Clinical, radiographic, endoscopic, bronchoalveolar lavage and lung biopsy findings in horses with exercise-induced pulmonary hemorrhage

A clinical study was performed to determine whether clinical, endoscopic, radiographic, bronchoalveolar lavage (BAL) cytological, and pulmonary biopsy findings could be correlated in horses with exercise-induced pulmonary hemorrhage (EIPH) compared with controls. Racing standardbred horses were selected as either EIPH (n = 10) or control (n = 10), based on repeated postexertional endoscopy of the lower airways. Complete physical and respiratory examinations were performed and blood samples were submitted for arterial blood gas analysis, hematologic study, and fibrinogen determination. Bilateral chest radiographs were taken with the horse standing, and a BAL sample was obtained for cytological examination. Lung was biopsied transcutaneously. Weighted scores were calculated for clinical, radiographic, and pulmonary biopsy findings. The conclusion was that only routine physical examination may help the clinician when EIPH is suspected in horses, especially when there are abnormal findings on percussion of the caudodorsal areas of the chest.     

Sex variation in the prevalence of exercise-induced pulmonary haemorrhage in racing quarter horses

Post race endoscopy was carried out on 255 two-year-old quarter horses and exercise-induced pulmonary haemorrhage (EIPH) was diagnosed in 166 (65 per cent) of them. Visible epistaxis was seen in a higher proportion of geldings than in either mares or stallions. The prevalence of EIPH was similar in mares (73 per cent) and in geldings (74 per cent). A significantly lower prevalence (49 per cent) was noted in stallions (P less than 0.01). It was concluded that a sex variation in the prevalence of EIPH exists in two-year-old quarter horses.     

Equine bronchoalveolar lavage cytology: survey of Thoroughbred racehorses in training

SUMMARY Sixty-two Thoroughbred horses aged between 1 and 7 years in training in Sydney had bronchoalveolar lavage (BAL) samples collected for cytological examination. All horses, except the yearlings and those with a cough, had raced at the time of the examination and the trainers reported satisfactory performance. Free erythrocytes were found In 73% of samples and haemoslderophages In 90% of the samples, Indicating Immediate or past occurrences of exercise-Induced pulmonary haemorrhage (EIPH). Bronchoalveolar fluid from the yearlings contained significantly less (P / 0.05) erythrocytes and haemosiderophages than samples from horses In other age groups. In the older horses, there was also more haemosiderln within the macrophages. No differences In BAL cytology could be attributed to gender, and there was no relationship between BAL cytology and racing performance. The main cytological findings were (mean ± sd): total nucleated cells - 832 ± 578/μL with the main cell types being: macrophages - 59 ± 10% (haemosiderophages - 20 ± 24%); neutrophlls - 9 ± 6%; lymphocytes - 31 ± 9%. The erythrocyte count was 10.3 ± 17.7% of the total cell count. Horses with chronic coughing had a higher proportion of macrophages and a lower proportion of lymphocytes in the leucocytes obtained from BAL. There was a higher occurrence of EIPH detected In BAL findings than that previously reported when endoscopic examination has been used to diagnose EIPH. The occurrence and severity of EIPH as Indicated by the BAL findings was found to be related to exercise Intensity. The cytological findings were similar to those reported in horses in the northern hemisphere. We conclude that BAL cytology may be useful In the diagnosis of various lower respiratory tract disorders and that exercise-induced pulmonary haemorrhage occurs In virtually all horses In race training.     

Regional pulmonary veno-occlusion: a newly identified lesion of equine exercise-induced pulmonary hemorrhage

Exercise-induced pulmonary hemorrhage (EIPH) is common in horses following intense exertion, occurring in up to 75% of racing Thoroughbreds and Standardbreds. In spite of this, the pathogenesis of EIPH is poorly understood. In 7 racing Thoroughbred horses with EIPH, 6 sections were collected from the left and right lung, representing the cranial, middle, and caudal region of the dorsal and ventral lung (84 sites total). Grossly, both right and left lungs had numerous dark brown to blue-black foci along the caudodorsal visceral pleura. Tissue sections were stained with hematoxylin-eosin, Masson's trichrome, and Prussian blue. Verhoeff Van Gieson and immunohistochemistry for alpha-smooth muscle actin were used to assess the pulmonary vasculature. Histologic scores (HS = 0-3) were assigned to each region/slide for the presence and severity of 5 findings: interstitial fibrosis, hemosiderin accumulation, pleural/interlobular septal thickness, arterial and venous wall thickness, and evidence of angiogenesis (maximum cumulative HS = 15). Thirty-nine of the 84 (46%) sections were histologically normal (HS = 0); 33/84 (39%) were mildly to moderately affected, with small amounts of hemosiderin and fibrosis (HS = 1-9) while 12/84 (14%), primarily from the dorsocaudal lung, had severe vascular remodeling, fibrosis, and hemosiderin accumulation (HS = 10-15). In the latter, veno-occlusive remodeling of the intralobular veins colocalized with hemosiderosis, fibrosis, hypertrophy of vessels within the pleura, and interlobular septa and bronchial neovascularization. We propose that regional veno-occlusive remodeling, especially within the caudodorsal lung fields, contributes to the pathogenesis of EIPH, with the venous remodeling leading to regional vascular congestion and hemorrhage, hemosiderin accumulation, fibrosis, and bronchial angiogenesis.     

