establish communication between loculae. Pleural lavage is typically performed by using the same chest tube for infu­sion and drainage of lavage fluid. Alternatively, lavage can be performed by infusing fluid through a dorsally positioned tube and draining it through a ventrally positioned tube. Approximately 5 L of sterile, warm isotonic fluid solution is infused by gravity flow. After infusion, the chest tube is reconnected to the unidirectional valve for drainage. Allowing the horse to walk once the continuous flow has stopped will often result in drainage of additional fluid. Pleural lavage is probably contraindicated in horses with bronchopleural com­munications because it may result in spread of septic debris up the airways and into normal areas of the lungs.⁵⁹ Coughing and drainage of lavage fluid from the nose during infusion suggest the presence of a bronchopleural fistula (see the Complications of Pleuropneumonia section later).

USE OF FIBRINOLYTIC AGENTS. The use of intrapleural fibrinolytics in the treatment of loculated pleural effusion remains controversial in human medicine.⁶⁰ Various fibrinolytic agents such as streptokinase, urokinase, and recombinant tissue plasminogen activators (rtPA; e.g., alteplase, tenecteplase) have been used.⁶⁰ Numerous case series in humans have shown that such therapy is fairly safe and may facilitate drainage. However, controlled trials are scarce and have given contradictory results regarding the ultimate benefit of intrapleural fibrinolytic therapy. Subgroup analysis of a meta-analysis of randomized controlled trials indicated that intrapleural fibrinolysis with urokinase might reduce the need for surgery and shorten hospital stay, whereas similar effects were not documented with streptokinase or rtPA.⁶¹

In recent years, rtPA has been increasingly used as an adjunct therapy for selected human patients with loculated pleural effusion. Several case series indicate a potential benefit for rtPA,^62,63 but large randomized placebo-controlled clinical trials are lacking.^60,62 Fibrinolytic drugs have negligible effects on decreasing the viscosity of purulent material because the viscos­ity of pus is attributable mainly to its DNA content. Therefore intrapleural recombinant human deoxyribonuclease I (rhDNase I) is sometimes used in combination with rtPA.⁶⁴ In a random­ized clinical trial, intrapleural rtPA-rhDNase I therapy improved fluid drainage in humans with pleural infection and reduced the frequency of surgical intervention and the duration of the hospital stay, whereas treatment with rhDNase I alone or rtPA alone was ineffective.⁶⁵

Given that fibrinous pleural effusion has been associated with a less favorable outcome in horses with pleuropneumonia,⁶⁶ there has been increasing interest in the use of intrapleural fibrinolytic agents. There are isolated reports and a retrospective case series reporting the use of rtPA in horses with fibrinous pleuropneumonia. In one report, 12 mg of alteplase was infused in volumes of 0.9% saline ranging between 250 mL and 2 L.⁶⁷ In another report, tenecteplase (12 to 30 mg in 500 mL of 0.9% saline) was administered alone or in combination with 25 mg of rhDNase I.⁶⁸ In both reports there was a subjective decrease in the amount of fibrin in the thoracic cavity as assessed by ultrasonography. In a retrospective case series of 25 horses, altepase (0.4 to 20 mg per hemithorax in 10 mL to 2 L of polyionic solution) was administered 1 to 4 times per horse with no evidence of adverse effects and a subjective decrease in the amount of fibrin as assessed by ultrasonography after approximately 65% of treatments.⁶⁹ Controlled studies evaluat­ing the safety and efficacy of rtPA and rhDNase I are required before the widespread use of these agents can be recommended in the management of fibrinous pleuritis in horses. the nature and location of the lesion must be thoroughly characterized by ultrasonography or thoracoscopy to determine the ideal surgical site. When there is bilateral disease, thora­cotomy is performed on the most severely affected side. If necessary, a second thoracotomy can be performed on the opposite side at a later time. Thoracotomy is typically performed with the horse standing. It is common practice to place a large chest tube into the targeted cavity and leave it open to atmo­spheric air for several hours before thoracotomy.⁵⁹ The onset of respiratory distress indicates development of bilateral pneumothorax, in which case standing thoracotomy is contra­indicated and pneumothorax must be corrected. Unilateral pneumothorax is usually well tolerated.

The standing surgical procedures most commonly performed include thoracotomy via an intercostal approach and thora­cotomy with rib resection. For the intercostal approach, the surgical site is prepared and infiltrated with local anesthetic. The lateral thoracic vein must be identified to avoid inadvertent incision. A vertical incision is made through the skin, intercostal musculature, and pleura. The length of the incision is dictated by the size of the pleural abscess and consistency of its content. When necessary, partial excision of the intercostal muscles will facilitate manual exploration of the cavity and removal of fibrin and necrotic debris. The major advantage of the inter­costal approach over rib resection is preservation of thoracic wall integrity and compliance. Thoracotomy with rib resection is elected when the cavity is very large and it is anticipated that extensive manual debridement of fibrin and necrotic debris will be necessary. The advantage of the thoracotomy with rib resection over the intercostal approach is improved access to the thorax, allowing manual removal of large fibrin clumps and necrotic debris (Fig. 31.11). With both rib resection and intercostal approaches, the incision is left open and irrigated once or twice daily with a sterile isotonic fluid solution. The cavity is periodically debrided via gentle massage, taking care not to disrupt mature adhesions. After adequate formation of granulation tissue, tap water may be used. Depending on the size of the incision, it may take a few weeks to 2 to 3 months for complete closure by second intention. Complications during thoracotomy may include bilateral pneumothorax (if the cavity is not walled off and the mediastinum is not complete) and cardiac dysrhythmias (if the lesion is in close proximity to the heart). The most common long-term complication is the formation of a chronic draining fistula, but by itself this complication does not prevent the horse from returning to its usual occupation. In a retrospective study of 16 horses that had a standing lateral thoracotomy for pleural disease, 14 horses (88%) survived to discharge, and 46% of horses that survived returned to their previous level of athletic activity.⁷¹ In another retrospective study, 9 of 11 horses that had a thoracotomy

THORACOSCOPY. Thoracoscopy allows direct evaluation of the lungs and pleural cavity. In selected cases, thoracoscopy may be a useful tool to facilitate placement of thoracic drains in abscesses, transect pleural adhesions, and disrupt loculations.⁴⁶ The technique can also be used to biopsy or aspirate specific lesions affecting the periphery of the lungs. The procedure can be performed in the standing sedated horse with local anesthesia and is usually very well tolerated.⁷⁰

Adhesions between the lung and the diaphragm or body wall may prevent complete collapse of the lung, making insertion of the instruments more difficult and sometimes limiting the view of the pleural cavity. Adhesions are easier to disrupt during the first week after formation when the tissue is fibrinous rather than fibrous. Transection of mature adhesions is more difficult and can result in severe hemorrhage. Creation of a pneumothorax is a necessary feature of thoracoscopy, and horses should be monitored carefully throughout the procedure. Transient exacerbation in clinical signs of pulmonary disease due to pneumothorax can be alleviated by reinflation of the lungs.

THORACOTOMY. Surgical intervention in horses with pleuropneumonia is not a substitute for adequate medical management, but in carefully selected cases it can save the lives of horses that would have to be euthanized otherwise. The criteria for surgical interventions include (1) failure to respond to antimicrobial therapy, pleural drainage, and pleural lavage; (2) stable systemic medical condition; (3) presence of a large amount of fibrin, debris, or pus in the pleural space; and (4) either a walled-off lesion or presence of a complete mediastinum to avoid creation of a bilateral pneumothorax. Surgical intervention is most beneficial in chronic cases with large unilateral localized pockets of thick debris, especially if there is resolution or at least a significant improvement of the disease in the opposite hemithorax. Before surgical exploration,

FIG. 31.11 Thoracotomy with rib resection in a Thoroughbred filly with a large pleural abscess in the right hemithorax communicating with another abscess cranial to the heart. In this case, a rib resection was done to facilitate manual removal of purulent material and necrotic debris from both abscesses.

4 survived.⁴

■ Complications of Pleuropneumonia Several com­plications may occur during medical therapy of severe pneu­monia and pleuropneumonia. These complications include jugular vein phlebitis or thrombosis from catheter placement, diarrhea resulting from antimicrobial therapy, pneumothorax or cellulitis secondary to thoracocentesis, coagulopathy, pleural abscesses, bronchopleural fistulas, pericarditis, and laminitis.

Pleural abscesses refractory to antimicrobial therapy are treated by thoracocentesis or thoracotomy (see earlier sections). In some cases of pleuropneumonia, the heart may act as a valve to trap effusion and inflammatory debris in the cranial thorax. Small cranial thoracic masses may result in nonspecific clinical signs such as fever, tachycardia, and sternal edema. Larger abscesses may result in jugular vein distention, forelimb extension (pointing), and caudal displacement of the heart. Diagnosis is made by ultrasonography (Fig. 31.12). It is neces­sary to pull a forelimb forward to successfully image the cranial thorax. Most horses with cranial thoracic masses will respond to conservative therapy with antimicrobial agents.⁷² Drainage of the abscess should be performed in cases refractory to medical therapy or when the mass interferes with normal cardiac function. Cranial thoracic abscesses may be sampled in the standing horse with a front limb pulled forward. However, when drainage and lavage are required, the procedure is best performed under short-term general anesthesia to avoid damage to the heart and major blood vessels present in the cranial

FIG. 31.12 Sonogram of the cranial thorax obtained from the third intercostal space of a 2-year-old Thoroughbred filly with distended jugular veins and pitting edema of the ventral thorax. A large, fluid-filled, loculated abscess was imaged.

thorax. It is not recommended to leave an indwelling tube because the triceps musculature often causes the tube to kink. Repeated drainage and lavage may be necessary in some cases.

Bronchopleural fistulas develop when necrosis of lung tissue leads to direct communication between the airways and the pleural cavity. Diagnosis may be confirmed by thoracoscopy or by injecting sterile fluorescein dye into the pleural fluid and looking for presence of dye at the nostrils or within the trachea by endoscopy. Horses with bronchopleural fistula will often cough during pleural lavage, and lavage fluid may be noted at the nostrils. Most bronchopleural fistulas seal spontane­ously as a result of adhesions to the chest wall or fibrin deposi­tion on the visceral pleura. Sealing may occur rapidly or take several weeks. In one case, partial pneumonectomy was suc­cessful at resolving a chronic bronchopleural fistula and pul­monary abscess.⁷³

■ Prognosis The prognosis for survival and return to normal athletic function depends on the severity and duration of clinical signs prior to therapy. Horses with septic inflammatory airway disease or mild to moderate bronchopneumonia have a very good prognosis for a return to previous athletic performance. The prognosis for horses with pulmonary abscesses and without concurrent pleuritis is also good. In one study, 45 of 50 (90%) adult horses with pulmonary abscesses survived.⁷⁴ In the same study, 92% of Standardbreds and 52% of Thoroughbreds raced after treatment of pulmonary abscesses.⁷⁴ For horses that returned to racing, performance after successful treatment of lung abscesses was not significantly different from that before the illness.⁷⁴

In a retrospective study of 327 horses with pneumonia or pleuropneumonia, the overall survival rate was 75%. In the same study, the survival rate for 81 horses from which anaerobic bacteria were cultured was only 38.3%.3 However, other studies failed to identify an association between the presence of anaerobic bacteria and decreased survival.^4,75 In cases of pleuropneumonia, retrospective studies have shown survival rates ranging from 43.3% to 87.6%.^4,14,75,76 Differences in survival rates among studies may reflect differences in referral populations as well as advances in therapy in more recent years. Many horses that would have been euthanized because of chronicity and lack of response to medical therapy several years ago are now successfully treated with the surgical approaches described above. This is evidenced by a retrospective of 153 horses with pleuropneumonia in which the survival rate was 95.7% when horses electively euthanized were excluded.⁷⁶

The effect of pleuropneumonia on subsequent racing performance has not been examined extensively. In one ret­rospective study, 43 of 70 (61%) horses that had recovered from pleuropneumonia returned to racing, and 24 of those 43 (56%) won at least one race.⁷⁷ In the same study, horses that required placement of an indwelling thoracic drain did not have a worse prognosis for return to performance compared to horses that did require placement of a drain.⁷⁷ In contrast, horses that developed complications such as pulmonary abscesses, cranial thoracic masses, or bronchopleural fistulas were significantly less likely to return to racing.⁷⁷

Rhodococcus equi Infections

FIG. 31.13 A, Left lung from a foal with severe pneumonia caused by Rhodococcus equi. There is marked consolidation of the cranial portion of the lung, and multiple abscesses are present caudodorsally. B, Cross-section of the same lung showing multiple abscesses filled with caseous material. (Courtesy Dr. William Castleman., University of Florida, College of Veterinary Medicine, Gainsville, Fla.)

the clinical disease in foals is endemic and devastating on some farms, sporadic on others, and unrecognized on still others. At farms where the disease is endemic, costs associated with veterinary care, long-term therapy, and mortality of some foals may be very high. This text reviews the clinical manifestations, pathogenesis, epidemiology, diagnosis, treatment, and control of infections caused by R. equi in foals.

