Smoke Inhalation

Peggy S. Marsh

Smoke inhalation injury is typically associated with exposure to fires, and there are often concurrent problems in other body systems from thermal injury. Thermal injury, along with smoke inhalation, leads to both local and systemic lesions.

Insult to the respiratory system by smoke inhalation depends on the fuels that burned, the completeness of combustion, the heat intensity, and the duration of exposure. In general, lesions are initiated by three mechanisms. The first is direct thermal injury, which is often limited to the upper respiratory tract by laryngeal reflexes and efficient heat exchange within the nasal passages.² Second, toxic chemicals in the smoke can cause damage, both directly and indirectly, through inflammatory mediators. Carbon monoxide intoxication is commonly associ­ated with human injuries from smoke and is a product of incomplete combustion.³ Finally, with combustion there is consumption of oxygen, and the resulting low PaO₂ can lead to pulmonary vasoconstriction as well as generalized hypoxia.

Three phases of pulmonary dysfunction have been described in the horse.^4,5 The first stage is acute pulmonary insufficiency caused by several mechanisms. Carbon monoxide may be present in sufficiently high concentration to cause toxicity within a short time after exposure. Carbon monoxide combines with hemoglobin to form carboxyhemoglobin. Hemoglobin has a 200 to 250 times greater affinity for carbon monoxide as compared with its affinity for oxygen.³ High levels of circulating carboxyhemoglobin result in a shift of the oxyhemoglobin dissociation curve to the left, thereby decreasing oxygen release at the tissue level and leading to tissue hypoxia.

These insults produce the second stage: formation of pulmonary edema, lower airway obstruction, and pulmonary parenchymal lesions. Within 48 to 72 hours after exposure, driven by pulmonary macrophages, neutrophils are called into the area of insult. They release cytokines, proteolytic enzymes, and oxygen-derived free radicals. Expression of the inflammatory cascade in excess of balance causes microvascular damage, leading to increased extravascular lung water. Local insult also results in the release of tissue factor, initiating the coagulation cascade to produce fibrin. Debris from the inflammatory cascade, along with fibrin and material directly deposited from smoke inhalation, create pseudomembranous casts, which may obstruct the small airways. Widespread plugging of the airways may significantly increase airway pressure, causing barotrauma and alveolar damage.⁶ The last stage, bronchopneumonia, occurs as a result of local and systemic impairment of the host immune system. This phase may take up to 1 to 2 weeks after the initial injury to develop.

It is common for horses with smoke inhalation injury to also have burn injury. A complete description of this process in horses has been reported. Thermal injury causes a local response that includes microvascular insult and direct tissue coagulation leading to inflammation, local edema, and finally necrosis. Extensive local injuries will drive a systemic response. An initial decrease in systemic organ blood flow is followed by the formation of generalized edema. The pathophysiology of edema formation is complex and involves protein shifts, endothelial damage, and alteration of the interstitial architecture, all leading to a net accumulation of fluid in the interstitial spaces.

Horses exposed to smoke will have a variety of clinical signs depending on the duration and type of exposure and the length of time from the insult. Acutely in more severe cases, within the first 6 hours, signs of carbon monoxide toxicity and shock may occur. The patient shows signs of severe hypoxemia and may be depressed, disoriented, irritable, ataxic, or even mori­bund. As edema and necrosis progress in the upper respiratory tract, dyspnea and stridor may develop. Auscultation of the thorax may reveal decreased air movement, crackles, or wheezes, but these may not become apparent for 12 to 24 hours. If edema of the airways is sufficiently severe, airflow may be severely restricted. Edema fluid may be visible at the nostrils and later may be replaced by inflammatory exudate. Concurrent edema formation may be occurring systemically. Hypoxia and generalized edema may lead to dysfunction of distant organs such as the kidneys and muscles. Signs of infection may be difficult to ascertain from other signs. All that may be noticed is a fever and a worsening of respiratory signs after initial improvement. As noted in a recent report, clinical signs may become more complex when there is thermal injury. Exposure to smoke without thermal injury can also occur, as with unhealthy air created by large wildfire smoke and associated debris. Clinical signs of an irritated respiratory tract are seen, similar to horses with chronic inflammatory airway disease.

Diagnosis is typically based on history and physical examina­tion. A normal initial examination does not rule out exposure because the onset of clinical signs may be delayed several days.

Treatment depends on the stage and extent of injury. Initially, oxygen support is of benefit. It is a treatment for carbon monoxide toxicity and helps reduce hypoxemia. Humidified oxygen can be supplied by nasal insufflation or via transtracheal catheter. Upper respiratory tract obstruction may necessitate a tracheostomy. Attention should be paid to keeping the airways clear, and nebulization may help. Bronchodilators may be useful in counteracting reflex bronchoconstriction. Decreasing inflammation and pulmonary edema may necessitate the use of diuretics and NSAIDs. Use of corticosteroids is controversial because of the potential for immunosuppression and laminitis.⁴ Where available, hyperbaric oxygen therapy has been used.

Secondary complications require attention. Cardiovascular compromise and burn-induced edema may necessitate the judicious use of intravenous fluids. To prevent infection, strict hygiene, meticulous nursing care, and optimal nutritional support should be provided.^1,7 Prophylactic antimicrobial use is not recommended in human patients. Documented infection should be treated with appropriate antimicrobial agents based on results of culture and sensitivity patterns.^1,7 Proper manage­ment of chronic smoke exposure includes limiting exercise and a full veterinary examination of the respiratory tract before return to work.¹³

Recurrent Airway Obstruction either by natural challenge (hay feeding) or by nebulization of a hay dust solution or a solution containing fungal spores, endotoxin, and silica microspheres.^13-17

Development of Pulmonary Inflammation

Within hours of hay dust exposure, airway neutrophilia and excessive mucus production are evident endoscopically and cytologically.^18,19 Alveolar macrophages, isolated from RAO- susceptible horses just 6 hours after nebulization of a hay dust solution,²⁰ have upregulated gene expression of TNF-α, IL-1β, and IL-8.