Effect of airway disease on blood gas exchange in racehorses

Inflammatory Airway Disease (IAD), exercise-induced pulmonary hemorrhage (EIPH), and upper airway obstruction (UAO) are common respiratory tract diseases that can decrease performance. The purpose of this retrospective study was to compare bronchoalveolar lavage fluid cytology and arterial blood gas analysis during a treadmill test by poorly performing racehorses presented to Purdue University. One hundred thirty-two horses with a history of poor performance were included in this study. Ten horses with no history or diagnosis of EIPH, IAD, or UAO served as controls. Horses were evaluated by rhinolaryngoscopy for upper airway abnormalities and underwent a standardized treadmill test, and samples were collected for blood gas analysis. Horses with IAD or EIPH had a more severe exercise-induced hypoxemia, (mean+/-SD; 84.8+/-1.5 and 86.0+/-1.7 mm Hg average Pao2, respectively), than horses in the control group (92.8+/-2.1 mm Hg). The average Pao2 of horses with only UAO (88.3+/-3.3 mm Hg) was not significantly different from control horses. Gas exchanges were the most severely impaired in horses affected with both EIPH and UAO because they exhibited the lowest Pao2 and highest Paco2 values (66.5+/-15.2 and 52.2+/-6.3 mm Hg, respectively).     

Efficacy of water vapor-saturated air in the treatment of exercise-induced pulmonary hemorrhage in Thoroughbred racehorses

The efficacy of water vapor-saturated air as a treatment for horses with exercise-induced pulmonary hemorrhage (EIPH) was studied. Horses selected for study (n = 14) had grade 1 or greater hemorrhage in the trachea after a minimum of 4 breezes between 0.8 and 1 km, as determined by endoscopy. Nine horses were treated with water vapor-saturated air; 5 horses were not treated. When the mean and maximal EIPH scores from the pretreatment period were compared with the mean and maximal EIPH scores from the treatment period in both treated and nontreated groups, there was no significant difference between groups. There was a suggestion of a linear relationship between exercise speed and the mean EIPH score of the first 4 breezes in all 14 horses.     

Validation of a Rapid Bronchoalveolar Lavage Fluid Colorimetric Evaluation for Assessing the Severity of Exercise-Induced Pulmonary Hemorrhage in Horses in Field Conditions

Diagnosis and assessment of severity of exercise-induced pulmonary hemorrhage (EIPH) relies on postexercise visualization of fresh blood in the airways via tracheobronchoscopic examination (TBE) and/or counting erythrocytes in bronchoalveolar lavage fluid (BALFRBC). Determining the BALFRBC is more sensitive than TBE but its usefulness is hampered by the need to have BALFRBC counted at a laboratory. We explored the feasibility of evaluating the severity of EIPH by using a color chart comprised of five shades of red and matching those colors with the color of BALF immediately following collection. To validate the technique, sets of ten BALF samples with known BALFRBC numbers were created and randomly shown to two groups of 18 observers who independently matched the color of the BALF with one of the shades of red displayed on the screen of a smartphone. Interobserver and intra-observer agreements regarding colors were > 0.9. The utility of the color chart was further validated under field conditions at two barrel racing events where 63 BALF samples were collected from 21 horses and BALF color was graded independently by three observers. The number of BALFRBC in the 63 samples ranged from 25–1,100,000/μL. EIPH was diagnosed in 39 samples based on the detection of color, and all 5 colors were matched multiple times with BALF samples. Overall, the color of the BALF was related to the number of BALFRBC. Colorimetric evaluation of BALF represents a practical and reliable option for rapid postexercise assessment of the presence and severity of EIPH.     