■ Clinical Manifestations

PULMONARY DISEASE. The most common clinical mani­festation of disease caused by R. equi infections in foals is a chronic suppurative bronchopneumonia with extensive absces­sation (Fig. 31.13). The slow spread of the lung infection combined with the remarkable ability of foals to compensate for the progressive loss of functional lung make early clinical diagnosis difficult. Early clinical signs often consist of a mild fever, occasional cough, or slight increase in respiratory rate that may not be apparent unless foals are exercised or stressed by handling. If pneumonia progresses, clinical signs may include decreased appetite, lethargy, fever, cough, tachypnea, and labored breathing characterized by nostril flaring and increased abdominal effort. Bilateral nasal discharge is an inconsistent finding. In a recent report of 161 foals affected with R. equi pneumonia, the most common clinical signs were cough (71%), fever (68%), lethargy (53%), and increased respiratory effort (43%).3 A smaller percentage of affected foals present with a more devastating subacute form. These foals might be found dead or, more commonly, are presented in acute respiratory distress with a high fever and no previous history of clinical respiratory disease. Foals with the subacute form of the disease have a poor prognosis despite appropriate therapy.

Because ultrasonographic screening for early detection of pneumonia caused by R. equi has become routine practice at many endemic farms, the most frequently recognized form of R. equi infection at those farms is a subclinical form in which foals develop ultrasonographic evidence of peripheral pulmonary consolidation or abscessation without manifesting clinical signs.^4,5 At those farms, the cumulative frequency of ultraso­nographically visible areas of focal pulmonary consolidation or abscessation considerably exceeds the historical frequency of clinical pneumonia attributed to R. equi, indicating that many subclinically affected foals might spontaneously recover without therapy. In two independent studies at breeding farms endemic for R. equi infections, 80% to 90% of foals with ultrasonographically visible lesions recovered without antimi­crobial therapy.^6,7 The proportion of subclinically affected foals that progress to clinically apparent disease might vary by farm, geographic region, and age at which foals are examined.

Extrapulmonary DISORDERS. Extrapulmonary manifesta­tions of rhodococcal infections are common.⁸ In a study of 150 foals with R. equi infections admitted to a teaching hospital, at least 1 of 39 different extrapulmonary disorders (EPDs) were recognized in 74% of foals, although some EPDs, par­ticularly abdominal lesions, can be difficult to recognize antemortem and were recognized only during necropsy.⁸ Survival was significantly lower among foals with EPDs (43%; 48 of 111) than among foals without EPDs (82%; 32 of 39).⁸ Some EPDs (e.g., polysynovitis, uveitis) might be the first clinical manifestation of R. equi-associated disease, detected prior to signs of pneumonia.

Intestinal lesions are present in approximately 50% of foals with R. equi pneumonia presented for necropsy.⁹ However, the majority of foals with R. equi pneumonia do not show clinical signs of intestinal disease. In the same study, only 4% of the foals had intestinal lesions without pneumonia.⁹ The intestinal form of R. equi infection is characterized by a multifocal ulcerative enterocolitis and typhlitis over the area of the Peyer's patches with granulomatous or suppurative inflammation of the mesenteric and/or colonic lymph nodes.⁹ In some cases, the only abdominal lesion noted may be a single large abscess (usually in a mesenteric lymph node) that often adheres to the large or small bowel (Fig. 31.14). Foals with abdominal abscesses have a poor prognosis.⁸

Clinical signs associated with the abdominal form of the disease may include fever, depression, anorexia, weight loss, colic, and diarrhea.^8,10 Diarrhea might occur in foals infected with R. equi either as an EPD caused by pyogranulomatous typhlocolitis or as a result of antimicrobial treatment. Of 31 foals with ulcerative enterotyphlocolitis identified at necropsy, only 12 foals had diarrhea, whereas 9 foals had diminished growth.⁸ Marked GI lymphatic obstruction associated with increased protein concentration in the peritoneal fluid and systemic hypoproteinemia may lead to ascites, giving affected foals a pot-bellied appearance. Such foals have a poor prognosis because of the extensive granulomatous inflammation of the colonic mucosa and submucosa, and mesenteric lymph nodes.

Intermittent or persistent bacteremia with R. equi may be more common than recognized and may result in metastatic spread of infection. The proportion of foals that are blood culture-positive and the stages of disease during which bacteremia is most likely to occur are poorly defined. In one study, 6 of 10 foals with naturally acquired pneumonia had positive blood cultures.¹¹ More recently, R. equi was isolated from the blood of 11 of 19 foals, and foals with positive blood culture results were less likely to survive than foals that were culture-negative.⁸

FIG. 31.14 Large abdominal abscess from a 4-month-old foal with a history of weight loss, fever, and mild intermittent episodes of colic. A portion of the small intestine and a portion of the small colon are adhered to the abscess. Culture yielded pure growth of Rhodococcus equi.

Polysynovitis, characterized by effusion of multiple synovial structures and absence of lameness, occurs in approximately one fourth to one third of foals with naturally acquired R. equi infections.^8,12 Uveitis has also been described in foals naturally infected with R. equi? Foals with uveitis might display epiphora, photophobia, aqueous flare, hypopyon, iris discoloration, miosis, and other ophthalmic abnormalities. Detection of immuno­globulins within the synovial membrane or iris by immuno­fluorescence in a small number of affected foals combined with the fact that R. equi is rarely isolated from these sites at the time of diagnosis or necropsy ^8,12-14 have led to the wide­spread hypothesis that polysynovitis and uveitis are immune- mediated disorders.^13-15 However, experimental infection with virulent R. equi consistently results in polysynovitis and uveitis, and development of these conditions is associated with the severity of pulmonary disease.^16,17 Culture of synovial fluid and aqueous humor within days of the onset of clinical signs in these foals often yields R. equi, and histologic examination of the synovial membrane reveals suppurative inflammation with bacteria.^16,17 Therefore polysynovitis and uveitis are most likely septic processes resulting from bacteremia. However, the infection is eventually cleared from the synovial structures and the eye, resulting in chronic nonseptic inflammation at these sites at the time of diagnosis or necropsy. Regardless of the inciting cause, local therapy of the affected joints in clinical cases of R. equi-associated polysynovitis is usually not indicated because the effusion resolves without any apparent consequences if the primary infection responds to appropriate antimicrobial therapy. The presence of polysynovitis or uveitis in a foal between 3 weeks and 6 months of age is highly suggestive of R. equi infection and deserves further investigation (see the Diagnosis section later).

Bacteremic spread of the organism from the lungs or GI tract may occasionally result in severe septic arthritis of a single joint or, more commonly, osteomyelitis. However, foals can occasionally develop R. equi septic arthritis or osteomyelitis without apparent lung involvement or other source of infection. Foals with septic arthritis typically show moderate to severe lameness, whereas foals with polysynovitis are not lame or only minimally so. In addition to appropriate antimicrobial therapy (see the Treatment section later), foals with R. equi septic arthritis and osteomyelitis often require aggressive local therapy. R. equi vertebral osteomyelitis or diskospondylitis resulting in spinal cord compression has also been reported.^18-20 Clinical signs of vertebral osteomyelitis or diskospondylitis may include stiff gait, reluctance to move, palpable pain, and sometimes soft-tissue swelling associated with paravertebral abscessation.⁸ If the infection or associated swelling extends to the epidural space, neurologic signs of spinal cord disease or nerve root compression might be apparent.⁸ Specific neu­rologic signs vary according to the specific region of the spinal cord that is affected. Foals with vertebral osteomyelitis generally have a poor prognosis. However, there are reports of successful treatment with prolonged antimicrobial therapy alone or in conjunction with surgical debridement of infected bone.¹⁸

Other less common EPDs associated with metastatic spread of R. equi include pericarditis, endocarditis, cellulitis, dermatitis, subcutaneous abscesses, peripheral lymphadenopathy, guttural pouch empyema, pleuritis, sinusitis, ulcerative lymphangitis, myositis, stomatitis, pyometra, and omphalitis.^8,21 Other immune-mediated EPDs include immune-mediated hemolytic anemia, immune-mediated thrombocytopenia, and telogen effluvium.^8,22

ADULT HORSES. Disease caused by R. equi in adult horses is rare, but there are sporadic reports of infection involving primarily the lungs or abdominal lymph nodes (similar to infection observed in foals) and, less common, of wound infection. In one adult horse, acquired combined immunode­ficiency of unknown origin led to R. equi septicemia and lung abscessation.²³ Occasionally the microorganism has also been isolated from infertile mares and aborted fetuses.^9,24,25

OTHER DOMESTIC ANIMAL SPECIES. R. equi has been isolated from many species other than people and horses. R. equi is frequently cultured from the submaxillary lymph nodes of pigs with granulomatous lymphadenitis. However, R. equi can also be isolated from the submaxillary lymph nodes of 3% to 5% of apparently healthy pigs, and experimental infection studies in pigs have failed to reproduce granulomatous lymphadenitis.^26,27 The causative role of R. equi in granulomatous lymphadenitis in pigs remains unproven. Isolation of R. equi from other domestic animal species is rare. R. equi can be isolated from lymph node granulomas in 0.008% of cattle at abattoir post­mortem inspection.²⁸ R. equi has been cultured also from rare cases of bronchopneumonia, mastitis, metritis, ulcerative lymphangitis, and septic arthritis in cattle. In goats, R. equi has been cultured from cases of liver abscesses, bronchopneu­monia, subcutaneous abscesses, osteomyelitis, and disseminated infections involving multiple sites.^29,30 In dogs and cats, R. equi has been isolated from cases of pneumonia as well as from wound infections, subcutaneous abscesses, vaginitis, endocarditis, endophthalmitis, hepatitis, osteomyelitis, myositis, and joint 3132

infections.^31,32 A severe systemic form with suppurative to necrotizing pneumonia and multiple caseating abscesses in mediastinal lymph nodes, liver, and spleen has been observed in camelids (llama and dromedary).^33,34

Virulence

The ability of R. equi to induce disease in foals likely depends on both host and microbial factors. R. equi is a facultative intracel­lular pathogen, and its ability to persist in and eventually destroy alveolar macrophages seems to be the basis of its pathogenicity. Knowledge of the virulence mechanisms of R. equi was largely speculative until the discovery of a virulence plasmid in the early 1990s.^35,36 Unlike most environmental R. equi, isolates from pneumonic foals typically contain an 80- to 90-kb plasmid. Plasmid-cured derivatives of virulent R. equi strains lose their ability to replicate and survive in macrophages.¹⁶ Plasmid-cured derivatives also fail to induce pneumonia and are completely cleared from the lungs of foals within 2 weeks following heavy intrabronchial challenge, confirming the absolute necessity of the large plasmid for the virulence of R. equi.^16,3

Sequencing and annotation of the virulence plasmid obtained from isolates of R. equi cultured from pneumonic foals revealed 73 coding sequences^38,39 divisible into four discrete areas based on open reading frame (ORF) amino acid sequence similarity and predicted protein function. The “backbone” sequence of the plasmid is highly similar to that of a plasmid found in the environmental microorganism Rbodococcus erytbropolis and consists of regions for replication or partitioning, conjugation, and unknown functions.³⁸ The fourth plasmid region is a 21-kb pathogenicity island (PAI) that is crucial for virulence.^38,39 The 26 coding sequences of the PAI include the unique and R. equi-specific virulence-associated protein (Vap) family.³⁹ There are six full-length vap genes, (vapA, vapC, vapD, vapE, vapG, and vapH) and three truncated vap pseudogenes (vapF, vapI, and vapX)3 To date, vapA, which encodes an immunodominant, temperature-inducible, and surface-expressed lipoprotein,^40,41 is the only vap gene with a demonstrated role in virulence,⁴² whereas vapsC, D, E, F, G, I, and X are dispensable.^42,43 VapA is required for intracellular growth in macrophages,⁴² where it disrupts endolysosomes and contributes to preventing maturation of the phagosome to the stage of fusion of R. equi-containing vacuoles with lysosomes.^44,45 The functions of the other Vap proteins are unknown.