MACROPHAGES. The initial upregulation of macrophage- derived proinflammatory gene transcription is to be transient, with mRNA levels returning to preexposure levels 24 hours after hay dust challenge.^20,23,24 With continuous, chronic hay dust exposure, protein and the gene expression levels of IL-8, IL-1β, and TNF-α in the BAL cells remain upregulated, which probably reflects contributions from recruited neutrophils and other cells.^25-28 The net effect of these proinflammatory cytokines (and others such as IL-17, see below) is to enhance the chemokinetic gradient existing within the airways and to promote neutrophilic extravasation.^29,30 The neutrophil intracel­lular signaling pathways required for extravasation into the airways involve mitogen-activated protein kinases (MAPKs) and/or phosphoinositide 3-kinase (PI3K) as inhibition of these pathways attenuates neutrophil chemokinesis.³⁰ In RAO-affected horses, at least one mechanism that tempers neutrophil che­motaxis, secretoglobin production, is attenuated. These are mucosal epithelial-derived proteins that reduce IL-8-stimulated chemotaxis of blood-derived neutrophils. They also sequester calcium and phosphatidylcholine, a cofactor and substrate, respectively, both of which are required for phospholipase A2 activity.³¹ RAO-affected horses have reduced secretoglobin concentrations in their BAL, suggesting an absence of an inflammatory braking mechanism.^31-33

LYMPHOCYTES. The role of lymphocytes, especially T-helper cells, in the development of the pulmonary inflammatory response remains uncertain.

Other data suggest that RAO does not reflect a polarized Th2 immune response. There are no significant differences in IgE-protein-positive cells in the lung tissue samples or in the BAL cells of RAO-affected horses compared with control horses.^42,43 Second, in those two well-studied Swiss Warmblood families classified as healthy or RAO-affected (based on a horse owner assessed respiratory signs index [HOARSI]), no significant differences in serum IgE concentrations against recombinant A. fumigatus antigen were found.⁴⁴ Third, results of IgE serologic and intradermal testing failed to detect a difference between RAO-affected and control horses.⁴⁵ Fourth, cytokine profiles of BAL cells^27,28 or of pulmonary CD4+ and CD8+ cells⁴⁶ isolated from RAO-affected horses during the first 24 or 48 hours after hay dust exposure fail to demonstrate an upregulation of IL-4, IL-5, or IL-13. Furthermore, chronic (3-week) hay dust exposure is associated with an increase in the gene expression of IFN-γ in the BAL fluid cells,^26,27 although in one trial, but not in the second trial of the same chronically affected horses, the gene expression of IL-4 was upregulated.²⁶ Fifth, GATA3, the transcription factor required for IL-4 gene expression, is not upregulated in the bronchial cells of RAO-affected horses.⁴⁷ Sixth, IL-4 gene expression in the pulmonary lymph nodes isolated from chronically affected RAO horses is significantly downregulated compared to controls.⁴⁸ Importantly, these data do not provide evidence that RAO is a Th1 immune response because (1) concomitant increases in the gene expression of T-bet, the transcription factor for IFN-γ expression in CD4+ cells, does not occur.²⁷ Interestingly, in other species TLR stimulation leads to the production of proinflammatory factors, including IL-12, and subsequent IFN-γ synthesis.^49,50

In RAO-affected horses there is evidence that, as occurs in human asthmatics, Th17 cells (a subset of T-helper cells) contribute to the immunopathogenesis of RAO.^28,29,48 IL-17, secreted by Th17 cells and some granulocytes, promotes maturation, chemotaxis, and activation of neutrophils. IL-17 also upregulates the gene expression of IL-8 in the airway epithelium and promotes airway neutrophilia. Following chronic (≥2 weeks) hay dust exposure, IL-17 gene expression in BAL cells is upregulated in RAO-affected horses.^28,29 Increased IL-17 protein expression, detected immunohistochemically, is also found in the pulmonary lymph nodes of RAO-affected horses compared to controls.⁴⁸

Currently little is known about the role of regulatory T cells (Tregs) in the immunopathogenesis of RAO. In mice and humans, this subset of CD4+ cells helps control immune responses to viral, bacterial, fungal, and parasitic infections.⁵¹ In a study of RAO-affected horses during remission and exacerbation (4 days of hay challenge), the percentage of Tregs (CD4+, CD25+, Foxp3+) in the BAL increased from a baseline value (remission) of 1.5% to 4% (exacerbation).⁵² Investigators hypothesized that the increase in Tregs may help resolve immune-mediated bronchial inflammation.

EPITHELIAL CELLS. The airway epithelium not only serves as a physical barrier but also plays a key role in the secretion of cytokines needed to recruit inflammatory cells, to direct tissue-resident or circulating antigen-presenting cells, and/or to promote lung remodeling. It is likely that the inhaled microbial components bind to the airway epithelium via pattern recognition receptors and initiate an innate immune response with synthesis of inflammatory cytokines and chemokines.¹⁷ In RAO-affected horses, bronchial epithelial IL-8 gene and protein expression is upregulated at least 24 hours after hay challenge,^28,53 potentially contributing to airway neutrophilia. Bronchial gene expression of other chemokines (CXC che- mokine ligand 1 [CXCL1], granulocyte-macrophage colony­stimulating factor [GM-CSF], granulocyte colony-stimulating factor [G-CSF]) or cytokines (IL-5, IL-6, IL-10, IL-17, and transforming growth factor [TGF]-β1) is not different between 2853

control and diseased horses.²⁸, In a recent transcriptome analysis of endobronchial biopsies from affected horses, investigators reported an upregulation of genes involved in neutrophil chemotaxis (IL-8, CXC chemokine receptor type 2 [CXCR2]) and immune and inflammatory responses (matrix metalloproteinase [MMP]-1, MMP-8), suggesting that bronchial epithelium is an important contributor to the inflammatory process of RAO.⁵⁴

OTHER INFLAMMATORY MEDIATORS. As airway neutrophils and other cells accumulate, other inflammatory mediators are released and contribute to the inflammation. Histamine, released by degranulated mast cells, increases vascular permeability and produces bronchoconstriction. Although mast cell degranulation is typically attributed to IgE-mediated crosslinking of high- affinity receptors (FceRI) on mast cells, leukotrienes, prosta­glandin E, acetylcholine, substance P, endothelin, and peptidoglycan induce mast cell degranulation.^55,56 Despite increased histamine concentrations in the BAL of RAO-affected horses, histamine-1 receptor antagonists provide little thera­peutic relief for affected horses, suggesting that it is not a predominant inflammatory mediator in RAO.⁵⁷ The potential role of prostaglandins, with their bronchoconstrictive (PGF_2α, PGD₂, thromboxane A2 [TXA2]) and/or bronchodilatory capabilities (PGE2, PGI₂), has also been examined. In a dif­ferential gene expression study of lung tissue from RAO-affected horses, PGF synthase, an enzyme that reduces PGD₂ and PGH₂ to PGF_2α, and PTGDR, a PGD₂ receptor, were upregulated.⁵⁸ In diseased horses, increases in BAL PGE2 and PGF_2α^59,60 and in serum TXB2 (a metabolite of TXA2) occur.⁶¹ Although administration of a cyclooxygenase inhibitor prevents the increase in the serum TXB2 in RAO-affected horses, it does not inhibit the development of airway hyperresponsiveness or bronchospasm in susceptible horses. This suggests that eico­sanoids are not the major contributor to clinical signs. Lipoxygenase metabolites have also been investigated in RAO-affected horses. In humans, these compounds are bronchoconstrictors (leukotriene [LT] B4, LTC4, and LTD4), mucus secretagogues (LTC4 and LTD4), and chemokines (LTB4).⁶² Inhalation of LTB4 or LTD4 induces airway neu­trophilia and bronchoconstriction, respectively, in normal and RAO-affected horses.⁶³ BAL cells isolated from RAO-affected horses during the first 48 hours of hay dust exposure and stimulated ex vivo produce increased concentrations of LTB4 and LTC4.⁶² However, actual BAL concentrations of LTB4 and LTC4 are not elevated in chronically affected horses⁶⁰ despite the finding of the enhanced gene expression of LTA4 hydrolase (which metabolizes LTA4 to LTB4) in RAO-affected horses.⁵⁸ Pretreatment of RAO-susceptible horses with either a 5-lipoxygenase inhibitor or a LTD4 receptor antagonist before hay dust exposure does not prevent the development of airway neutrophilia or alterations in lung mechanics.^64,65 In addition, treatment of clinically affected horses with the leukotriene receptor antagonist montelukast (LTC4, LTD4, LTE4 antago­nist) also fails to improve pulmonary function test results, clinical scores, or arterial blood gas tensions.⁶⁶