Pulmonary response to airway instillation of autologous blood in horses

REASONS FOR PERFORMING STUDY: Exercise-induced pulmonary haemorrhage (EIPH) occurs in the majority of horses performing strenuous exercise. Associated pulmonary lesions include alveolar and airway wall fibrosis, which may enhance the severity of EIPH. Further work is required to understand the pulmonary response to blood in the equine airways. OBJECTIVES: To confirm that a single instillation of autologous blood into horse airways is associated with alveolar wall fibrosis, and to determine if blood in the airways is also associated with peribronchiolar fibrosis. METHODS: Paired regions of each lung were inoculated with blood or saline at 14 and 7 days, and 48, 24 and 6 h before euthanasia. Resulting lesions were described histologically and alveolar and airway wall collagen was quantified. RESULTS: The main lesion observed on histology was hypertrophy and hyperplasia of type II pneumocytes at 7 days after blood instillation. This lesion was no longer present at 14 days. There were no significant effects of lung region, treatment (saline or autologous blood instillation), nor significant treatment-time interactions in the amount of collagen in the interstitium or in the peribronchial regions. CONCLUSION: A single instillation of autologous blood in lung regions is not associated with pulmonary fibrosis. POTENTIAL RELEVANCE: Pulmonary fibrosis and lung remodelling, characteristic of EIPH, are important because these lesions may enhance the severity of bleeding during exercise. A single instillation of autologous blood in the airspaces of the lung is not associated with pulmonary fibrosis. Therefore the pulmonary fibrosis described in EIPH must have other causes, such as repetitive bleeds, or the presence of blood in the pulmonary interstitium in addition to the airspaces. Prevention of pulmonary fibrosis through therapeutic intervention requires a better understanding of these mechanisms.     

The Effectiveness of Immunotherapy in Treating Exercise-Induced Pulmonary Hemorrhage

Inflammatory airway disease has been linked to exercise-induced pulmonary hemorrhage (EIPH), and consequently, we hypothesized that immunomodulation via concentrated equine serum (CES) treatment would reduce EIPH as evidenced by red blood cell (RBC) concentrations in bronchoalveolar lavage fluid (BALF). Separate trials were conducted on Thoroughbred horses treated with either CES (n = 6) or placebo (0.9% saline; n = 4). All horses completed pre-treatment and post-treatment (2 and 4 weeks after initiating treatment) maximal exercise tests on a 10% inclined treadmill (1 m/s/min increments to fatigue) over a 10-week period (2−3 weeks between tests), with bronchoalveolar lavage (BAL) performed 30 minutes after exercise. Treatment ensued 10 days after the pre-treatment exercise test, with horses receiving a series of five CES or placebo injections 24 hours apart (20 mL intratracheal and 10 mL intravenously), with subsequent weekly injections for 5 weeks thereafter. After CES treatment, both EIPH (RBC in BALF) and inflammation (white blood cell concentration [WBC] in BALF) were significantly diminished by the 4-week posttreatment run, demonstrating 46 ± 12% and 24 ± 11% decreases, respectively (P < 0.05). In contrast, EIPH was elevated significantly at the 4-week time point, and inflammation remained constant in the placebo trial. In conclusion, these preliminary data suggest that therapeutic intervention involving immunomodulation may represent a viable approach to reducing the severity of EIPH.     

Exercising Blood Gas Analysis, Dynamic Upper Respiratory Tract Obstruction, and Postexercising Bronchoalveolar Lavage Cytology—A Comparative Study in Poor Performing Horses

Respiratory abnormalities are common causes of decreased performance in horses presumably because of impaired pulmonary gas exchange. The objectives of the present study were to describe respiratory abnormalities in poorly performing horses and to investigate the relationships between dynamic upper respiratory tract (URT) videoendoscopy, postexercising bronchoalveolar lavage (BAL) cytology, and exercising arterial blood gas analysis. Medical records of 93 horses with exercise intolerance, which presented for treadmill evaluation, were reviewed. Relationships between horse demographics, treadmill endoscopic findings, exercising blood gas values, and BAL cytology results were examined. A total of 25 (27%) horses had a URT obstruction and 91 (98%) horses had abnormal BAL cytology; 73 (78%) had evidence of inflammatory airway disease (IAD) and 83 (89%) had exercise-induced pulmonary hemorrhage (EIPH). In all, 39 (42%) horses had abnormal blood gas values. Dynamic URT obstruction was significantly associated with exercising hypoxemia (P = .036). There were no significant relationships between gas exchange and IAD or between EIPH. Out of 24 (26%) horses with combined URT obstruction and abnormal BAL, horses with URT obstruction and EIPH were more likely to be hypoxic during exercise (P = .037). It was concluded that horses with dynamic URT abnormalities are likely to have exercising hypoxemia. Although IAD and EIPH were commonly indentified in poor performers, they were not significantly associated with abnormal exercising blood gas analysis.     