Although vapA is necessary for virulence, it is not sufficient.¹⁶ Two additional regulator genes of the PAI, virR and virS, are required for intracellular growth in macrophages and virulence.^46-49 Loss of either regulator results in decreased transcription, not only of vapA, but also of several chromosomal genes.^46-49 These findings demonstrate the presence of molecular “crosstalk” between the virulence plasmid and the R. equi chromosome. All of the vap genes as well as five other ORFs within the PAI are upregulated when R. equi is grown in macrophage monolayers.⁵⁰ Regulation of expression of the genes in the PAI is complex and depends on at least five environmental signals: temperature, pH, oxidative stress, magnesium, and iron.^50-52 The precise role of each of these genes in the pathogenesis of R. equi infections remains to be determined. The entire virulence plasmid can be transferred by conjugation from plasmid-containing strains of R. equi to plasmid-free R. equi strains at a high frequency, and plasmid transmission restores the capacity for intracellular growth in macrophages.⁵³ Deletion of a putative relaxase-encoding gene (traA) located in the proposed conjugative region of the plasmid prevents plasmid transfer.⁵³

Virtually all isolates of R. equi cultured from diseased foals contain the large circular plasmid described earlier (designated pVapA because it encodes vapA), whereas most isolates from swine with granulomatous lymphadenitis contain a similar large circular plasmid (designated pVapB) that encodes a 20-kDa antigen (VapB) that is related to but distinct from VapA.⁵⁴ Most isolates from cattle with granulomatous lymphadenitis and from goats with various infections contain a linear plasmid (designated pVapN, with “N” standing for “no-A no-B”) unrelated to the circular virulence plasmids pVapA and pVapB.^55,56 In contrast, many environmental isolates of R. equi do not contain any of these three plasmids and are referred to as avirulent because they lack the ability to replicate in macrophages and in experimentally infected mice and they cannot induce disease in foals. Isolates carrying pVapB or pVapN have not been isolated from foals with naturally acquired R. equi infections. Experimentally, heavy intrabronchial challenge of foals with R. equi carrying pVapB results in pneumonia but at a dose much higher than that required for induction of pneumonia with strains containing pVapA.⁵⁷ The observation that all R. equi isolates obtained from foals carry pVapA, most isolates from swine carry pVapB, and most isolates from ruminants carry pVapN-type plasmid has led to the hypothesis that plasmid type might dictate host species tropism. However, R. equi isolates from pigs, carrying pVapB, are able to replicate in a plasmid-dependent manner in macrophages obtained from a variety of species (murine, swine, and equine).⁵⁸ Similarly, equine isolates carrying pVapA are capable of replication in swine macrophages.⁵⁸ Plasmid swapping between equine and swine strains through conjugation does not alter the intracellular replication capacity of the parental strain.⁵⁸ These results demonstrate that, although distinct plasmid types exist among R. equi isolates obtained from different species, host tropism is not determined exclusively by plasmid type.

The plasmid profile of R. equi isolated from other species is not as straightforward as it is in horses, swine, and ruminants. Analysis of 65 R. equi isolates from humans with and without AIDS reveals that approximately 15% of isolates contain pVapA, approximately 17% contain pVapB, and approximately 25% contain pVapN.⁵⁹ Most isolates of R. equi cultured from humans do not contain pVapA, pVapB, or pVapN.^59-62 Therefore the pathogenesis of R. equi infection in immunocompromised human patients appears to be different from the pathogenesis in foals, in which pVapA is always found. Similarly, isolates from cats and dogs can be avirulent or contain any of the three plasmids.^31,63

Another major advance in the understanding of R. equi virulence was the sequencing and annotation of the genome of R. equi⁶ The 5.04-Mb R. equi genome is most similar to that of the environmental bacterium Rbodococcus jostii and then to Nocardia farcinica and Mycobacterium tuberculosis.^6,4 As with these other actinomycetes, a large number of the 4598 genes of R. equi appear to be involved in lipid metabolism. Like M. tuberculosis, lipids are a key component of the outer cell envelope of R. equi.⁶⁵ This mycolic acid-containing glycolipid barrier might serve to protect the peptidoglycan and plasma membrane from the damaging effects of host-generated enzymes and immune-mediated reactive intermediates. Cell envelope mycolic acid carbon chain length varies among R. equi isolates, and notably it was observed that strains with longer mycolic acids were more lethal to mice.⁶⁶ This might be explained by the fact that mycolic acid chain length plays a key role in diversion of macrophage phagosome trafficking by R. equi.⁶⁵ Although it has been firmly established that the virulence plasmid is essential for infection of foals, several chromosomally encoded genes have also been proven to play essential roles in intracel­lular survival and virulence, notably those involved in various metabolic functions such as fatty acid metabolism, anaerobic metabolism, aromatic amino acid biosynthesis, regulation of secreted proteins, iron acquisition, hydrogen peroxide resistance, and cholesterol catabolic pathways.^64,68-76

■ Pathogenesis Inhalation of virulent R. equi is the major route of pulmonary infection in foals. The incubation period following experimental intrabronchial challenge varies from approximately 9 days after administration of a heavy inoculum to approximately 2 to 4 weeks when a lower inoculum is administered.^16,77 Lung consolidation can be detected as early as 3 days following heavy intrabronchial challenge.¹⁶ However, the incubation period under field conditions is unknown and likely varies depending on several factors, including the number of virulent bacteria in the air, the age of the foal, and host defense mechanisms. Ingestion of the organism is an important route of exposure, and likely also of induction of an immune response, but rarely leads to hematogenously acquired pneu­monia unless a foal has multiple exposures to large numbers of bacteria.⁷⁸ Indeed, intragastric administration of live virulent R. equi to foals on day 2 and day 7 of life is highly effective in preventing development of pneumonia following subsequent heavy intrabronchial challenge.⁷⁹ Despite conferring protection, oral vaccination with live virulent R. equi would not be an acceptable strategy to immunize foals, since it would lead to progressive contamination of the environment with virulent bacteria. However, it undoubtedly shows that newborn foals have the ability to mount protective immune responses against R. equi (see the Immunity section later).

Epidemiologic evidence suggests that most foals on endemic farms become infected early in life.⁸⁰ The median age at the time of diagnosis is approximately 35 to 50 days on most endemic farms.^3,81 Given the fairly long incubation period of the disease, this finding would also support the fact that many foals become infected early in life. In one study, foals between 3 and 13 days of age (mean of 6.4 days) were more susceptible to experimentally induced R. equi pneumonia than foals between 14 and 36 days of age (mean of 25 days).⁸² In another study, foals between 3 and 42 days of age were experimentally infected with various doses of virulent R. equi (10³ to 10⁷).⁸³ Whereas the youngest foals were highly susceptible to infection, the oldest foals were resistant to thousandfold higher challenge.⁸³ Collectively, these findings indicate that many foals on endemic farms become infected at a young age and that younger foals are more susceptible than older foals to infection caused by R. equi. However, these findings do not necessarily indicate that foals are only susceptible to R. equi during the neonatal period. Older foals are also susceptible to experimental infection with R. equi, but they require a higher inoculum. In one study, intratracheal administration of R. equi to 10 foals between 27 and 67 days of age (mean of 49 days) resulted in disease in all foals.³⁷ The relative resistance of older foals to infection caused by R. equi compared with younger foals is likely the result of immunologic priming after natural exposure to the pathogen.

■ Epidemiology Although R. equi can be cultured from the environment of virtually all horse farms, the clinical disease in foals is endemic and devastating on some farms, sporadic on others, and unrecognized on still others. This might reflect differences in environmental and management conditions, as well as differences in the virulence of isolates, but the exact basis for the difference in the prevalence of the disease among farms remain unknown. Breeding farms with a large acreage, a large number of mares and foals, a high foal density, and a large population of transient mares and foals have greater odds of being affected by R. equi pneumonia.^84,85 In contrast, R. equi pneumonia does not appear to be associated with lack of attention to routine preventive health practices.⁸⁶

R. equi is a soil microorganism with simple growth require­ments. The highest numbers of R. equi are found in surface soil, and the organism cannot be found at depths exceeding 30 cm.⁸⁷ R. equi can be cultured not only from the soil of all horse farms but also from the soil of areas not inhabited by horses. In one study, R. equi was isolated from approximately 74% of soil samples collected from 115 parks and 49 household yards in Japan. The number of R. equi in those samples ranged from 10 to 10⁵ colony-forming units (CFU) per gram of soil. None of the 1294 isolates from those samples expressed VapA or VapB.²⁷

R. equi can be isolated from the feces of most adult horses at concentrations ranging from 10 to 10⁴ CFUs/g of feces. The number of R. equi in the feces of mares on endemic farms does not increase in the weeks before or after foaling.⁸⁸ In one study, virulent R. equi could be isolated from the feces of all mares sampled, demonstrating that mares can be an important source of virulent R. equi for their surrounding environment.⁸⁹ However, dams of affected foals did not shed more R. equi in feces than did dams of unaffected foals, indicating that heavier shedding by particular mares does not explain infection in their foals.⁸⁹ In another study, R. equi could be isolated from the feces of only 19% of the foals at 1 week of age and from all foals by 4 weeks of age.⁸⁸ Foals can shed high concentrations of R. equi in their feces, reaching numbers up to 10⁵ CFUs/g of feces.⁹⁰ Virulent R. equi present in the sputum of pneumonic foals will be swallowed, and as a result the manure of R. equi-affected foals is likely a major source of progressive contamination of the environment with virulent organisms.

In a survey of the prevalence of R. equi at horse breeding farms in Japan, the organism was isolated from almost all soil samples, at numbers of 10² to 10⁵ CFUs/g of soil.⁹¹ Although virulent R. equi isolates containing pVapA were cultured from 24 of 31 farms examined, the vast majority of these isolates did not contain plasmids and were avirulent. On those farms, virulent R. equi represented 1.7% to 23.3% of all isolates.⁹¹ However, neither the presence nor the concentration of virulent R. equi in soil is positively associated with increased cumulative incidence of R. equi pneumonia at breeding farms in North America or Australia.^92-94 In addition, soil geochemistry does not appear to be associated with occurrence of R. equi pneu­monia at farms.⁹⁵

In contrast, airborne concentration of virulent R. equi is associated with disease incidence at Thoroughbred breeding farms in Australia.⁹⁴ Factors associated with increased air concentration of R. equi at these farms included site (higher in holding pens and lanes relative to paddocks), warmer ambient temperature, less soil moisture, reduced grass height, and later date during the foaling season.⁹⁴ In Kentucky, concentrations of airborne R. equi were lower between midnight and 6 AM, and the presence of horses at the sampling site appeared to increase the airborne concentrations of virulent R. equi.⁹⁶ These findings suggest that airborne concentrations may be increased by greater activity and density of horses at sites where foals are housed. Also in Kentucky, foals were more likely to be exposed to airborne virulent R. equi when housed in stalls versus paddocks and earlier (January and February) versus later (May and June) during the foaling season.⁹⁷ Similarly, the odds of detecting airborne virulent R. equi at three breeding farms in Ireland were approximately 17 times greater in stables (stalls) than in paddocks of a given farm, and concentrations of virulent R. equi were also significantly higher in stables than in pad- docks.⁹⁸ In a recent study at a breeding farm in Texas, exposure of foals to airborne virulent R. equi during the first 2 weeks after birth was significantly associated with subsequent develop­ment of R. equi pneumonia.⁹⁹ The fact that exposure to R. equi preceded development of pneumonia suggests that the higher concentration of airborne virulent R. equi is likely the cause of pneumonia rather than an effect of it.

In one study, air samples from the breathing zone of pneumonic foals had higher concentrations of virulent R. equi than environmental air samples collected from lanes and pens at the same farms.¹⁰⁰ However, concentrations of virulent R. equi from air samples collected from the breathing zones were not significantly different between pneumonic foals and healthy controls, indicating that affected foals do not represent a greater risk than other foals for aerosol transmission. In addition, there was no association between virulent R. equi in the breathing zone of foals and the subsequent diagnosis of rhodococcal pneumonia.¹⁰¹ Thus, to date, there is no compelling evidence that R. equi infection is contagious among foals and that affected foals should be isolated from other foals.

Considerable genotypic variability has been found among isolates of R. equi obtained from the same farm and among isolates 102104

from various countries or continents.^102-104 Interestingly, multiple different R. equi strains were isolated in five of the six cases in which more than one isolate from a single foal was examined.¹⁰³ Similarly, six distinct genotypes were identified among nine R. equi isolates from one foal with pneumonia and concurrent abdominal lesions.¹⁰⁵ None of the four pulmonary isolates were identical; however, a pulmonary isolate was found to be identical to an isolate recovered from a small intestinal lymph node, and a second pulmonary isolate was identical to an isolate from a mesenteric lymph node.¹⁰⁵ Therefore foals can be infected with multiple strains of virulent R. equi, and with current techniques it is not possible to link infections to a given geographic site or region on the basis of analysis of isolate genotyping.