Reactive oxygen species derived from pulmonary macro­phages and granulocytes contribute to the inflammatory process of RAO.⁶⁷ Oxidative stress occurs when enzymatic (superoxide dismutase, glutathione peroxidase, glutathione S-transferase) and nonenzymatic antioxidants (glutathione [GSH], ascorbate, α-tocopherol, bilirubin, albumin, and transferrin) are depleted.⁶⁸ Evidence of an oxidant stress within the airways of RAO-affected horses is based on the findings of (1) increased levels of oxidized glutathione (glutathione disulfide [GSSG]), (2) increased glutathione redox ratios (GSSG/GSSG + GSH), and (3) decreased ascorbic acid concentrations in the pulmonary epithelial lining fluid of RAO-affected horses.^67-69 Tan and colleagues also reported that both red blood cell and white blood cell glutathione peroxidase activities were significantly increased in RAO-affected horses during exacerbation compared with controls, suggesting a systemic response to the pulmonary inflammation.⁶⁸ (Systemic responses are discussed below.) Changes in these indices of oxidant stress correlate with lung dysfunction parameters (pulmonary [R_L], C_dyn, and PaO₂) in RAO-susceptible and RAO-affected horses.⁷⁰ Although reactive oxygen species (ROS) are well recognized for their microbicidal activities, they also upregulate proinflammatory gene expression by activating NF-κB.⁷¹

Proteases derived from airway phagocytes, lymphocytes, and epithelial cells may propagate the inflammatory reaction of RAO. MMPs with elastolytic and collagenolytic activities have been identified in the respiratory tract secretions of RAO-affected horses and include MMP-1, MMP-8, MMP-9, and MMP-13.^72-77 The degradative effect of MMP-9 on the basement membrane and on the extracellular matrix components facilitates the extravascular movement of neutrophils into the airways. The collagenase activity of MMP-8 and MMP-13 contributes to airway remodeling during disease exacerbation.⁷⁶ In vitro, tetracycline derivatives effectively inhibit MMP activity detected in tracheal epithelial lining fluid,⁷⁷ but their efficacy in ameliorating or preventing the inflammatory response of RAO-susceptible horses exposed to hay dust remains to be determined.

■ Epidemiology and Genetics RAO occurs in both ponies and horses, but no gender or breed predisposition has been identified.^78,79 The prevalence of RAO is dependent on the geographic region examined. Bracher reported that 60% to 80% of Swiss horses older than 8 years of age suffer from some degree of this disease. In North America and in Europe the prevalence of RAO is estimated to range from 12% to 50%.⁷⁸ Although exposure to hay particulates is integral to the development of the disease, the role of a prior viral (or bacterial) infection in inducing susceptibility to hay dust hypersensitivity remains to be established. Horses with IAD (mild to moderate equine asthma) share some of the clinical features of RAO,⁸⁰ yet the relationship of IAD to the development of RAO is unknown. Currently there is no evidence that IAD progresses to RAO.^1,3

A heritable component of RAO disease susceptibility exists. Marti and colleagues examined families of RAO-affected warmblood and Lipizanner horses and found that the risk of developing RAO was 3.2 times higher when one parent (dam or sire) had RAO and 4.6 times higher when both parents did.⁸¹ In segregation and genomic analyses of two families of Swiss Warmblood horses, the mode of inheritance of RAO is characterized by major gene effects but the potential candidate genes differ between the two families.^39,82 In one family of Swiss Warmblood horses, RAO was transmitted in an autosomal recessive mode with the major association of microsatellites or single nucleotide polymorphisms and affected cases being found on equine chromosome 13 (ECA13).⁸³ In the other family of Swiss Warmblood horses, an autosomal dominant mode of inheritance was found, with the major association found on ECA15.⁸² Investigators suggested that the observed genetic heterogeneity between the two families of Swiss Warmblood horses may reflect etiologically distinct subgroups of RAO that are not distinguished phenotypically. Proposed potential candidate genes on ECA13 were those encoding receptors for IL-4 (IL4R) and IL-21³⁹; on ECA15, a potential gene of interest was SOCS5 (suppressor of cytokine signaling 5). Using a high-density SNP chip to re-genotype a subset of that same Swiss Warmblood family, Schnider identified an association (single SNP) located in an intron of TXNDC11; this locus was separate from the other previously identified regions.⁸⁴ TXNDC11 encodes a thioreductase involved in the degradation of glycoproteins and, in cell cultures, mucin expression by MUC5AC gene. However, this SNP association failed to reach Bonferroni corrected genome-wide significance threshold. Although the results are intriguing, it remains to be determined if RAO genetic susceptibility maps to loci on ECA13 and ECA15 in other horse breeds. Lastly, as copy number variations may help identify candidate genes involved in disease processes, Ghosh and colleagues investigated this in a genomic analysis of RAO-affected and control horses (multiple breed types). Although 20 RAO-specific copy number variant regions were identified in 9 of 16 horses, they were distributed on 15 autosomes, and none was on ECA13.⁸⁵