Cytologic examination of specimens obtained by means of tracheal washes performed before and after high-speed treadmill exercise in horses with a history of poor performance

OBJECTIVE: To evaluate results of cytologic examination of specimens obtained by means of tracheal washes (TW) in 42 horses with a history of poor performance. DESIGN: Cross-sectional case series. ANIMALS: 42 horses with a history of poor performance. PROCEDURE: A TW was performed via endoscopy before and after horses exercised on a high-speed treadmill, and specimens were evaluated microscopically and graded. RESULTS: Ten (24%) horses were considered to be clinically normal before and after exercise. Pulmonary hemorrhage was diagnosed in 8 (19%) horses. One horse had evidence of exercise-induced pulmonary hemorrhage (EIPH) before exercise and pulmonary hemorrhage and allergic or inflammatory airway disease (IAD) after exercise. Five (12%) horses had IAD, and 1 had IAD and pulmonary hemorrhage after exercise. Seven (17%) horses had evidence of EIPH and IAD in both specimens. Four (10%) horses with EIPH had an increase in the proportion of hemosiderophages in the specimen obtained after exercise. Specimens obtained before exercise in 6 (14%) horses were not representative of the respiratory tract and could not be compared with specimens obtained after exercise. CLINICAL IMPLICATIONS: Interpretation of TW specimens obtained before and after exercise differed for only 5 of 36 (14%) horses. Specimens obtained after exercise were more likely to reveal airway disease. All specimens obtained after exercise adequately represented the respiratory tract, whereas 6 specimens obtained before exercise did not. Specimens obtained after exercise contained more airway secretions and had less cytologic evidence of pharyngeal contamination. Therefore, we recommend that TW samples be obtained after exercise in horses.     

Effects of furosemide on the racing times of Thoroughbreds

The effects of furosemide on the racing times of 79 horses without exercise-induced pulmonary hemorrhage (EIPH) and 52 horses with EIPH were investigated. Racing times were adjusted to 1-mile equivalent racing times by 2 speed handicapping methods, and analysis of covariance was used to adjust actual racing times by winning time and distance for each race. All 3 methods of determining racing time indicated that geldings without EIPH had significantly faster racing times (P less than 0.05) when given furosemide before racing than when furosemide was not given before racing. Females and colts without EIPH were determined to have faster racing times when furosemide was given before racing, but the difference was not significant. Geldings with EIPH had significantly faster racing times (P = 0.0231) when given furosemide before racing, as determined by one of the speed handicapping methods. There was a strong correlation (range 0.9314 to 0.9751) between the 1-mile equivalent racing times, as determined by the 2 speed handicapping methods for horses with and without EIPH. Furosemide failed to prevent the development of EIPH in many horses that were previously considered to be EIPH-negative. When given furosemide, 62 (25.3%) of 235 EIPH-negative horses were EIPH-positive after racing. Furosemide had questionable efficacy for prevention of EIPH in known EIPH-positive horses. Thirty-two (61.5%) of 52 EIPH-positive horses given furosemide before a race remained EIPH-positive after that race.     

Exercise-induced pulmonary hemorrhage.

EIPH is a condition affecting virtually all horses during intense exercise worldwide. The hemorrhage originates from the pulmonary vasculature and is distributed predominantly bilaterally in the dorsocaudal lung lobes. As the condition progresses, the lung abnormalities extend cranially along the dorsal portions of the lung. An inflammatory response occurs in association with the hemorrhage and may contribute to the chronic sequela. Although conflicting opinions exist as to its affect on performance, it is a syndrome that is thought to increase in severity with age. The most commonly performed method to diagnose EIPH at the present time is endoscopy of the upper airway alone or in combination with tracheal wash analysis for the presence of erythrocytes and hemosiderophages. Because horses may not bleed to the same extent every time and the bleeding may originate from slightly different locations, these diagnostic procedures may not be extremely sensitive or quantitative. At this time, there is no treatment that is considered a panacea, and the currently allowed treatments have not proven to be effective in preventing EIPH. Future directions for therapeutic intervention may need to include limiting inflammatory responses to blood remaining within the lungs after EIPH.     