■ Rhodococcus equi-Phagocytic Cell Interactions Once inhaled, R. equi is taken up by alveolar macrophages

through a process of receptor-mediated phagocytosis. One of the receptors used by macrophages to engulf complement- opsonized R. equi is complement receptor 3 (CR3 or Mac-1).¹⁰⁶ In addition, R. equi might use the macrophage mannose receptor for entry, which may recognize lipoarabinomannan (LAM), an outer surface component of the bacterium, either directly or via mannose binding protein or surfactant molecules adhered to LAM.¹⁰⁷ Once engulfed by resident macrophages, virulent R. equi are able to modify the phagocytic vacuole to prevent acidification and subsequent fusion with lysosomes.^44,108-110 The ability of R. equi to disrupt endolysosome function depends on VapA,⁴⁵ and addition of soluble recombinant VapA restores the ability of avirulent R. equi mutant lacking the vapA gene to survive and replicate in macrophages.^111,112 Uncontrolled intracellular replication of R. equi leads to necrosis of the macrophage.¹¹³ If bacteria are opsonized with R. equi-specific antibody, presumably promoting bacterial entry via the mac­rophage Fc receptor, the fate of the R. equi-containing phago­some is altered and lysosome fusion occurs.¹¹⁴ This may explain how the presence of R. equi antibody might aid in the prevention of infection. Killing of R. equi by mouse macrophages is dependent on the presence of IFN-γ, which activates macro­phages to produce both reactive oxygen and reactive nitrogen intermediates. These two radicals combine to form peroxynitrite, which efficiently kills R. equi.¹¹⁵ Neither reactive oxygen nor reactive nitrogen intermediates alone are sufficient to mediate killing of R. equi.¹¹⁵ Macrophage activation and nitric oxide synthesis act at least in part by increasing the ability of mac­rophages to store iron in a way that it cannot readily be accessed by R. equi A⁶ Additional cytokines such as tumor necrosis factor (TNF)-α might have similar effects on restricting intracellular growth of R. equi in macrophages because both IFN-γ and TNF-α are required for clearance of virulent R. equi in mice.¹¹⁷ In a recent study, priming of equine monocyte-derived mac­rophages IFN-γ, TNF-α, or IL-6 decreased intracellular survival of R. equi, whereas IL-1β or IL-10 enhanced intracellular survival.¹¹⁸ Neutrophils also play an important role in early host defense against virulent R. equi.¹¹⁹ As opposed to macro­phages, neutrophils from foals and adult horses are fully able to kill R. equi.^120-122 As seen with macrophages, killing of R. equi by neutrophils is considerably enhanced by specific opsoniz- 123124

ing antibody.^123,124

■ Immunity

ANTIBODY-MEDIATED IMMUNITY. Immunity to R. equi pneumonia in foals likely depends on both the antibody and cell-mediated components of the immune system, but its exact basis remains to be determined. The strongest evidence for a role of antibody in protection against R. equi is the partially protective effect of passively transferred anti-R. equi hyperim­mune equine plasma (HIP; see the Control and Prevention of R. equi Infections on Farms Where the Disease Is Endemic section later) and the fact that maternal vaccination against poly-N-acetylglucosamine protects foals against intrabronchial challenge with R. equi.¹¹⁵ The mechanisms by which antibody confers protection against facultative intracellular bacterial pathogens is an area of active investigation. Opsonization of R. equi with specific antibody has been shown to promote phagocytosis and killing of R. equi by alveolar macrophages, identifying antibody as a critical component of HIP.^114,126 In most studies evaluating the protective effect of HIP, donors were immunized with whole cell vaccines or a mixture of several soluble antigens, making it impossible to determine the role of antibody against defined antigens of R. equi.

A number of studies have focused more specifically against the role of antibody against plasmid-encoded Vaps. First, mono­clonal antibody to VapA and serum from horses immunized with partially purified VapA have opsonizing activity.¹²⁷ Moreover, purified immunoglobulins obtained from horses vaccinated with partially purified VapA protected mice against intraperitoneal challenge with virulent R. equi compared with mice administered immunoglobulins from nonimmunized horses.¹²⁸ Finally, intra­venous administration of purified immunoglobulins obtained from horses immunized with recombinant VapA and VapC to foals was found to reduce the severity of pneumonia following heavy experimental challenge with R. equi.¹²⁹ In the same study, the degree of protection conferred by purified anti-VapA and anti-VapC immunoglobulins was similar to that provided by commercially available HIP.¹²⁹

In adult horses, the concentrations of R. equi- and VapA- specific IgGa and IgGb antibodies (the IgG isotypes that preferentially opsonize and fix complement in horses) are dramatically enhanced following intrabronchial challenge with virulent R. equi. This occurs in conjunction with clearance of the bacteria from the lungs.¹³⁰ In foals, production of antibody to VapA and VapC, but not that to other Vaps, increases fol­lowing natural exposure to R. equi.¹²⁹ Characterization of the subisotype response in naturally infected foals revealed mainly an increase in IgGa, IgGb, and IgG(T).^129,131,132 Experimental infection of young foals with a low inoculum of virulent R. equi that resulted in subclinical pulmonary lesions caused a marked increase in serum IgGa and IgGb levels, resulting in concentrations that were significantly higher than those of adult horses undergoing the same challenge.¹³³ Infection of foals of the same age with a higher inoculum leading to severe respiratory disease resulted in a switch from a predominant IgGa response to a striking IgG(T) response, indicating that the size of the R. equi inoculum can modulate immune responses.¹³⁴

CELL-MEDIATED IMMUNITY. Because of the facultative intracellular nature of R. equi, cell-mediated immune mecha­nisms are thought to be of major importance in resistance to infection. A large part of the knowledge of cell-mediated immunity to R. equi infections comes from infection of mice. Deficiencies in the complement component C5 and in natural killer (NK) cells in mice do not impair the pulmonary clearance of virulent R. equi.^v3 In contrast, functional T lymphocytes are absolutely required for the clearance of virulent (plasmid- and VapA-positive) R. equi in mice.^136-138 However, athymic nude mice (lacking functional T lymphocytes) clear plasmid- cured derivatives from their lungs within 1 week of infection, suggesting that, as opposed to virulent organisms, clearance of avirulent plasmid-negative strains in mice does not require functional T lymphocytes and depends mainly on innate defense mechanisms.¹³⁷

The two major mechanisms by which T lymphocytes mediate clearance of intracellular pathogens are secretion of cytokines and direct cytotoxicity. Although both CD4+ (helper) and CD8+ (cytotoxic) T cells contribute to host defense against R. equi in mice, CD4+ T lymphocytes play the major role and are absolutely required for complete pulmonary clearance.^138-140 Studies in mice have clearly shown that a type I response, characterized by IFN-γ production by T-helper lymphocytes, is sufficient to effect pulmonary clearance of R. equi, whereas a type II response, characterized by IL-4 production, is detrimental.^136,141

As opposed to foals, adult horses are typically resistant to R. equi infections. Thus immune adult horses have been used as a relevant model to better understand what responses are necessary for immunologic protection. Clearance of virulent R. equi in adult horses is associated with a significant increase in BAL fluid CD4+ and CD8+ lymphocytes, lymphoproliferative responses to R. equi antigens, development of R. equi-specific cytotoxic CD8+ T lymphocytes (CTLs), and IFN-γ induction by CD4+ and CD8+ lymphocytes.^130,142,143 R. equi-specific CTLs, which are apparently present in all immune adults, are major histocompatibility complex (MHC) class I-unrestricted and appear to recognize unique bacterial lipids from the cell wall.^143,144 The current thought is that these lipid antigens might be presented to T lymphocytes via the CD1 system, as has been well described in M. tuberculosis.¹'⁴⁵ How these findings in mice and adult horses relate to the foal is an area of active research.

IMMUNITY TO R. EQUI IN FOALS. Virtually all newborn mammals are immunologically immature and have an assortment of immunologic deficits compared with older animals. In general, neonates and perinates have diminished innate immune responses and decreased antigen-presenting cell function and are less able to mount type I immune responses.¹⁴⁶ As a result, neonates of various species demonstrate an increased susceptibil­ity to certain infections.

A number of relative immunologic deficits have been demonstrated in foals. In conjunction with the lack of immu­nologic memory in newborns, these deficits are postulated to account for the unique age-associated susceptibility of foals to rhodococcal pneumonia. Age-related deficiencies in R. equi-specific CTL activity has been documented in 3-week-old foals.¹⁴³ Activity of CTLs is improved by 6 weeks of age and is similar to that of adult horses by 8 weeks.¹⁴³ Antigen­presenting cells from foals have significantly lower CD1 and MHC class II expression compared with adult horses.^145,147 In addition, several studies have demonstrated that the ability of equine lymphocytes and neutrophils to produce and/or upregulate various cytokines is strongly influenced by age.^146-151 In particular, the finding that young foals are deficient in their ability to produce IFN-γ in response to mitogens has led to the hypothesis that an IFN-γ deficiency and type II bias might be at the basis of their peculiar susceptibility to R. equi infec­tions.^148,149 However, foals are also deficient in their ability to produce IL-4 in response to stimulation with mitogens and stimulation with R. equi and after vaccination with a killed adjuvanted vaccine, suggesting that a clear polarization toward a type II response is unlikely in neonatal foals.^152-155 Consistent with these findings, experimental infection of young foals with virulent R. equi results in IFN-γ induction and antibody responses similar to, or greater than, those of adult horses undergoing the same experimental challenge.¹³³

Several studies have investigated various immunostimulants that might enhance host defense mechanisms during the rela­tively narrow period of susceptibility to R. equi. Inactivated Parapoxvirus ovis, Propionibacterium acnes, and unmethylated CpGs all enhance ex vivo or in vitro phagocytic cell function or cytokine induction in foals.^156-159 However, despite successfully enhancing IFN-γ production in foals, inactivated P ovis failed to decrease the cumulative incidence of pneumonia at an R. equi-endemic farm. In the same study, IFN-γ and IL-4 secretion at birth was not associated with subsequent development of pneumonia.¹⁶⁰

Spontaneous resolution of R. equi pneumonia after experi­mental challenge has been recognized.^82,161,162 Many foals on farms where the disease is endemic do not develop disease or develop subclinical disease that resolves without interven­tion.^7,163,164 In addition, intragastric administration of live, virulent R. equi to newborn foals confers complete protection against subsequent heavy intrabronchial challenge.^79,165 Oral inoculation with virulent R. equi results in accelerated develop­ment of R. equi-specific CTL,¹⁶⁶ providing a potential mecha­nism for the protection conferred by oral inoculation. Collectively, these findings unequivocally demonstrate that most foals have the ability to mount protective immune responses to R. equi. The basis for the peculiar susceptibility of foals to infection with R. equi is likely complex and multifacto­rial rather than involving a simple and single explanation. The number of virulent bacteria in the environmental air, the age of the foal, the level of prior exposure to R. equi, and genetic susceptibility likely play a role. Genetic susceptibility to R. equi is likely complex, with modest contributions from numerous genes involving various biological pathways and processes.¹⁶⁷

■ Diagnosis The distinction between lower respiratory tract infections caused by R. equi and those caused by other pathogens is problematic, especially at farms without previous history of R. equi infections. Measurement of WBC or fibrinogen concentrations, ultrasonography, and radiography may help raise the degree of suspicion that pneumonia in a given foal may be caused by R. equi rather than another microorganism. However, the definitive diagnosis of bronchopneumonia caused by R. equi should be based on bacteriologic culture or amplifica­tion of the vapA gene by PCR from a TBA obtained from a foal with (1) clinical signs of lower respiratory tract disease,

(2) cytologic evidence of septic airway inflammation, and/or

(3) radiographic or ultrasonographic evidence of bronchopneu­monia. Amplification of vapA by PCR is done preferably in conjunction with bacterial culture because the former test does not permit identification of other bacterial pathogens and in vitro antimicrobial susceptibility testing of R. equi isolates.¹⁶⁸ The definitive diagnosis of extrapulmonary infections (e.g., abdominal abscess, osteomyelitis) caused by R. equi must rely on bacteriologic culture or PCR amplification of vapA from samples from the site of infection. The diagnosis of extrapul- monary disorders from sites at which R. equi cannot be detected (e.g., uveitis or polysynovitis) should be based on isolation of R. equi from TBA or other primary sites of infection. The diagnosis of enterocolitis caused by R. equi is problematic because isolation of R. equi from feces cannot be taken as evidence of enterocolitis caused by R. equi.⁴⁶⁹

CLINICAL LABORATORY TESTS. Hyperfibrinogenemia is the most consistent laboratory finding in foals with R. equi pneu­monia, although rare cases may have normal fibrinogen concentrations. Neutrophilic leukocytosis with or without monocytosis is also common. One study showed significantly higher fibrinogen concentrations and WBC counts in non­survivors than in survivors,¹⁶⁹ whereas other studies showed no difference between the two groups.^12,170 In one study, WBC count higher than 20,000 cells∕μL, fibrinogen concentration higher than 700 mg/dL, and evidence of pulmonary abscessation were more likely to be found in foals with pneumonia caused by R. equi than in foals with pneumonia caused by other bacteria.¹⁷¹ However, there is considerable overlap in distribu­tions, which precludes the use of fibrinogen concentrations and WBC counts for diagnosis or prognosis for an individual foal.¹⁷¹ Similarly, serum amyloid A (SAA) concentrations are significantly higher in pneumonic than in healthy foals.¹⁷² However, many foals with severe R. equi pneumonia have SAA concentrations within the reference interval, and there is a poor association between SAA concentration and the severity of pulmonary disease as assessed by radiography.¹⁷²