■ Clinical Signs and Differential Diagnosis The severity of clinical signs in RAO-affected horses is variable and is exacerbated by stabling, exposure to dusts, hay, and ammonia fumes. Winter is considered a risk factor for disease exacerbation. However, RAO-susceptible horses kept outdoors during hot, humid summer weather—which may reflect a period of high concentrations of airborne pollen and spores—also exhibit a worsening of clinical signs.⁸⁶ In a study of 148 cases of RAO presented for evaluation to a veterinary referral hospital in the United Kingdom, the chief complaints (median duration of which was 7 months) were coughing (84% of cases), bilateral nasal discharge (54%), exercise intolerance (51%), and postex­ercise breathing difficulty (23%).⁸⁷ In a review of 16 cases presented to a veterinary teaching hospital in western Canada, the most common historical complaint (88% of cases) was coughing.⁸⁸ In a review of 65 cases presented to a veterinary teaching hospital in the northeastern United States, the primary problems were abnormal breathing effort (68% of cases), nasal discharge (50%), and a spontaneous cough (46%).⁸⁹

On physical examination, mildly affected horses are typically afebrile and exhibit a bright demeanor and a normal or slightly accentuated (abdominal) breathing effort, respiratory rate, and heart rate. Nasal discharge may not be evident in all cases, perhaps because horses swallow excessively the expectorated exudate. When the horse is exercised, a thick white nasal discharge and a cough become apparent. The frequency of coughing increases when horses are stabled, often exceeding 10 to 15 coughs per hour.⁹⁰ More severely affected horses exhibit tachypnea (respiratory rate >40 breaths/min), nostril flaring, and a markedly accentuated expiratory effort (heave). The thoracic expiratory component may precede that of the abdominal compartment, producing a “double effort.”⁹¹ Abdominal pressure swings may be large enough to cause the anus to protrude with expiration. Paroxysmal coughing, bilateral nasal discharge, anxious facial expression, flatulence during coughing, reluctance to move, inappetence, and weight loss occur in the most severely affected cases.

Although RAO is a diffuse lung disease, auscultation of mild cases may not reveal abnormal lung sounds unless a rebreathing bag is used to induce a hyperpnea. In one survey, abnormal thoracic (47% of cases) and tracheal (63% of cases) auscultation occurred.⁸⁷ In the latter, abnormal lung sounds and “mucus clicks” (tracheal exudate) were detected by tracheal auscultation. In mildly affected horses, only regional areas of adventitious lung sounds (crackles or high-pitched expiratory wheezes) may be detected. In contrast, in severely affected horses, wheezes and mucus clicks may be heard at the horse's nostrils without the use of a stethoscope. Depending on the severity of the disease, thoracic percussion may be normal or may reveal expansion of the caudal lung fields beyond the normal line of pleural reflection.

Researchers have developed clinical scoring systems to categorize the severity of RAO. In moderately to severely affected horses, there is a very good correlation between the clinical signs and alterations in RL or Cdyn,⁹² but in mild cases, clinical scores underestimate the severity of changes in the pulmonary function tests. In another study of 30 RAO-affected horses and 10 controls, investigators found that the severity of some clinical signs (e.g., cough, nostril flare, abdominal lift but not breathing frequency, thoracic auscultation, exercise intolerance) correlated with tracheal mucus scores, BAL neutrophil percentages, and thoracic radiographic evidence of bronchial and tracheal thickening.⁹³

In rare cases, horses with long-standing RAO develop cor pulmonale, evidenced by jugular distention, pulsation, ventral edema, and tachycardia.^94,95 Although RAO-susceptible horses exhibit echocardiographic alterations during acute episodes of RAO (characterized by an increase in pulmonary artery diameter, a decrease in left ventricular diameter, and an abnormal septal 96 motion), serum troponin measurements remain normal.⁹⁶ Because the echocardiographic parameters (reductions in the circumferential ratio of the aorta to the pulmonary artery and in the mass ratio of the left ventricle and septum to the right ventricle) return to normal when horses are in remission, it is likely that the reversibility of the airway disease reduces the risk of permanent myocardial disease.^95,9

■ Laboratory Aids and Diagnostic Tests The diagnosis of RAO is based on the history, the clinical signs, demonstration of the reversibility of airway obstruction following broncho­dilator administration, and the response to therapy. A definitive diagnosis may be challenging in asymptomatic or mildly affected cases necessitating a “hay challenge” to precipitate clinical signs. Endoscopic examination of the respiratory tract of RAO-affected horses demonstrates excessive amounts of tra­cheobronchial secretions (mucus) that originate from most of the bronchial segments. Distal airways are edematous and inflamed, and airways easily collapse during expiration. Expec­torated secretions from the lower respiratory tract coat the pharynx, larynx, and other structures of the upper respiratory tract. Cytologic analysis of lower respiratory tract secretions (BAL) is helpful in symptomatic cases. During periods of clinical remission, BAL cell populations obtained from RAO-susceptible horses consist predominantly of mononuclear cells (90% macrophages and lymphocytes) and do not differ from those found in healthy horses.⁹⁸ BAL from symptomatic horses contains more than 25% nondegenerative neutrophils and often exceeds 90% of the total nucleated cell count. Yet the extent of pulmonary neutrophilia alone is not correlated with severity of clinical signs.⁹⁹ In most cases the percentage of eosinophils or mast cells is not increased in affected horses. Thus cytologic analysis of BAL in RAO-affected horses reflects a suppurative nonseptic inflammation with increased mucus concentrations, distinguishing this disorder from lungworm infections (pul­monary eosinophilia), bacterial pneumonia (degenerative neutrophils, intracellular bacteria), or fungal pneumonia (degenerative neutrophils, hyphae, erythrocytes). In rare cases, infectious bronchitis may complicate RAO.

In general, the complete blood cell count (CBC) and serum biochemistry profiles are usually normal in RAO-affected horses.⁸⁹ In severely affected horses, a mature neutrophilia and lymphopenia may be evident on the leukogram. The “stress leukogram” reflects the effects of increased plasma cortisol concentrations that occur at least 6 hours after RAO-susceptible horses are exposed to a high-dust environment.¹⁰⁰ Significant increases in serum haptoglobin, amyloid A, and functional fibrinogen concentrations in RAO-affected horses (maintained in dusty environments for >30 days) occur, challenging the conventional idea that RAO is a compartmentalized inflam­matory response.^101,102

Arterial blood gas analyses are normal in approximately 25% of the cases¹⁰³; other cases exhibit hypoxemia (PaO₂ although not typically performed to diagnose RAO, demonstrate bronchiolar goblet cell metaplasia, bron- chiolar luminal exudate, peribronchiolar lymphoplasmacytic cell infiltration, and accumulations of neutrophils within the airways.⁸⁸ The potential disadvantages of performing a lung biopsy include the risk of inducing hemorrhage (hemothorax, epistaxis) and the chance of obtaining a nonrepresentative tissue sample.