Pulmonary vascular pressures of strenuously exercising Thoroughbreds after administration of varying doses of frusemide

The frusemide dose-response for attenuation of exercise-induced pulmonary capillary hypertension was studied in 7 healthy, exercise-conditioned Thoroughbred horses using previously described haemodynamic procedures. Four different doses of frusemide were tested: 250 mg regardless of bodyweight (amounting to 0.56 +/- 0.03 mg/kg bwt), 1.0 mg/kg bwt, 1.5 mg/kg bwt and 2.0 mg/kg bwt. Frusemide was administered i.v., 4 h before exercise. Haemodynamic data were obtained at rest and during treadmill exercise performed at 14.2 m/s on a 3.5% uphill grade; this workload elicited maximal heart rate of horses. Airway endoscopy was performed post exercise to detect exercise-induced pulmonary haemorrhage (EIPH). In standing horses, frusemide administration resulted in a significant (P<0.05) decrease in mean pulmonary arterial, pulmonary capillary and pulmonary artery wedge pressures, but significant differences among the various frusemide doses were not observed. In the control experiments, exercise caused significant increments in the right atrial as well as pulmonary arterial, wedge, and capillary pressures, and all horses experienced EIPH. Following frusemide administration, the exercise-induced rise in right atrial and pulmonary vascular pressures was significantly attenuated, but significant differences between the frusemide doses of 250 mg, 1.0 mg/kg, and 1.5 mg/kg were not discerned and all horses remained positive for EIPH. Although a further significant (P<0.05) attenuation of the exercise-induced rise in pulmonary capillary blood pressure occurred when frusemide dose increased from 250 mg to 2.0 mg/kg bwt, all horses still experienced EIPH. It is concluded that a linear response to increasing frusemide dosage in terms of attenuation of the pulmonary capillary hypertension does not exist in strenuously exercising Thoroughbred horses.     

Efficacy of furosemide in the treatment of exercise-induced pulmonary hemorrhage in Thoroughbred racehorses

The repeatability of endoscopic observations of exercise-induced pulmonary hemorrhage (EIPH) and the efficacy of furosemide as a prophylactic treatment of horses with EIPH were studied in Thoroughbred race horses after consecutive breezes (at or near maximum speed, approx 16 m/s). Of 56 horses examined greater than or equal to 2 times, 21 (38%) had identical EIPH scores, whereas 26 (46%) and 9 (16%) had scores that differed by greater than or equal to 1 grade. In 56 nontreated horses, there was good agreement between 2 consecutive observations (K = 0.59, Z = 4.54, P less than 0.001). Similar comparisons after placebo (saline solution) treatment of 21 horses yielded fair to good agreement, whereas poorer agreement was seen after furosemide treatment of 23 horses. Comparison of average and maximum EIPH scores of 44 horses with a minimum of 4 observations (2 nontreated, 1 saline-treated, and 1 furosemide-treated) indicated that although furosemide did not stop EIPH, it did reduce the EIPH score in 28 (64%) horses.     

Association between exercise-induced pulmonary hemorrhage and performance in Thoroughbred racehorses

OBJECTIVE: To determine whether exercise-induced pulmonary hemorrhage (EIPH) was associated with racing performance inThoroughbred horses not medicated with furosemide and not using nasal dilator strips. DESIGN: Observational cross-sectional study. ANIMALS: 744 two- to 10-year-old Thoroughbred horses racing in Melbourne, Australia. PROCEDURE: Horses were enrolled prior to racing, and a tracheobronchoscopic examination was performed after 1 race. Examinations were recorded on videotape, and presence and severity (grade 0 to 4) of EIPH were subsequently determined by 3 observers blinded to the horses' identity. Race records were abstracted for each horse examined. RESULTS: Overall, 52.1% of horses eligible for participation in the study were examined, and horses that were examined did not differ from horses that were not examined in regard to age, sex distribution, or proportion of horses that won or finished in the first 3 positions. Horses with EIPH grades < 1 were 4.0 times as likely to win, 1.8 times as likely to finish in the first 3 positions, and 3.03 times as likely to be in the 90th percentile or higher for race earnings as were horses with grades > 2. Horses with EIPH grades > 1 finished significantly farther behind the winner than did horses without EIPH. However, odds that horses with grade 1 EIPH would win or finish in the first 3 positions were not significantly different from odds for horses without EIPH. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that EIPH is associated with impaired performance in Thoroughbred racehorses not medicated with furosemide and not using nasal dilator strips.     