IMAGING TECHNIQUES. Thoracic radiography is useful in evaluating the severity of pneumonia and in assessing response to therapy. A prominent alveolar pattern characterized by ill-defined regional consolidation is the most common radio­graphic abnormality.¹⁶⁹ The consolidated lesions are often seen as more discrete nodular and cavitary lesions consistent with pulmonary abscessation (Fig. 31.15). Severity of alveolar pattern and number of cavitary lesions are the radiographic findings significantly associated with a poor outcome in foals with R. equi pneumonia.¹⁷³ Although nonsurvivors tend to have more severe radiographic lesions than survivors, many survivors have very severe radiographic lesions; thus radiographs should not be used as the sole criterion for prognostication and euthana­sia.^12,170,173 Ultrasonography is a helpful diagnostic tool when lung involvement includes peripheral areas but may not be as useful as radiography to evaluate the full extent of lung lesions because abscesses with overlying aerated lung will not be detected by ultrasonography. However, in most horses and foals with pulmonary abscessation the periphery of the lung is affected, enabling the ultrasonographer to successfully image some of the abscesses.¹⁷⁴ Early ultrasonographic lesions are nonspecific and may include only irregularities of the pleural surface. These lesions may progress to form focal areas of consolidation of various sizes (Fig. 31.16, A). In more chronic cases, well-circumscribed, encapsulated abscesses can be detected (Fig. 31.16, B). Ultrasonography is very useful in evaluating the severity of pneumonia and in assessing response to therapy, especially for equine practitioners who do not have access to thoracic radiography. Ultrasonography is also a useful tool for detecting some abdominal abscesses (Fig. 31.17) and for screening for R. equi-infected foals on farms where the disease is endemic (see the Control and Prevention of R. equi Infections on Farms Where the Disease Is Endemic section later). Ultrasonographic or radiographic detection of lung abscesses raises the degree of suspicion that pneumonia in a given foal is caused by R. equi. However, detection of pulmonary abscesses, while commonly used as a screening test (see the Screening for Earlier Detection of Affected Foals section later), is not a definitive diagnostic test.

SEROLOGY. Currently available serologic tests should not be used for the diagnosis of pneumonia caused by R. equi. Independent studies evaluating the performance of serologic tests available for diagnosis of infection caused by R. equi at endemic farms have demonstrated that these tests have low sensitivity, low specificity, or both.^175-178 Improving either sensitivity or specificity of ELISA assays by changing the cutoff value corresponding to a positive or negative test result could only be done to the detriment of the other. The presence of antibodies indicates exposure, subclinical infection, or maternal transfer of antibodies, but it does not necessarily indicate infection leading to clinical disease.

FIG. 31.15 Thoracic radiograph from a 2-month-old foal with pneumonia caused by Rhodococcus equi. The study shows a combination of alveolar and nodular interstitial patterns as well as cavitary nodular infiltrates, primarily in the cranioventral and caudoventral thorax. A mild interstitial infiltrate is present in the caudodorsal lung. The trachea is displaced dorsally, suggesting tracheobronchial lymphadenopathy. There is slight narrowing of the tracheal lumen at the thoracic inlet, likely indicating the presence of tracheal secretions.

FIG. 31.17 Sonogram of the abdomen obtained from a 3-month-old Thoroughbred foal with severe pneumonia caused by R. equi and multiple abdominal abscesses. The sonogram shows an abdominal abscess (arrow). The pneumonia resolved with therapy, but the abdominal abscesses persisted. The foal was eventually euthanized because of weight loss and intermittent episodes of colic. Postmortem examination confirmed resolution of pneumonia and the presence of abdominal abscesses caused by Rhodococcus equi.

FIG. 31.16 A, Sonogram of the right thorax in a foal with mild pneumonia caused by Rhodococcus equi. There is an approximately 1 cm² area of focal consolidation. B, Sonogram of the left thorax in a foal with severe pneumonia caused by R. equi. There is consolidation of the ventral aspect of the lung surrounding a large encapsulated abscess.

CYTOLOGY, CULTURE, AND PCR AMPLIFICATION. Bacte­riologic culture or PCR amplification of vapA from a TBA are the only acceptable ways of establishing a diagnosis of R. equi pneumonia. In one study, only 7 of the 11 (64 %) foals with positive R. equi culture at necropsy and 57 of the 89 (64 %) foals with radiographic evidence of lung abscessation yielded R. equi from culture of a TBA.¹⁷⁹ However, in two other studies, all 17 foals in which R. equi was isolated from lung parenchyma at necropsy yielded the organism from culture of TBAs obtained antemortem, suggesting that culture of a TBA is a valid method for diagnosing R. equi pneumonia.^180,181 Larger case series are required to more accurately estimate the sensitivity of TBA culture for diagnosis of R. equi pneumonia in foals. Specificity of the test must also be considered. Foals without clinical disease may have R. equi in their trachea as a result of subclinical disease or, incidentally, from inhaling R. equi in contaminated environments. At a farm with endemic R. equi pneumonia, 77 of 216 (35%) foals sampled had positive TBA cultures but remained free of clinical signs of respiratory disease throughout the season.¹⁸² For this reason, culture or PCR amplification of R. equi from a TBA should always be interpreted in the context of cytologic and clinical findings. Detection of R. equi from a foal without clinical signs of respira­tory disease, cytologic evidence of septic airway inflammation, or ultrasonographic or radiographic evidence of pulmonary lesions is likely an incidental finding, due to the ubiquitous nature of the organism on horse farms. Amplification of vapA by PCR has been shown to be more sensitive than bacterial culture in most but not all studies.^183-185 However, increased sensitivity may also result in a higher incidence of false-positive results due to the detection of very small numbers of R. equi present as environmental contaminants. Moreover, culture offers the advantage of detecting other bacterial pathogens present and permits in vitro susceptibility testing of recovered pathogens. As a result, PCR amplification of vapA may be done in conjunction with, but should not replace, bacterial culture. The use of PCR assays based on genes other than vapA is not recommended, as these assays would also detect environmental isolates lacking the virulence plasmid, which are not known to cause disease in foals.

Isolation or detection of R. equi from nasal or fecal swabs cannot be taken as evidence of disease due to infection with R. equi. R. equi can be cultured from the feces of healthy horses even if they live at farms without history of R. equi pneumonia.^90,186,187 Quantitative culture of the feces of foals at weekly intervals has been advocated to aid in early diagnosis of R. equi infections because the fecal concentration of R. equi increased at the same time as clinical signs of respiratory disease appeared in a few foals.¹⁸⁷ Similarly, concentrations of R. equi in feces at the time of clinical diagnosis, as assessed by quantitative PCR amplification of the vapA gene, were significantly higher in clinically affected foals than in healthy foals or foals with subclinical lesions at the same farm.¹⁸⁸ However, the assay is not commercially available and it is unknown if the cut-point used in the aforementioned study would be applicable to foals at other farms. Quantification of R. equi in a single fecal sample from a foal has little diagnostic value because of individual and farm-to-farm variation in the number of R. equi in feces.^90,186,187 Furthermore, a negative fecal culture may not be helpful in excluding R. equi infection because only 5 of 30 (17%) foals with confirmed R. equi pneumonia had positive fecal cultures.¹⁸² Similarly, bacterial culture and PCR amplification of nasal 183184189 swabs are insensitive for diagnosis of R. equi pneumonia.¹⁸³,¹⁸⁴,¹⁸⁹

■ Treatment Because ultrasonographic lesions in many foals that are clinically healthy will resolve without treatment (see the Screening for Earlier Detection of Affected Foals section later), it is important to consider separately treatment of subclinically affected foals and foals with clinical disease.

FOALS WITH SEVERE CLINICAL PNEUMONIA. A wide variety of antimicrobial agents are active against R. equi in vitro. However, many of these drugs are reported to be ineffective in vivo, possibly because of poor cellular uptake and resulting low intracellular concentrations. Randomized, controlled clinical trials comparing the efficacy of various antimicrobial treatments in foals with severe clinical pneumonia caused by R. equi are lacking. The only data available regarding treatment compari­sons are from retrospective cohort studies and studies using historical controls. In a retrospective cohort of 48 foals referred to a veterinary teaching hospital between 1978 and 1985, all 10 foals treated with erythromycin-rifampin survived compared to 2 of 4 with TMS, 2 of 6 with chloramphenicol, 0 of 4 with oxytetracycline, and 0 of 17 with penicillin-gentamicin.¹² In the aforementioned study, the combination of erythromycin­rifampin was introduced in 1981, and it is unknown if disease severity influenced treatment allocation. In another report, 50 of 57 (88%) foals with culture-confirmed pneumonia caused by R. equi and treated with erythromycin and rifampin survived between 1981 and 1986. These reports, in association with the previously reported in vitro activity and synergism of erythromycin and rifampin,¹⁹¹ became the basis of recommended therapy with erythromycin and rifampin. Since the early 2000s, azithromycin and clarithromycin have replaced erythromycin in the combination with rifampin because of their higher oral bio availabilities and considerably higher concentrations in pulmonary epithelial lining fluid (PELF) and bronchoalveolar cells.^192-196 The recommended dosages are listed in Table 31.4. Several formulations of erythromycin are commercially available. Although they all show slight differences in bioavailability and elimination, they all result in therapeutic concentrations at recommended dosages. However, the bioavailability of eryth­romycin in foals is considerably lower when foals are not fasted (mean ± SD: 26 ± 15% when fasted and 8 ± 7% when fed).¹⁹⁴

Macrolides and rifampin are highly active against R. equi 197198

in vitro but only exert bacteriostatic activity.¹⁹',¹⁹⁸ As a result, macrolides and rifampin exert time-dependent activity against R. equi in vitro.¹⁹⁸ Of the three macrolides listed previously, clarithromycin is the most active against R. equi in vitro and 198199

against intracellular R. equi in cell culture.¹⁹⁸,¹⁹⁹ Ihe minimum inhibitory concentrations at which 90% of R. equi isolates are inhibited (MIC₉₀) are 0.06, 0.5, and 1.0 μg∕mL, for clarithro- 200201 mycin, erythromycin, and azithromycin, respectively.²⁰⁰,²⁰¹ Ihe combination of a macrolide and rifampin is synergistic both in vitro and in vivo, and the use of the two classes of drugs in combination reduces the likelihood of R. equi resistance to either drug.^191,197,198

Randomized, controlled, prospective studies comparing the relative clinical efficacy of erythromycin, clarithromycin, and azithromycin in pneumonic foals have not been reported. In a retrospective study, the combination clarithromycin-rifampin was significantly more effective than erythromycin-rifampin or azithromycin-rifampin, especially in foals with severe pulmonary lesions.²⁰² Although data analysis was adjusted for disease severity and age, these results must be interpreted with caution because foals were not randomly allocated to treatment groups and because of potential biases inherent to retrospective data. Nevertheless, these data are the best available evidence to guide macrolide selection in foals with R. equi pneumonia.

The discovery that concurrent therapy with rifampin considerably decreases plasma, PELF, and bronchoalveolar cell 203204 concentrations of clarithromycin³, has led to questions about the value of the combination. Despite the profound decrease in clarithromycin bioavailability caused by concurrent admin­istration with rifampin, concentrations of clarithromycin in PELF and bronchoalveolar cells are well in excess of the MIC₉₀ (0.06 μg∕mL) and even the concentration of clarithromycin that prevents emergence of resistant mutants (0.24 μg∕mL when combined with rifampin).^203,204 These findings likely

■ TABLE 31.4

Recommended Dosages, Oral Bioavailability, and Serum Half-Lives of Antimicrobial Agents Commonly Used for the Treatment of R. equi Infections in Foals

^aErythromycin base, stearate, phosphate, estolate, or ethylsuccinate. ^bMean (± SD) bioavailability with ad libitum access to milk and hay. ^cMean (± SD) oral bioavailability after fasting.

^dAdministration q24h for 5 days and q48h thereafter.