Intradermal skin testing of RAO-affected horses has been investigated extensively with the aim of diagnosing RAO and identifying the offending allergens. In separate studies, Evans and Jose-Cunilleras found that the total number of positive intradermal skin tests evaluated at 0.5, 4 to 6, and 24 hours after allergen injection was greater in RAO-affected horses compared to controls but that this difference in skin reactivity did not permit discrimination between a diseased horse and a healthy horse.^109,110 In contrast, others have failed to find a significant difference between RAO-affected and control horses in the number of positive skin responses to a variety of molds at either the early (0.5, 4 hours) or the late (24 hours) phases.^111-113 Wong and colleagues, evaluating the skin responses to intradermal histamine and Aspergillus antigens, found that compared with the controls, the RAO-affected horses had a heightened response to histamine (evaluated at 0.5 hours post injection) and a delayed resolution of the skin wheal (after 24 hours) to the Aspergillus antigens.¹¹⁴ In general, positive skin tests and serum precipitins to fungal and thermophilic acti- nomycete antigens are found in many normal as well as RAO-affected horses, reflecting a level of exposure of the horse rather than a susceptibility to disease. Skin testing does not allow identification of RAO-susceptible or RAO-affected horses or of the specific inhaled particulate that is inciting the pul­monary inflammation.¹¹⁵

Serum allergy tests such as the radioallergosorbent test (RAST) or IgE-based ELISA, which aim to identify the allergen that is inciting the hypersensitivity, are of limited value in the diagnosis of RAO. Lorch and colleagues compared the results of three serum IgE quantitation assays—the spot test (a RAST), an ELISA that uses a recombinant human FceRIα chain as the IgE readout reagent (Allercept [Heska, Loveland, Colo.]), and an equine ELISA that uses a polyclonal antiserum to detect equine IgE—with intradermal skin test responses in RAO- affected horses. They concluded that based on the poor performance of all three tests, none should be used as a screening test for allergen hypersensitivity.¹¹⁶ Similar conclusions were reached by Tahon and colleagues, who compared the Allercept test and the CAST test (Alpco [Windham, N.H.]), which measures IgE-mediated release of sulfidoleukotrienes from peripheral blood basophils, with intradermal testing in RAO- affected horses and controls. None of the tests showed a significant difference in responses between diseased and control horses, and thus these tests did not support an IgE-mediated basis for RAO.⁴⁵

Conventional pulmonary function tests (PFTs), used primar­ily in research investigations to confirm the diagnosis of RAO, require measurements of transpleural pressure changes (obtained via an intrathoracic esophageal balloon catheter system) either alone or in combination with airflow measurements (detected by a flow meter attached to a sealed face mask worn by the horse). Horses with RAO exhibit a decrease in dynamic lung compliance or lung distensibility (C_dyn), an increase in pulmonary resistance (R_L), an increase in maximum pleural pressure changes (ΔP_pι_max), and an increase in peak inspiratory and expiratory flow rates.¹¹⁷ In general, C_dyn reflects changes in peripheral airway function, whereas R_l reflects central, larger airway function. RAO-affected horses also exhibit bronchial hyper­reactivity: The dose of aerosolized histamine or methacholine required to reduce C_dyn to 65% of its baseline value is much less in diseased horses than in controls.¹¹⁸ During periods of clinical remission, PFTs are not different from those of healthy age-matched horses. Yet when using more sensitive tests of airway disease, horses in remission still maintain some degree of bronchial hyperreactivity and airflow obstruction,^119,120 demonstrating that conventional PFTs lack sensitivity.^121,122

The changes in respiratory mechanics and the development of clinical signs are consequences of the influx of inflammatory cells, the contraction of smooth muscle in the airways, the production and accumulation of mucus, and possibly remodeling of the lung parenchyma (see below). Bronchoconstriction probably results from the cumulative effects of (1) released inflammatory mediators (histamine, serotonin, eicosanoids, leukotrienes, endothelin) and (2) activation of muscarinic receptors on the smooth muscle by released acetylcholine.¹²³ In addition to their direct effects on smooth muscle tension, histamine and serotonin also augment the release of acetyl­choline from parasympathetic nerves.¹²⁴ Based on in vitro muscle bath studies, there is evidence that RAO-affected horses lack inhibitory nonadrenergic-noncholinergic innervation of the peripheral (but not central) airways, further contributing to the development of bronchospasm.¹²⁵ However, the expression or distribution of airway muscarinic receptor subtypes (M₁, M2, M3) is unaltered in RAO-affected horses.¹²⁶

As noted earlier, excessive mucus is evident endoscopically and cytologically in RAO-affected horses. This may be the result of increased mucus viscoelasticity^127,128 and an impairment of the mucociliary apparatus damaged from epithelial cell destruction.^129,130 An increase in the number of mucus cells and their delayed apoptosis may also be contributory.¹²² An upregulation of two mucin-producing genes, MUC5AC and MUC2, in airway tissues of RAO-affected horses has not been detected.^131,132

Airways that are narrowed, obstructed, or both (as a result of inflammatory cells, mucus, and alterations in pulmonary surfactant)^133,134 increase R_l and fail to participate in efficient gas exchange. Low ventilation-perfusion lung regions contribute to a widening of the alveolar-arterial oxygen-difference (AaDO₂) gradient and the development of arterial hypoxemia.^135,136 High ventilation-perfusion regions (dead space) also increase in affected horses, but normal arterial PaCO₂ levels can be maintained in some horses through increases in the total minute ventilation.¹³⁷ Changes in minute ventilation are usually due to an increase in breathing frequency, as tidal volume changes little with the disease. In general, there are no significant changes in cardiac output, heart rate, and mean systemic arterial pressure, but pulmonary artery systolic and mean pressures and pulmonary vascular resistances are significantly increased in diseased horses.¹³⁷ The combination of airway inflammation and hypoxemia, if severe enough, stimulates the respiratory controller in the ventral medulla to increase the respiratory drive and to activate inspiratory and expiratory muscles (abdominal heave). Recruitment of respiratory muscles increases the maximum pleural pressure excursion (ΔP_pι_max), the work of breathing, and the total body oxygen consumption.¹³⁸ However, much of the increased expiratory effort (and increased intrathoracic pressures) collapses the noncartilaginous airways, further compounding gas trapping and impairing gas exchange.