The role of alveolar macrophages in the pathogenesis of recurrent airway obstruction in horses

When challenged with allergens and pro-inflammatory agents, such as Aspergillus fumigatus (AF), hay dust solution (HDS) and lipopolysaccharide (LPS), the innate immune response will not only activate the immune system but also increase the amount of pro-inflammatory cytokines in the bronchoalveolar space. The aim of this study was to assess the response of equine alveolar macrophages to different aerosolized challenges and to investigate the differences in this response between horses susceptible or nonsusceptible to recurrent airway obstruction (RAO). Seven susceptible and 5 nonsusceptible horses were challenged with saline, LPS, HDS, or AF, and bronchoalveolar lavage (BAL) cytology, total cell counts, and lung function were assessed. In addition, alveolar macrophages were isolated 6 and 24 hours after challenge, and macrophage mRNA expression of tumor necrosis factor (TNF)-alpha and interleukins (IL) IL-1beta, IL-6, IL-8, and IL-10 were measured by means of real-time (RT) polymerase chain reaction (PCR). There was a significant difference in lung function, neutrophil ratios, and total cell counts in the bronchoalveolar lavage fluid between RAO-susceptible and nonsusceptible horses. In addition, the expression of TNF-alpha, IL-1beta, and IL-8 by alveolar macrophages after challenges were higher in susceptible horses, than in nonsusceptible horses. In contrast, I1-6, considered an anti-inflammatory cytokine, showed a higher expression in nonsusceptible horses 6 hours after inhalation challenge with allergens and pro-inflammatory antigens. These data suggest that the differences between susceptible and nonsusceptible horses to RAO are not only dependent on adaptive immunity but also start with an innate immune response.     

Exercise-induced pulmonary hemorrhage in the racing appaloosa horse

A retrospective survey was carried out at a Florida racetrack in 1984, to determine the prevalence of exercise-induced pulmonary hemorrhage (EIPH) in Appaloosa horses. Of 94 horses examined endoscopically between 30 and 90 minutes after racing, forty-nine (52%) were found to exhibit some degree of EIPH. There was a significantly increased prevalence of EIPH with increasing age in this population. An increased prevalence of EIPH was noted with mares and geldings as compared with stallions, but this was not statistically significant. More horses bled from the right lung than from the left, but the difference was statistically insignificant. An increased prevalence of EIPH with distance run was evident, but this was not significant when animals of the same age were examined.     

Bronchoalveolar lavage findings in horses with exercise intolerance.

Significant differences were detected by bronchoalveolar lavage (BAL) between horses racing successfully and those showing exercise intolerance. Neutrophil percentage, haemosiderophage percentage and total bacterial numbers were significantly elevated in horses with exercise intolerance. BAL provided a more accurate indication of the incidence and extent of exercise induced pulmonary haemorrhage (EIPH) than visual inspection.     

Exercise-induced pulmonary haemorrhage in the horse: results of a detailed clinical, post mortem and imaging study. IV. Changes in the bronchial circulation demonstrated by C.T. scanning and microradiography

The purpose of this study was to use radiographic contrast techniques and special imaging methods to identify and high-light bronchial arterial involvement in lung lesions associated with exercise-induced pulmonary haemorrhage (EIPH) in horses. The lungs from four horses with histories of EIPH were prepared for computerised tomographic scanning and microradiography by perfusing the broncho-oesophageal artery with a mixture of red latex and either barium or iodine contrast materials while the pulmonary supply received only blue latex. Computerised tomographic scan slices of the prepared inflated lungs were obtained from the caudal tip of the lung to the hilus. Microradiography of selected lung slices was also performed on a Faxitron. Diffuse areas of increased density, with preferential bronchial arterial supply noted on the computerised tomographic scans were confirmed by microradiography. Dense focal and diffuse plexuses of markedly hypertrophied and highly branched bronchial arterial networks were identified, centred around certain small airways. The vascular supply to these plexuses was recruited predominantly from neighbouring bronchial vessels, and in some cases, from the enlarged vasa vasorum of pulmonary arteries sending anastomoses to the affected areas. The authors conclude that bronchial vascular lesions in EIPH cases are the likely origin of haemorrhage; that small airway disease is the probable initiating stimulus for bronchial vascular proliferation in these lesions; and that the morphology and nature of the neovascular tissue in these lesions provides the conditions leading to haemorrhage in the lungs of horses with EIPH.