IV, Intravenous; NA, not applicable; PO, by mouth.

explain the apparent clinical efficacy of the combination of macrolides with rifampin despite the decrease in bioavailability. Administration of rifampin 4 hours after administration of clarithromycin results in a statistically significant improvement in the bioavailability of clarithromycin, but the modest improve­ment is unlikely to be of clinical relevance.²⁰⁵

Studies in immunodeficient mice indicate that the combina­tion of a macrolide with rifampin is superior to either drug used in monotherapy. In one study, the combination of erythromycin with rifampin was significantly more effective than either drug used alone.²⁰⁶ Given that concurrent admin­istration of rifampin considerably decreases bioavailability of clarithromycin in foals, a similar mouse infection model was used to compare the in vivo activity of clarithromycin and rifampin alone or in combination. Despite significantly lower clarithromycin concentrations in mice treated with the combina­tion, treatment with clarithromycin and rifampin in combination significantly decreased the number of R. equi in the organs of nude mice compared to treatment with either drug alone.²⁰⁷ The numbers of R. equi in the organs of mice treated with clarithromycin or rifampin alone were either higher or not significantly different from that of saline controls.²⁰⁷

A well-designed, large-scale clinical trial is needed to determine the benefit (or detriment) of combining a macrolide with rifampin for treating foals with severe R. equi pneumonia. However, such a study is unlikely to be performed because of the very large sample size needed and the logistical and financial requirements involved. In the meantime, we have more than 30 years of experience and retrospective data as well as animal models documenting the superiority of the combination of a macrolide with rifampin, compared to a complete lack of evidence that macrolide monotherapy is effective in foals with severe clinical pneumonia. Until there is solid evidence demonstrating that macrolide monotherapy is not inferior, the combination of a macrolide (erythromycin, azithromycin, or clarithromycin) with rifampin remains the recommended treatment of choice.^168,208

FOALS WITH MILD TO MODERATE SUBCLINICAL LESIONS. Only studies enrolling foals with lesion scores of 10 cm or greater (see the Screening for Earlier Detection of Affected Foals section later) have documented a significant difference in recovery between foals treated with antimicrobials (i.e., azithromycin with rifampin) and the placebo groups.^164,209 Therefore apparent recovery of foals with mild subclinical lesions with various treatments in uncontrolled studies cannot be taken as evidence of efficacy. In two separate, blinded, placebo-controlled studies, azithromycin alone (10 mg/kg PO q24h) or gamithromycin alone (6 mg/kg IM q7d) were found to be noninferior to the combination azithromycin and rifampin.^164,209 Monotherapy with tulathromycin also was more effective than a placebo, but not as effective as the combination of azithromycin with rifampin.²¹⁰ Despite a statistically sig­nificant difference between macrolide monotherapy and the control group in these studies, it is important to note that up to 78% of the foals receiving the placebo recovered without the need for antimicrobial treatment.²⁰⁹ Therefore the apparent efficacy of macrolide monotherapy in foals with subclinical ultrasonographic lesions should not be interpreted as evidence of similar efficacy in foals with severe pneumonia. It is also unknown if macrolide monotherapy is more likely to induce and select for resistance than combination therapy in vivo. In vitro evidence indicates that resistance is less likely to occur with the combination of macrolide with rifampin.²¹¹

TREATMENT OF FOALS WITH MIXED BACTERIAL INFEC­TIONS. Various bacteria and fungi are often isolated from TBA along with R. equi. The most common concurrent bacterial isolates are β-hemolytic streptococci and E. coli²¹² Isolation of multiple types of bacteria or fungi from a TBA does not negatively affect survival, and mixed infections are significantly more likely to be encountered in TBA than in lung tissue, suggesting that many bacteria isolated from a TBA colonize the trachea without necessarily contributing to pulmonary pathology.²¹² Some clinicians advocate the addition of a third antimicrobial agent if heavy growth of a gram-negative pathogen is isolated along with R. equi. Retrospective data indicate that the use of additional antimicrobial agents with better gram­negative coverage (i.e., gentamicin or ceftiofur) is not associated with increased survival relative to the use of a macrolide with rifampin in foals from which both R. equi and gram-negative bacteria are isolated from a TBA.²¹² The combination of an aminoglycoside (gentamicin or amikacin) with macrolides or rifampin in vitro is antagonistic against R. equi compared with either drug alone. ^,, However, the clinical significance of this in vitro finding has not been established.

TREATMENT OF FOALS INFECTED WITH ISOLATES OF R. EQUI RESISTANT TO MACROLIDES AND/OR RIFAMPIN. Although the vast majority of R. equi isolates from foals are highly susceptible to macrolides and rifampin in vitro, strains resistant to either drug class have been encountered. Most foals infected with R. equi that do not respond to therapy are infected with isolates susceptible to macrolides and rifampin. Therefore failure to respond to antimicrobial therapy does not provide evidence that a given foal is infected with an isolate resistant to antimicrobial agents. The presence of an isolate resistant to macrolides or rifampin can only be confirmed by in vitro susceptibility testing done by a laboratory following the criteria established by the Clinical and Laboratory Standards Institute (CLSI).

In a recent study, the overall prevalence of macrolide- and rifampin-resistant isolates in Texas and Florida over a 10-year period was 4%.²¹³ In the same study, the odds of death were seven times higher in foals infected with resistant isolates.²¹³ In addition, the study demonstrated that isolates of R. equi susceptible to macrolides were sometimes misclassified as resistant; therefore it is reasonable to request retesting or validation of resistance by the testing laboratory. More recently, it has been documented that mass antimicrobial treatment of subclinically affected foals has selected for antimicrobial resistance over time, with isolates of R. equi resistant to all macrolides and rifampin being cultured from as many as 40% of the foals at some farms.¹⁰⁴ Macrolide resistance in R. equi is conferred by acquisition of a newly identified ribosomal RNA methylase gene designated erm(46)²¹⁴ This gene confers resistance to all macrolides, lincosamides, and streptogramins type B but not to other classes of antimicrobial agents.²¹⁴ The gene can be transferred from resistant to susceptible isolates of R. equi by conjugation.²¹⁴ Most macrolide-resistant isolates of R. equi that have been identified to date are also resistant to rifampin. Rifampin resistance in R. equi is the result of mutations in the rpoB gene.^201,215-217

Therapy of foals infected with isolates confirmed to be resistant to macrolides, rifampin, or both is problematic because of the limited range of effective alternatives. Macrolide- and rifampin-resistant isolates of R. equi are susceptible in vitro to fluoroquinolones, gentamicin, linezolid, and vancomycin.²¹³,²¹⁸,²¹⁹ In one study, 18 of 24 isolates were also susceptible to chlor­amphenicol, minocycline, and TMS.²¹³ Currently there are no data to indicate the preferred antimicrobial agents for the treatment of foals infected with isolates resistant to macrolides and rifampin. Vancomycin, imipenem, or linezolid in foals should be used only for the treatment of life-threatening R. equi infections caused by isolates confirmed to be resistant to all other possible alternatives. The use of systemic and/or nebulized gentamicin is a reasonable consideration for foals infected with macrolide- and rifampin-resistant isolates. Histori­cally, the efficacy of gentamicin in foals infected with R. equi has been widely reported as being poor. This belief is based on the results of a retrospective case series of foals with pneumonia caused by R. equi in which all 17 foals treated with gentamicin died, whereas all 10 foals treated with erythromycin in combination with rifampin survived.¹² In contrast, in another retrospective case series of 39 foals with pneumonia caused by R. equi, all 19 survivors were treated with gentamicin, whereas nonsurvivors were treated with a variety of other antimicrobial agents, including erythromycin, kanamycin, or chlorampheni­col.¹⁶⁹ These studies were not controlled and did not account for lesion severity at the time of initiation of therapy. In addition, the dosages of gentamicin used in those studies were lower than dosages currently recommended based on the fact that gentamicin effectiveness is now known to be concentration dependent. Gentamicin is one of the few drugs that is bacte­ricidal against R. equi, and it is highly active against intracellular R. equi.²²⁰ Intravenous or nebulized gentamicin administered at a dose of 6.6 mg/kg once daily results in pulmonary epithelial and bronchoalveolar cell concentrations well in excess of the MIC₉₀ for isolates of R. equi. Additional studies will be necessary to assess the clinical efficacy of gentamicin intravenous or nebulized in foals infected with R. equi.

DURATION OF THERAPY. Resolution of clinical signs and radiographic or ultrasonographic resolution of lung lesions are commonly used to guide the duration of therapy, which generally ranges between 2 and 12 weeks, depending on the severity of the initial lesions and response to therapy. Foals treated based on subclinical lesions identified during ultraso­nographic screening typically do not require as long a treatment period as foals in respiratory distress with severe pulmonary lesions. In fact, as addressed later, many foals with subclinical ultrasonographic lesions clear the infection without therapy.^7,163

ADVERSE EFFECTS. Although well tolerated by most foals, macrolides commonly cause diarrhea. The diarrhea is often self-limiting and does not necessitate cessation of therapy. However, affected foals should be monitored carefully because some may develop severe diarrhea, leading to dehydration and electrolyte losses that necessitate intensive fluid therapy and cessation of oral macrolides. The incidence of diarrhea in foals treated with erythromycin-rifampin has ranged between 17% and 36%.^202,221 In one study, foals treated with clar­ithromycin had a higher incidence of diarrhea (28%) than those treated with azithromycin (8%), although in most cases diarrhea was mild and self-limiting.²⁰² In the same study, the incidence of severe diarrhea necessitating administration of IV fluids was not significantly different between groups of foals treated with azithromycin-rifampin, clarithromycin-rifampin,

202

or erythromycin-rifampin.²⁰²

Hyperthermia and tachypnea have been described in foals treated with erythromycin during periods of very hot or humid weather.²²¹ Anecdotal reports suggest that these reactions occur occasionally with newer macrolides as well. Recent studies demonstrate that azithromycin, clarithromycin, and erythro­mycin suppress sweating in foals, with erythromycin inducing 222223

more severe sweating dysfunction.²²²,²²³ In contrast, rifampin does not impair sweating.²²²

Severe enterocolitis also has been reported in mares whose foals are treated with erythromycin, presumably caused by disruption of the mare's normal colonic microflora following ingestion of small amounts of active drug during coprophagia or by contamination of feeders or water buckets with drug present on the foal's muzzle. This complication seems to be rare. Enterocolitis in mares has been reproduced experimentally by administration of subtherapeutic doses of erythromycin.²²⁴ In some cases, the severe enterocolitis in the mares of treated foals is associated with Clostridium difficile²²⁵

ANCILLARYTHERAPIES. Nursing care, provision of adequate nutrition and hydration, and maintaining the foal in a cool and well-ventilated environment are important. Humidified oxygen delivered by pharyngeal insufflation in moderately hypoxemic foals or by percutaneous transtracheal oxygenation in severely hypoxemic patients is indicated.²²⁶ Judicious use of NSAIDs might reduce fever and improve attitude and appetite in febrile, lethargic, anorectic foals. Nebulization with saline, antimicrobial agents, or bronchodilators have been advocated, but there are no data to either support or refute these thera­peutic practices. In addition to appropriate systemic antimi­crobial therapy, foals with R. equi septic arthritis or osteomyelitis often require aggressive local therapy such as joint lavage, surgical debridement, and IV or intraosseous regional limb perfusion with antimicrobial agents.

■ Prognosis Prior to the introduction of the combination of erythromycin and rifampin as the treatment of choice in the early 1980s, the prognosis of R. equi-infected foals was poor, with survival rates as low as 20%.²²⁷ Using erythromycin and rifampin, Hillidge reported a successful outcome (as assessed by survival) in 50 of 57 (88%) foals with confirmed R. equi pneumonia.¹⁷⁹ Studies from referral centers, where severely affected cases are likely more prevalent, have revealed survival proportions ranging between 59% and 72%.^170,202,228 The prognosis for foals with abdominal abscesses is poor, although rare cases will respond to long-term antimicrobial therapy.^8,229 Surgical removal or marsupialization have been attempted in some foals, but abdominal adhesions usually result in inability to resect the abscess.

In contrast, the survival rate at farms that use a screening program to identify and treat foals with subclinical lesions has resulted in survival proportions of nearly 100%.^4,23⁰ However, recent evidence indicates that many of these foals would have recovered without therapy.^6,7,163,231 Controlled studies have shown that approximately 88% of foals with ultrasono­graphic lesion scores (sum of the diameter of each abscess) of 10 cm or less recover without antimicrobial therapy. In the same studies, treatment of affected foals with antimicrobial agents did not provide a clear benefit over administration of a placebo.⁷,¹⁶³

The impact of R. equi infections on future athletic perfor­mance has been examined. No significant differences in total earnings, average earning index, or age at first race were observed when comparing 30 horses that previously had R. equi pneumonia with either their dams' other progeny or the North American averages.²³² In another study, 54% of 83 foals that survived R. equi pneumonia had at least one racing start as compared with 65% of their birth cohort, suggesting that horses contracting R. equi pneumonia as foals might be some­what less likely to race as adults. However, consistent with a previous report,²³² the racing performance of those foals was not different from that of the U.S. racing population.¹⁷⁰ In summary, prognosis for performance after successful treatment of uncomplicated R. equi pneumonia should be regarded as very good.