■ Pathology The lungs obtained from RAO horses in remission at the time of death appear normal.¹³⁵ In symptomatic horses, the lungs are pink, soft, hyperinflated, and imprinted with rib impressions. Lungs fail to collapse when the chest is opened because of gas trapping distal to occluded airways.¹¹¹ Emphysematous bullae are usually lacking, but accumulations of mucus and cellular debris occlude the small airways, forming spiral-shaped casts known as Curschmann' spirals. Light and electron microscopic examinations of lung segments from diseased horses demonstrate that the peripheral airways are the most severely affected but central airways are not spared.^129,130 Histologic changes include goblet cell hyperplasia, epithelial cell damage, bronchial and bronchiolar epithelial cell hyper­plasia, smooth muscle hypertrophy and hyperplasia, increased collagen and elastin deposition in the lamina propria, and overinflation of the alveoli.^139-141 Pathologic changes occur in response to pulmonary inflammatory mediators (release of neutrophil-derived exosomes, ROS, MMPs)¹⁴² and mechanical loads on the airways. Peribronchiolar infiltrates consist of lymphocytes and plasma cells, along with collagen deposition¹⁴³; eosinophils are rarely observed either within the airways or in the lung parenchyma. Airway remodeling (changes in airway smooth muscle [ASM] mass and collagen) as a function of antigen exposure, avoidance (pasture), and/or antiinflammatory and bronchodilator therapy have been extensively studied. Compared to age-matched controls, the ASM mass in the peripheral airways of RAO horses in remission is increased two- to threefold.¹⁴⁴ When RAO-affected horses are treated with either pasture turnout (antigen avoidance) or daily inhaled fluticasone for 12 months, a 30% reduction in peripheral ASM mass is observed.¹⁴⁵ A similar reduction in ASM mass occurs when RAO-affected horses are maintained in an antigen avoidance environment (pasture) while receiving daily inhaled fluticasone/salmeterol for 3 months,¹⁴⁶ an effect attributed to the steroid and not the bronchodilator. In the central airways, fluticasone/salmeterol treatment reversed extracellular matrix remodeling (of which collagen is a component) both within the lamina propria (decreased thickness) and within the smooth muscle layer.¹⁴⁶

■ Treatment and Prognosis

ENVIRONMENTAL MANAGEMENT. Environmental man­agement remains the single most important therapeutic and prophylactic intervention. RAO-prone and RAO-affected horses should be kept outside at all times, blanketed during inclement weather as needed, and provided access to a well-ventilated, two- or three-sided shelter free of manure and urine (ammonia) fumes. When pasture is limited, the diet is supplemented with complete pelleted feed or hay cubes. Some horses will tolerate hay that has been soaked in water, but this practice is often forsaken when the weather becomes cold. Steaming hay can prevent exacerbations of RAO in susceptible horses,¹⁴⁷ as this reduces bacterial and mold counts of the hay.¹⁴⁸ In affected horses that have been turned out to pasture, clinical remission and normalization of pulmonary function test results usually take 4 to 8 weeks depending on weather conditions.¹⁴⁹ Reductions in peripheral airway smooth muscle mass take 12 months.¹⁴⁵ Acute exacerbations of RAO occur in pastured, susceptible horses during periods of hot, humid weather.^86,150 Owners should provide an alternative clean, cool environment during these periods to reduce the risk of pulmonary crises in susceptible horses.

In addition to alterations in forage, horses that are stabled should be bedded on a material that has a low respirable dust particle count. Cardboard bedding has a significantly lower particulate concentration (dust, A. fumigatus, F. rectivirgula, and T. vulgaris) than do wood shavings, wheat, or flax straw, although cardboard may be difficult to obtain and store.^149,151 Auger and colleagues found optimal reductions in airborne respirable dust in barns when horses were bedded on shavings and fed steamed hay.¹⁵² In affected horses that are bedded on shavings and fed pelleted rations, significant improvements in the PFT results occur within 3 days.¹⁵³ Complete resolution of clinical signs may take 30 days, although not all horses respond to only allergen reduction measures while stabled indoors.¹⁵⁴ When implementing changes for RAO-susceptible horses, attention should be given to making similar alterations to surrounding stalls within the common airspace.^122,155 At-risk horses should be removed from the barn when stalls or other areas are being cleaned. When environmental management changes are implemented, horses in remission can remain clinically normal for months and exhibit normalization of BAL fluid neutrophil percentages and other indices of airway inflam­mation such as exhaled ethane. Nevertheless, bronchial hyperreactivity and forced expiratory airflow in RAO-prone horses still remain abnormal.^119,156-158 Therefore these horses will always be more susceptible to disease development than healthy horses.

Environmental management considerations should also be implemented when horses are transported. Their heads should not be tied in close proximity to the hay net, as this practice places their breathing zone near a concentrated source of organic dusts and molds and restricts their head position so as to further decrease mucociliary clearance of respiratory secretions.¹⁵⁹

BRONCHODILATORS. During acute exacerbations of RAO, bronchodilators improve lung mechanics and are useful in establishing the reversibility (and thus diagnosis) of RAO. The three main classes of bronchodilators that have been used in RAO-affected horses are the anticholinergics (muscarinic antagonists), the β₂-agonists, and the phosphodiesterase (PDE) inhibitors.

Muscarinic antagonists such as atropine, ipratropium, and N-butyl-scopolammonium bromide block the smooth muscle­constricting effects of acetylcholine. These actions exert the greatest effects in the central as opposed to the peripheral airways. IV administration of 0.01 to 0.02 mg/kg of atropine sulfate improves clinical signs and pulmonary function param­eters (ΔP_plmax, respiratory rate, R_l, E_l) within 10 minutes of administration.^160-162 Atropine’s short duration of action (2 hours) and undesirable side effects (mydriasis, ileus, tachycardia, dry airway secretions) preclude its routine use in the treatment of RAO. In contrast, ipratropium bromide is poorly absorbed from either the respiratory or the GI tract, minimizing the development of systemic effects. The powder form is nebulized (2 to 3 μgZkg) and exhibits an onset of action within 30 minutes; effects last from 4 to 6 hours.¹⁶³ In resting horses it reduces R_l and ΔP_plmax and increases C_dyn but produces no further improvement if horses are exercised.¹⁶⁴ N-butyl-scopolammo- nium bromide, when administered at 0.3 mg/kg IV, improves clinical scores and pulmonary function tests of RAO-affected horses within 10 minutes of its administration; effects last for 30 to 60 minutes.^162,165 No clinically adverse effects were noted following its use in research trials.

The β₂-adrenergic agonists—clenbuterol, albuterol, leval- buterol, and salmeterol—produce relaxation of smooth muscle by increasing the intracellular levels of cyclic adenosine monophosphate (cAMP). The bioavailability of clenbuterol is greater than 90%, and when administered orally at a dose of 1.6 pg/kg of body weight twice per day (bid), a significant improvement in the clinical signs of affected horses can occur.¹⁶⁶ The effective dosage of clenbuterol varies among horses. If clinical improvement is not observed within 3 days of initiating treatment, the dose is increased stepwise (e.g., 2.4 pg/kg for 3 days, 3.2 pg/kg for 3 days). Adverse side effects (anxiety, shivering, sweating, and tachycardia) occur at higher dosages but may be minimized if clenbuterol is increased in a stepwise manner. Reports of fatal overdoses of clenbuterol in horses estimated to have received 100 pg/kg orally of a compounded product have occurred.¹⁶⁷ In addition to its bronchodilatory effect, clenbuterol also exerts an antiinflammatory effect: IV administration (0.75 pg/kg bid) to RAO-susceptible horses prior to hay dust exposure attenuates BAL fluid neutrophilia and TNF-α and IL-1 gene transcription.¹⁶⁸ Prolonged admin­istration of clenbuterol (0.8 pg/kg bid for 3 weeks) may result in tachyphylaxis¹⁶⁹ due to downregulation of β-adrenoceptors, but this phenomenon may be reversed with concurrent cor­ticosteroid administration.¹⁷⁰

Aerosolized β₂-agonists are used to treat RAO-affected horses and permit direct delivery of the drug to the respiratory tract, minimizing the total amount of drug needed and reducing the prevalence of unfavorable systemic effects. The metered dose inhaler containing the drug is attached to a delivery system such as the Equine Haler (Equine Health Care, Inc., Horsholm, Denmark) or Aerohippus (Trudell Medical, Inc., London, Canada). Both devices exhibit similar efficacy in drug delivery and clinical improvement.¹⁷¹ Albuterol sulfate and levalbuterol improve pulmonary function by 60% to 70% in RAO-affected horses for approximately 1 to 3 hours, necessitating frequent administration of the drugs.^172,173 They are poorly absorbed, so systemic effects are uncommon. These drugs are typically used in “rescue” protocols to treat horses in respiratory distress. The recommended dose of albuterol is 1 to 2 pg/kg q1-4h and of levalbuterol is 0.5 pg/kg q4h.¹⁷⁴ Salmeterol, a long-acting β₂-agonist, is administered at a dose of 0.5 pg/kg q6-8h. Clinical improvement develops within 2 hours of administration, and significant decreases in pulmonary resistance last for 6 hours.¹⁷⁵

Aminophylline and pentoxifylline are examples of the PDE inhibitors that produce smooth muscle relaxation by inhibiting the breakdown of intracellular cAMP. PDE inhibitors also exert antiinflammatory properties.¹⁷⁶ Administration of 12 mg/ kg IV of aminophylline (in 1 L of 5% dextrose) improved clinical signs and reduced ΔP_pι_mat in approximately 50% of RAO cases¹⁶¹ but caused hyperexcitability, hyperesthesia. and trembling. Because of the narrow range between therapeutic efficacy (serum concentration of 10 pg/mL) and toxicity (serum concentration of 15 pg/mL), aminophylline is not routinely used for the treatment of RAO. Pentoxifylline is well absorbed after oral administration (32 mg/kg q12h), improves some indices of pulmonary function, but does not improve BAL fluid neutrophilia.¹⁷⁷

ANTIINFLAMMATORY DRUGS. NSAIDs, in contrast to corticosteroids, are not beneficial in the treatment of RAO.⁶¹ Of the glucocorticoids evaluated, oral prednisone is not used because of its poor bioavailability and lack of conversion to the active metabolite prednisolone.^178,179 Liquid and tablet forms of prednisolone are rapidly absorbed from the equine GI tract (bioavailability is approximately 50%), achieve peak concentrations 45 minutes post administration, and exert pharmacologic effects (based on changes in serum cortisol levels) for 8 hours.¹⁷⁹ Oral prednisolone (2 mg/kg PO once per day [sid] for 7 days) improved R_l, E_l, and ΔP_plmax within 3

days of its administration to RAO-affected horses stabled and continuously exposed to dusty hay. It was less efficacious than dexamethasone administered by intravenous (1 mg/kg q24h)¹⁷⁸ or oral (0.05 mg/kg q24h)¹⁸⁰ routes.

In RAO-affected horses receiving dexamethasone, within 2 hours of IV administration (0.1 mg/kg) there is a 37% decrease in ΔP_plmax that is further reduced by 4 hours post treatment.¹⁸¹ Oral administration of the parenteral formulation of dexamethasone (0.164 mg/kg q24h to account for bioavail­ability) improves lung function parameters within 6 hours of treatment, with peak effects occurring 24 hours later. When horses are maintained in dusty environments, IV dexamethasone (0.1 mg/kg q24h) or dexamethasone isonicotinate (0.04 mg/ kg IM every 3 days for three treatments) improves pulmonary function within 3 days of initiating therapy^178,181 and reduces airway inflammation (BAL fluid neutrophilia) within 10 days of treatment.^178,182 As might be predicted, a single dose of dexamethasone isonicotinate (0.06 mg/kg IM) administered to RAO-affected horses maintained in a dusty environment fails to provide discernible improvements in the clinical score, the PFT results, the BAL cytologic findings, or the activities of the transcription factors NF-κB or activator protein 1 (AP-1) evaluated 10 days post treatment.¹⁸³ In general, the rapid onset of action of IV dexamethasone makes it an antiinflammatory of choice in horses exhibiting respiratory distress caused by RAO.¹⁸⁴

Long-acting corticosteroids have also been evaluated in RAO-affected horses maintained in a dusty environment. A single IM dose of triamcinolone acetonide (0.09 mg/kg) mark­edly improved lung mechanics for 4 weeks. Cytologic changes (evaluated 14 days post treatment) demonstrated a reduction in airway neutrophilia (from 62% to 32%), but this change was not statistically significant.¹⁸⁵ Adverse effects of triamcino­lone administration were not reported. In another study, RAO-affected horses received isoflupredone acetate for 14 days (0.03 mg/kg IM q24h) while being maintained in a dusty environment. Significant improvements in pulmonary function indices occurred within 3 days of initiation of the treatment,¹⁸⁶ and beneficial effects continued 7 days beyond the cessation of drug therapy. The only adverse effect noted was the develop­ment of hypokalemia.