Control and Prevention of R. equi Infections on Farms Where the Disease Is Endemic

Methods for control and prevention include using screening tests for early detection of R. equi pneumonia, environmental management, chemoprophylaxis, and prevention through use of either passive or active immunization. Currently there is inadequate evidence to recommend environmental interventions to control or prevent R. equi pneumonia.²³³ Further evaluation of certain practices, such as foaling at pasture and reducing either airborne concentrations of virulent R. equi or density of mares and foals, is merited.

SCREENING FOR EARLIER DETECTION OF AFFECTED FOALS. R. equi pneumonia is often not recognized until it is well advanced and therefore difficult to treat. To a casual observer, even severely affected foals may appear to suckle and behave normally. The rationale for screening is the assumption that detecting foals in the early stages of disease and implementing appropriate treatment of affected foals will improve outcomes. It is important to emphasize that screening methods are not diagnostic tests. A useful screening test is one in which the probability of disease is high with a positive test result (high positive predictive value) and very low with a negative test result (high negative predictive value). The higher the prevalence of disease is at a given farm, the higher the positive predictive value of a given test will be. Therefore depending on the prevalence of R. equi infections at a given farm, a positive result on a screening test could be a basis from which to perform a diagnostic test (low to moderate prevalence) or to initiate therapy (high prevalence). Uncontrolled studies suggest that screening may reduce the mortality attributable to R. equi pneumonia.^4,234 Thus in the absence of a vaccine or other effective methods for preventing this devastating disease, screening is a reasonable approach to control R. equi at endemic farms. Implementing screening testing at farms that are only sporadically affected by R. equi pneumonia may not be warranted.

A variety of screening techniques performed serially have been described, including visual inspection of foals for clinical signs of pneumonia, rectal temperature monitoring, hematologic variables, serology, and thoracic imaging using either radi­ography or ultrasonography, with empirical recommendation that screening begin around 3 weeks of age.²³⁵ Systematic comparisons of these tests have not been performed. Thus a specific recommendation for any particular screening test cannot be made, and it is likely that the optimal approach for screening may vary among farms on the basis of cumula­tive incidence of disease, resources available for control and prevention, and preferences of the attending veterinarians and farm management.

Studies have shown that serial PCR amplification of vapA in feces and serum measurement of serum concentrations of either antibodies against R. equi, serum amyloid A, or plasma fibrinogen are not useful screening tests.^{172,176,178,236,237} WBC concentrations performed at monthly intervals had clinically acceptable sensitivity and specificity for early detection of R. equi pneumonia at one farm (sensitivity of 79% and specificity of 91% at a cutoff value of 15,000 cells^L).¹⁷⁵ However, serial measurement of WBC or neutrophil concentrations resulted in limited performance for prediction of subsequent develop­ment of clinically apparent R. equi pneumonia at another farm.²³⁸ In addition to conflicting data regarding the benefit of this approach, additional limitations of using WBC concentration for screening are the time and effort required to collect blood samples from foals, the lag from time of submission to avail­ability of results, and the potential for false-positive results attributable to stress-associated leukocytosis or other infectious or inflammatory conditions that may be common among foals.

Over the past decade, control of R. equi infections at many farms where the disease is endemic has relied on early detection of subclinical pulmonary disease using thoracic ultrasonography and initiation of treatment with antimicrobial agents prior to development of clinical signs.^4,5,239 Ultrasonography of the chest offers several advantages over other screening tests: (1) results are specific for the presence of pulmonary pathology; (2) the procedure can be performed relatively quickly for an individual foal; (3) results are available immediately; and (4) the procedure may be more sensitive than radiography for detecting lesions in their early stages of development or in certain regions where soft-tissue structures are superimposed over regions of the lung.¹⁷⁴ Periodic ultrasonography of the chest has been reported to decrease mortality due to R. equi pneumonia at some farms relative to historical controls, but contemporaneously controlled studies are lacking.^4,5,239 Historically, antimicrobial treatment was recommended for all foals with pulmonary lesions 1 cm or greater in diameter.^4,5,239 This approach of ultrasonographic screening resulted in a considerabe increase in the number of foals treated for presumptive R. equi pneumonia. The temporal association between widespread use of macrolides and rifampin consequent to ultrasonographic screening and a perceived increase in the frequency of detection of antimicrobial-resistant isolates of R. equi in the last decade suggests that this practice is not innocuous. Emergence of widespread macrolide and rifampin resistance at a farm after widespread use of these drugs was implemented on the basis of an ultrasonographic screening program has been documented.¹⁰⁴

Several uncontrolled studies have reported the apparent efficacy of various antimicrobial agents in foals with mild subclinical lesions. Recently reported blinded, randomized, placebo-controlled studies, however, indicate that approximately 88% of foals with lesion scores (defined as the sum of the largest diameter of all pulmonary lesions ≥1 cm) between 1 and 10 cm recover without antimicrobial therapy.^7,163 Further­more, antimicrobial treatment of foals with such ultrasono­graphic lesions does not significantly hasten lesion resolution compared to administration of a placebo.^7,163 In another study, 270 foals at an endemic farm were subjected to thoracic ultrasonography every 2 weeks for the entire season. The veterinarian making treatment decisions was unaware of ultrasonographic findings. Forty-six foals (17%) developed clinical signs of pneumonia and were treated; all 46 foals had ultrasonographic lesions.⁶ Fifty-four foals (20%) remained clinically healthy and free of ultrasonographically visible pulmonary lesions, whereas 170 foals (63%) remained clinically healthy but had pulmonary lesions of various degrees of severity.⁶ Optimal cut-point of the sum of the maximal diameters of all lesions observed sonographically at a given examination to maximize sensitivity and specificity in that study was 20 cm. This high sum of lesion diameters as a cut-point for lesion score would have yielded a relatively high sensitivity (89%) but poor specificity (62%) for the diagnosis of clinical pneu­monia.⁶ Despite using such a high cut-point to initiate therapy, the number needed to treat was 3, meaning that approximately three foals would be treated with antimicrobial agents unneces­sarily for every one foal benefiting from therapy.

Because ultrasonography only allows the detection of lesions at the periphery of the lung and because of individual variation in disease susceptibility and host defense mechanism, there is no lesion score cut-point that will perfectly differentiate foals requiring therapy from foals that will spontaneously recover. Although it is probably indicated to treat foals with clinical signs of lower respiratory tract disease, the evidence reviewed earlier indicates that treatment of all foals with small subclinical pulmonary lesions is unnecessary. By being more selective and treating only foals with larger lesion scores (e.g., >10 cm) and monitoring the foals weekly or twice weekly both clinically and ultrasonographically, one farm was able to decrease antimicrobial drug use considerably without an increase in mortality.^209,210 Additional studies at multiple other endemic farms will be required to determine whether this approach should be widely recommended.

PASSIVE IMMUNIZATION. Intravenous administration of HIP obtained from horses vaccinated against R. equi using various antigens has generally proved effective in significantly reducing the severity of R. equi pneumonia in foals following experimental challenge.^{79,127,240,241} However, studies evaluating the efficacy of various HIP preparations under field conditions have given equivocal results. Although the data are conflicting and not all trials have shown a statistically significant reduction in the cumulative incidence of R. equi pneumonia, 5 of 7 studies have demonstrated reduction of relative risk, suggesting some benefit of HIP.³,⁸¹,¹⁸¹,^242-245

Intravenous administration of purified immunoglobulins obtained from horses immunized with recombinant VapA and VapC to foals reduced the severity of pneumonia following high-dose experimental challenge with R. equi to a degree similar to that provided by commercially available HIP.¹²⁹ This demonstrates that immunoglobulins, predominantly against VapA and VapC, are responsible for the protection conferred by HIP. Therefore administration of plasma obtained from horses that are not hyperimmunized against R. equi is not recommended for protection against R. equi. Because endemicity of R. equi at farms does not appear to be explained by farm-specific isolates and because foals may be infected with multiple strains at once,¹⁰² there is no need to administer plasma that is produced from immunizing horses with isolates of R. equi obtained from a given farm; transfusion of plasma from horses that have been immunized against any virulent (i.e., VapA-positive) strain of R. equi should suffice.²³³ Use of HIP licensed as an aid in the control of R. equi pneumonia is recommended (rather than plasma simply obtained from horses hyperimmunized against R. equi) because licensure ensures standards of potency, purity, and safety. Currently there is insufficient information to recommend one brand of licensed antibody product over another.

The optimal amount of plasma to be transfused and the optimal age at which transfusion should occur remain to be determined. Administration of HIP 9 days after aerosol infection of foals with R. equi did not confer protection,²⁴⁶ suggesting that administration of HIP prior to infection is important. Because of evidence that many foals become infected early in life,⁸⁰ it is commonly recommended that foals receive transfusion of at least 1 liter of HIP no later than the second day of life. Because early administration may result in the decline of passively transferred antibody to a nonprotective level at a time when foals are still susceptible to R. equi and when environmental challenge is high, it is common practice to administer a second liter of HIP at 2 to 4 weeks of age.

Transfusion of HIP is not completely effective and therefore does not eliminate the need for careful monitoring of foals at risk. In addition to being incompletely effective, transfusion of HIP carries some risk to foals, such as trauma that may occur during handling and adverse reactions to transfusions. The process is also time intensive, labor intensive, and expensive. Thus the cost-effectiveness of transfusion depends on the value of the foals and the prevalence of disease at a given farm.

ACTIVE IMMUNIZATION. It would be convenient to control R. equi pneumonia on endemic farms by active immunization of mares or foals with a protective antigen. To date, however, this approach has been largely unrewarding. The role of antibody in partial protection against R. equi infection suggests that vaccination of mares could confer at least some degree of protection. However, in both a field study and an experi­mental challenge, vaccination of mares with live or killed whole cell R. equi did not provide protection against R. equi pneumonia despite a significant increase in colostral R. equi-specific antibody and transfer of these antibodies to foals.^244,247 Vaccination of a small number of mares with VapA associated with a water­based nanoparticle adjuvant led to high anti-VapA IgG con­centrations in mares and foals and might have conferred protection against natural challenge compared with nonvac­cinated controls.²⁴⁸ More recently, vaccination of mares against poly-V-acetylglucosamine protected foals against intrabronchial challenge with R. equi.¹¹⁵ Large-scale studies at endemic farms will be necessary to confirm these preliminary findings before widespread vaccination of mares can be recommended.

Because cell-mediated immunity is of paramount importance for protection against R. equi, active immunization of foals likely will be required for complete protection. Oral or nasal immunization with S. enterica serotype Typhimurium expressing the VapA antigen protects mice against R. equi infection.^249,250 Studies in mice indicate that DNA immunization with vapA protects against R. equi infection and that the IgG subisotype response is consistent with a type I-based immune response.²⁵¹ A similar DNA vaccine containing the vapA gene induced strong cell-mediated immune responses in adult horses, but responses were poor in foals.²⁵² Intrabronchial immunization of neonatal foals with a live, fully attenuated, riboflavin auxotroph strain of R. equi stimulated immune responses but did not confer protection against subsequent intrabronchial challenge with live R. equi.²⁵³ The fact that live, avirulent plasmid-cured R. equi does not elicit adaptive immunity and is cleared rapidly indicates that bacterial replication is needed for induction of strong cell-mediated immune responses.²⁵⁴ Intrabronchial immunization with a deletion mutant of R. equi lacking the chromosomal genes isocitrate lyase (icl) and cho­lesterol oxidase (choE) conferred protection against subsequent challenge in three foals.²⁵⁵ However, two foals developed pneumonia caused by the mutant strain.²⁵⁵ More recently, oral vaccination of foals with a live attenuated R. equi strain lacking genes involved in the steroid catabolic pathway elicited sub­stantial protection against subsequent intrabronchial challenge with live virulent R. equi.⁷³ Collectively, the aforementioned studies indicate that there is a fine line between sufficient replication of R. equi for induction of strong cell-mediated immune responses and development of disease from the vaccine strain. Additional challenges are that immunization in foals will need to be initiated very early in life and an effective vaccine will have to overcome the relative immaturity of the naive neonatal immune system. Currently there is inadequate evidence to recommend active immunization of mares or foals to control or prevent R. equi pneumonia, but promising vaccine candidates are under development.