Inhaled corticosteroids such as beclomethasone and fluti­casone have been used with success to treat RAO. In a 3-week study, fluticasone proprionate (2 mg q12h) was administered to RAO-affected horses while they were stabled and fed “good”- quality hay. By day 21, marked improvements in PFTs and normalization of BAL neutrophil counts occurred.²⁶ Robinson and colleagues compared the effects of inhaled fluticasone (6 mg bid) and IV dexamethasone (0.1 mg/kg q24h) on PFTs and BAL cytologies in RAO-affected horses.¹⁸⁴ They found that (1) no adverse effects, including laminitis, were associated with a 3-day course of fluticasone at 6 mg bid; (2) greater reductions in ΔP_plmax in RAO-affected horses were achieved with dexamethasone than with fluticasone treatment; and (3) prophylactic (3-day) and continual (4-day) administration of fluticasone (6 mg bid) to RAO-susceptible horses subjected to stabling and dusty hay exposure were more efficacious in preventing the disease than was dexamethasone. They recom­mended that dexamethasone should be used for “rescue” therapy and fluticasone used as a preventive agent. Fluticasone is approximately 18 times more potent than another inhaled corticosteroid, beclomethasone dipropionate. Nasal or oral administration of fluticasone has no effect on circulating glucocorticoid levels, suggesting that its bioavailability across nasal and intestinal epithelium is poor.¹⁸⁷ However, daily aerosolization of fluticasone (dose of 1.5, 3, or 6 mg bid) significantly reduces serum cortisol measured 72 hours after initiating therapy.^184,187 The effect of long-term fluticasone administration on immunologic parameters has also been investigated. Administration of fluticasone (2 mg bid for 11 months) to RAO-affected or susceptible horses had no significant effect on (1) peripheral blood leukocyte, neutrophil, or lym­phocyte numbers; (2) percentages of lymphocyte phenotypes; (3) blood neutrophil chemokine (IL-8, TNF-α) and gluco­corticoid receptor gene expression; (4) IgG vaccine titers induced by immunization against tetanus toxoid or an inactivated infectious bovine rhinotracheitis (IBR) vaccine; or (5) mitogen- induced lymphocyte proliferative responses.¹⁸⁸ These data suggest that fluticasone is a relatively safe drug for use in the treatment of RAO.

Another inhaled corticosteroid that has been widely inves­tigated in RAO-affected horses is beclomethasone dipropionate. When administered at a dose of 1320 to 1500 μg bid for 7 to 10 days to diseased horses continuously stabled and fed dusty hay, investigators noted that by day 7 of treatment there was a significant improvement but not a complete normalization of pulmonary function test results and BAL fluid neutrophil percentages.^182,189

Inhaled steroids are often used in conjunction with inhaled bronchodilators for maximal therapeutic effects. When separate formulations are used, it is recommended that the bronchodilator be administered first so as to improve subsequent glucocorticoid deposition in the peripheral airways.¹⁹⁰ As suggested earlier, when administering glucocorticoids, adrenal gland atrophy, immunosuppression, and precipitation of laminitis remain a concern. All of the previously discussed parenteral and inhaled glucocorticoids reduce endogenous serum cortisol levels, but despite prolonged treatment protocols with glucocorticoids, adrenal gland responsiveness to adrenocorticotropic hormone administration remains.^{185,186,189,191} Despite the improvement in clinical signs and in PFTs achieved with glucocorticoid therapy, BAL neutrophilia remains evident until environmental management changes are implemented.

OTHER THERAPIES. In horses that develop respiratory distress because of the bronchospasm and airway inflammation, additional therapies are required. Administering nasal oxygen at flow rates as low as 5 L/min improves arterial oxygen tension by as much as 30 mm Hg in some severely affected horses.¹⁹² Flow rates of 30 L/min (delivered by 2 nasal cannulae at 15 L/ min) are associated with coughing and gagging in horses. Nasal oxygen supplementation does not reduce breathing frequency, suggesting that stimulation of vagally mediated afferents (perhaps responding to inflammatory mediators) is responsible for the tachypnea.

Furosemide (1 mg/kg IV or nebulized) provides beneficial effects to RAO-affected horses within 20 minutes of its administration, decreasing R_L and increasing C_dyn without affecting PaO₂.¹⁹³ The beneficial effects appear to be mediated by PGE2, derived from either the renal or the airway epithelium, which promotes smooth muscle relaxation. Prior treatment with the cyclooxygenase inhibitor flunixin meglumine prevents the furosemide-induced bronchodilation.¹⁹⁴ Furosemide may have a local effect on vagal sensory afferent discharge inde­pendent of reflex anticholinergic bronchodilation.¹⁹⁵

Another therapeutic agent proposed for the treatment of RAO is tamoxifen. This is a selective estrogen receptor modula­tor that, when administered at 100 mg PO every other day for three treatments, reduced airway neutrophilia in RAO- affected horses.¹⁹⁶ Tamoxifen also enhances apoptosis of peripheral blood and BAL neutrophils. Long-term studies of its efficacy and safety remain to be conducted.

Finally, Barussi and colleagues reported that in a clinical trial with only four horses, intratracheal instillation of autolo­gous bone marrow-derived mononuclear cells improved clinical scores and reduced BAL neutrophilia.¹⁹⁷ Larger studies need to be completed before this approach can be recommended.

Additional therapeutic approaches that have not been found to be efficacious in the treatment of RAO-affected horses include a single acupuncture treatment,¹⁹⁸ the oral administration of an herbal preparation containing thyme and primula,¹⁹⁹ the rapid IV administration of 30 L of isotonic saline,²⁰⁰ treatment with a constant rate infusion of IV lidocaine,²⁰¹ and sound-wave physiotherapy.²⁰²

■ Prognosis RAO-prone horses are likely to develop acute exacerbations of the disease when husbandry practices lapse or when weather conditions become adverse. Horses do not “outgrow” the disorder, and in fact some clinicians believe that the inflammation becomes more difficult to manage as the horse ages. The basis for the lack of clinical response is unknown but may be the result of lung remodeling or potential downregulation of the glucocorticoid receptor.

If proper environmental management changes are imple­mented and if the appropriate therapy is initiated, clinical remission occurs in most cases. In a follow-up survey conducted by Naylor of RAO-affected horses that had been examined and treated at a referral center 2 to 4 years previously, it was found that 20% of the original cases (3 of 15) had been eutha­nized, 33% were still receiving bronchodilators on an as-needed basis, and athleticism had not decreased in 92% of the horses.⁸⁸ Similarly, in a follow-up survey conducted by Aviza and col­leagues,⁸⁹ 13% of the horses that had been diagnosed with RAO 3 to 4 years earlier had been euthanized. However, in contrast to Naylor's findings, Aviza noted that more than 50% of the respondents in that survey stated that the athletic performance of the horse had been compromised by the disease. Perhaps this difference between the two surveys simply reflected a failure of the owners to comply with environmental manage­ment recommendations. When queried about the husbandry practices, 77% of the respondents still stabled the horse for part of the day, and 84% still fed the horse dry hay.⁸⁹

Because the effects of repetitive episodes on the development of irreversible changes in lung structure have yet to be deter­mined, clients should be encouraged to implement effective husbandry changes that minimize the recurrences of RAO and prevent pulmonary parenchymal damage.