CHEMOPROPHYLAXIS. Because an effective vaccine is lacking, prophylactic administration of antimicrobial agents to all foals during the period when they are most susceptible to infection has been suggested as another approach for the prevention of R. equi infections on endemic farms. Azithromycin is an attractive choice for chemoprophylaxis because of good oral bioavailability, long half-life, and high and sustained concentrations in PELF, bronchoalveolar cells, and neutrophils. Two studies have evaluated the use of azithromycin for che­moprophylaxis. In a randomized, controlled trial conducted in the United States among 338 foals at 10 farms, the cumulative incidence of R. equi pneumonia was reduced from 21% among untreated foals to 5% among foals that received azithromycin (10 mg/kg PO q48h for the first 14 days of life).²⁵⁶ In contrast, in a study conducted among 70 foals at a large breeding farm in Germany, the incidence of abscessing pneumonia was not significantly different between foals that received azithromycin (10 mg/kg PO q24h for the first 28 days of life) for prevention of abscessing pneumonia (cumulative incidence = 60%) and foals that did not receive azithromycin for chemoprophylaxis (cumulative incidence = 69%); however, the age at onset of abscessing pneumonia was delayed in treated foals.²³⁰ The reason for the discrepancy between studies is unknown. Regard­less, use of azithromycin is not considered an acceptable approach for chemoprophylaxis because widespread use of this drug would create greater pressure for emergence of macrolide resistance among bacteria.¹⁶⁸

Gallium maltolate is a metal-based compound with antimi­crobial properties that has been demonstrated to reduce replica­tion of R. equi in pure culture²⁵⁷ and within macrophages,²⁵⁸ to reduce tissue concentrations of R. equi in mice following experimental infection,²⁵⁷ and to be bioavailable²⁵⁹ and safe in foals.²⁶⁰ Unfortunately, chemoprophylaxis with gallium maltolate (30 mg/kg PO q24h for the first 14 days of life) failed to reduce the incidence of R. equi pneumonia in a placebo-controlled trial of 438 foals at 12 farms in the United States.³

Pneumonia in Foals

Daniela Bedenice

Equine pneumonia is a significant cause of morbidity and mortality in both the newborn and the older foal.^1-4 Although the causative agents and the mode of lung infection are generally different for each age-group, foal pneumonia is commonly associated with a compromise in the immunologic protection of the host. The exact immunologic basis of the foal's inherent susceptibility to opportunistic respiratory pathogens needs to be further elucidated.

Failure of passive transfer of immunity (FPT) is a well-known risk factor of neonatal infection, including respiratory disease,⁴ because not only are affected foals deprived of specific maternal antibody protection but their neutrophil function is also seri­ously impaired.⁵ Although foals can respond immunologically in utero to bacterial or viral infections, their ability to do so is less than that of adults. Reduced complement values and defective chemotaxis (directed migration of neutrophils or macrophages) and killing capacity of the neonatal neutrophil contribute to a relatively decreased defense against invading bacteria. Furthermore, an immature ciliary apparatus and the presence of fewer alveolar macrophages in neonates in com­parison with adult horses lead to decreased bacterial clearance from the lungs.⁶ The number and distribution of the bron­choalveolar cells approach adult levels only at approximately 3 to 6 weeks of age.⁷

It has been suggested that a natural cellular immunodefi­ciency may also occur in foals at 2 to 4 months of age.⁸ Although the exact association between age-related cellular immunode­ficiency and the defense against intracellular pathogens such as Pneumocystis carinii (Pneumocystis jiroveci) and R. equi of foals still needs to be clarified, CD4+ and CD8+ T-lymphocytopenia was documented in a filly with P carinii pneumonia.⁹

■ Etiology

NEONATAL PNEUMONIA (identified, suspected causes include viruses, P. carinii, heat stroke, envi­ronmental toxins, and bacteria.^26-29 Enteric gram-negative organisms, R. equi, Pseudomonas aeruginosa, and P. carinii have all been cultured from the lungs of affected foals.

The reported cases are usually sporadic, affecting a single foal within a herd,²⁶ although rare clusters of cases have also been described.²⁸ The disease is generally rapidly progressive and may lead to sudden death as a result of fulminant respiratory failure. Severe respiratory distress, hypoxemia, hypercapnia, and respiratory acidosis are the most striking clinicopathologic signs. The hypoxemia of bronchointerstitial pneumonia is relatively resistant to supplemental oxygen therapy, indicating a pulmonary shunt. Thoracic radiographs commonly demonstrate caudodorsally distributed interstitial and bronchointerstitial pulmonary opacities. With advanced disease, the radiographic pattern progresses to include patches of a coalescing alveolar nodular pattern with air bronchograms.³⁰ As with bacterial pneumonia, many foals demonstrate hyperfibrinogenemia and neutrophilic leukocytosis on hematology.

Therapy is symptomatic, including antiinflammatory therapy, broad-spectrum antibiotics, thermoregulatory control, bron­chodilation, and supplemental oxygen. Surviving foals may develop a proliferative epithelial and interstitial response, including bronchiolar and alveolar epithelial hyperplasia, type II cell hyperplasia, and hyaline membrane formation.³⁰

Chronic pulmonary disease (1 to 3 months' duration) with marked radiographic interstitial opacities was reported in 12 foals, ages 3 to 9 months, that were examined at a veterinary teaching hospital.³¹ Evaluation failed to yield a consistent pathogen from tracheobronchial aspirates. The authors proposed that chronic interstitial pneumonia occurring in foals is associ­ated with a good prognosis and considered corticosteroid therapy as a potentially useful treatment.

PNEUMOCYSTIS JIROVECI (CARINII) INFECTION. Rare opportunistic fungal pathogens of the lung may include P carinii (renamed P jiroveci in humans), Aspergillus spp., Candida spp., and Mucor spp. P. carinii is a unicellular eukaryote that has been classified as a fungus after DNA studies.³² It is most commonly seen in humans and animals that suffer from a concurrent immunodeficiency. It was first recovered in Arabian foals with SCID.³³ It has since been recovered from foals of other breeds with no specific history of immunodeficiency, although for the most part no direct testing for immunocompetence was performed in the reported cases. However, one more recently documented case demonstrated a low number of circulating CD4+ and CD8+ lymphocytes in a filly with P carinii.³³

In several reports the onset of clinical signs in affected foals ranged from a 3-week history of weakness, weight loss, and nasal discharge to acute dyspnea. Half of the reported non-Arabian cases had concurrent infections with R. equi, Enterobacter cloacae, E. coli, or S. equi subsp. Zooepidemicus. A severe interstitial, sometimes miliary and alveolar pattern was seen on radiographs of the lungs. The majority of the reported foals died, and the diagnosis of P carinii was obtained at necropsy.^29,33-36 On postmortem examination the lungs were uniformly heavy and consolidated, and a diffuse interstitial pneumonia was found. Trophozoites and cysts were observed in the alveolar epithelial cells, and macrophages were observed within the alveoli. Fluorescent in situ hybridization, silver staining, and streptavidin-biotin immunolabeling of histologic sections provide better identification of the organisms in the pulmonary tissue.³⁴ A few foals were also diagnosed by visualization of the organism on a BAL sample in the latter report. Two surviving foals responded to TMS therapy.

■ Clinical Presentation and Diagnosis of Foal Pneumonia The clinical manifestation of pneumonia in the newborn foal can be variable and depends on the disease severity and underlying or associated problems. Early in life, localizing clinical signs of respiratory infection may be absent even in the presence of extensive disease. Dyspnea may be seen in severely affected foals and may manifest as an increase in respiratory rate, effort, or thoracoabdominal asynchrony (paradoxical breathing). However, signs of respiratory distress and hypoxemia are frequently vague. Even some severely hypoxemic foals may show only restlessness, considerable resistance, and struggling when being handled or restrained. Abnormal respiratory sounds (crackles or wheezes) may be heard on auscultation of ill foals. However, even normal foals may show crackles on the down lung after having remained in lateral recumbency for a prolonged period of time. Fur­thermore, foals with no auscultable abnormalities may still have extensive pulmonary disease., Similar to other species, cyanosis is a sign of severe hypoxemia in foals. The arterial PaO₂ (partial pressure of oxygen), however, must reach very low levels (35 to 45 mm Hg) before clinical cyanosis is observed. Approximately 5 g/dL of unoxygenated hemoglobin in the capillaries generates the dark-blue color that is appreciated as cyanosis in humans. Therefore severely anemic foals may never appear cyanotic even in the face of profound hypoxemia. Weakness, depression, anorexia, weak or absent suckle reflex, dehydration, and fever may also be noted in newborn foals with respiratory disease. Cough and nasal discharge are usually absent in the early stages of neonatal pulmonary disorders.¹⁰ The older foal, on the other hand, will generally exhibit clinical signs that focus on the respiratory tract, such as abnormal auscultation of the lungs, nasal discharge, cough, fever, tachy­pnea, and increased respiratory effort.

Thoracic radiographs and arterial blood gas analysis are important in confirming the presence and evaluating the extent and distribution of foal pneumonia. Because radiographic findings may precede or lag behind alteration of respiratory function, repeated diagnostic imaging is recommended. A recent study documented that radiographic abnormalities involving the caudodorsal lung region were most common in neonatal foals admitted to a referral center.³⁹ Bedenice and colleagues showed that the presence of dyspnea, tachypnea, a fibrinogen concentration greater than 400 mg/dL, systemic inflammatory response syndrome (SIRS), hypoxemia, or FPT in neonatal foals may be a predictor of underlying respiratory disease and should prompt an early radiographic evaluation of the thorax, even in the absence of other localizing clinical signs.^4,39 Thoracic ultrasonography has become a popular diagnostic tool for the assessment of equine lung and pleural disease and is particularly useful in determining the side(s) of the thorax affected, as well as the precise location of the lesion in many cases.⁴

Diagnostic sampling of the respiratory tract is less commonly performed in neonatal foals compared with older foals and adult horses. However, transtracheal aspirate, BAL, protected catheter brush (PCB), or protected aspiration catheter (PAC) sampling is important in stabilized patients, irrespective of age, to direct antimicrobial therapy and evaluate therapeutic efficacy. Lung biopsies are rarely performed and are limited to patients with chronic disease or suspected fungal pneumonia.³⁸

■ Treatment The treatment goals in dealing with bacterial pneumonia in either age-group are similar and directed at maintaining adequate gas exchange to ensure patient survival, limit progressive pulmonary damage, and ultimately eradicate infection. Long-term goals should aim to maintain optimal pulmonary function and conserve adult athletic performance. Specific treatment strategies for septic pneumonia may include goal-directed antimicrobial therapy, treatment of inflammation independent of antimicrobial therapy, respiratory therapy (oxygen therapy, ventilation, bronchodilation), and cardiovas-

1 ₃7 ^,

cular support.³'

Broad-spectrum antimicrobial treatment, with consideration of the drugs’ anticipated sensitivity and lung penetration, should be initiated before culture results are available. Third- or fourth-generation cephalosporins (ceftiofur and cefepime, respectively) show good penetration into lung tissues.⁴⁰ Ceftiofur administered intramuscularly at 6.6 mg/kg in weanling foals provided plasma and pulmonary epithelial lining fluid concentra­tions above the therapeutic target of 0.2 μg∕mL for at least 4 days, and would thus be expected to be an effective treatment for pneumonia caused by susceptible pathogens such as S. equi subsp. Zooepidemicus.⁴ Alveolar delivery of antibiotics typically occurs via diffusion of a free, nonprotein-bound drug and is usually satisfactory if plasma concentrations and alveolar perfusion are adequate.⁴² Aminoglycosides penetrate lung tissue at 10% to 45% of serum levels. However, the acidic environment of infected airways may reduce the drug’s activity below levels sufficient to eradicate the offending bacteria in some cases.^43,44 Antibiotic therapy should not be discontinued prematurely in neonatal foals because pulmonary infection is commonly well established before its diagnosis.⁴⁵

■ Outcome The survival of neonatal foals with septic pneumonia is contingent on the underlying disease conditions, a prompt diagnosis, and the cause and severity of lung dysfunction. Few studies have objectively evaluated the outcome of foal pneumonia.³⁸ One report documented that 65% of foals with radiographic evidence of pulmonary disease were discharged alive from the hospital, in contrast to 86% survival of foals without respiratory illness. The presence of diffuse radiographic infiltrates (caudodorsal, caudoventral, and cranioventral lung involvement) or concurrent alveolar patterns within the caudodorsal and caudoventral lung indicated lower survival rates.³⁸ Pneumonia is also considered one of the greatest causes of morbidity and mortality in foals older than 1 month of age.² The long-term effects on athletic performance have thus far predominantly been evaluated in older foals that survived Rbodococcus pneumonia. R. equi infection in foals was associated with a decreased chance of racing as adults. However, the performance of foals that went on to race was not significantly different from that of the general U.S. racing population.⁴⁶ It remains speculative whether severe equine neonatal pneumonia will most likely limit future peak athletic performance. Human studies have shown that many hosts develop long-lasting or permanent lung changes after recovery from neonatal pneu­monia, which may significantly affect quality of life and sus­ceptibility to later infections. Case-based evaluations in horses have demonstrated that functional and structural pulmonary abnormalities may occur in survivors of severe neonatal pneumonia. However, the clinical significance of this finding currently remains unclear.