Very high pressures are required to cause stress failure of pulmonary capillaries in thoroughbred racehorses.

Thoroughbred horses develop extremely high pulmonary vascular pressures during galloping, all horses in training develop exercise-induced pulmonary hemorrhage, and we have shown that this is caused by stress failure of pulmonary capillaries. It is known that the capillary transmural pressure (Ptm) necessary for stress failure is higher in dogs than in rabbits. The present study was designed to determine this value in horses. The lungs from 15 Thoroughbred horses were perfused with autologous blood at Ptm values (midlung) of 25, 50, 75, 100 and 150 mmHg, and then perfusion fixed, and samples (dorsal and ventral, from caudal region) were examined by electron microscopy. Few disruptions of capillary endothelium were observed at Ptm < or = 75 mmHg, and 5.3 +/- 2.2 and 4.3 +/- 0.7 breaks/mm endothelium were found at 100 and 150 mmHg Ptm, respectively. Blood-gas barrier thickness did not change with Ptm. At low Ptm, interstitial thickness was greater than previously found in rabbits but not in dogs. We conclude that the Ptm required to cause stress failure of pulmonary capillaries is between 75 and 100 mmHg and is greater in Thoroughbred horses than in both rabbits and dogs.

Autologous blood instillation alters respiratory mechanics in horses.

To investigate physiological consequences of autologous blood instillation in the lungs of healthy horses, respiratory mechanics and bronchial response to histamine were studied in 8 Thoroughbreds before and after introducing autologous blood (n = 5) and sterile saline solution (n = 3) into their lungs. Blood instillation resulted in a decrease in dynamic compliance (Cdyn) and increased respiratory resistance (R). Bronchial sensitivity and reactivity were unchanged after blood introduction. There were no significant changes in pulmonary mechanics or bronchial response after saline instillation.

Morphologic and morphometric characterization of lung collagen content in clinically normal adult thoroughbreds in race training.

Interstitial and bronchointerstitial pulmonary patterns are commonly observed in thoracic radiographs of Thoroughbreds. Prominent interstitial and bronchointerstitial pulmonary patterns are observed in clinically normal horses, and in horses with respiratory tract disease. Until recently, the relevance of these pulmonary patterns was not known. Previous studies indicated that bronchiolitis, bronchiolar epithelial hyperplasia, epithelial metaplasia, and bronchial arteriolar recruitment correlated strongly with the prominence of the interstitial and bronchointerstitial pulmonary patterns observed radiographically. We examined the content and distribution of collagen in the lungs of 7 clinically normal Thoroughbreds in race training. After standardized fixation, lung tissue was treated with a compound that selectively stains collagen. Standard morphometric techniques were used to determine the volume density of parenchymal tissue and parenchymal airspace, mean linear intercept (estimate of alveolar size), alveolar surface area-to-volume ratio, percentage of parenchyma composed of collagen, percentage of airway wall composed of collagen, and airway wall thickness. These values were compared with radiographic and histopathologic scores obtained from the same horses. The volume density of parenchymal tissue and small airway wall thickness correlated strongly with the prominence of the bronchial and bronchointerstitial pulmonary patterns observed radiographically. Small airway thickness was also highly correlated with the perceived prominence of the interstitial pulmonary patterns observed radiographically, and morphometric estimates of parenchymal tissue and parenchymal collagen. There were also strong correlations between the volume density of parenchymal tissue, the percentage of parenchymal collagen, peribronchiolar mononuclear cell infiltrates, and bronchiolar mucosal plication estimates.(ABSTRACT TRUNCATED AT 250 WORDS)

Evaluation of a technique for detection of pulmonary hemorrhage in horses, using carbon monoxide uptake.

The diffusing capacity for carbon monoxide (DLCO) and the functional residual capacity (FRC) of the lung were measured in 5 healthy Thoroughbreds before and after instillation of autologous blood into their lungs, in an attempt to develop a method to quantitate extravascular blood in the lungs of horses with exercise-induced pulmonary hemorrhage. Mean (+/- SD) baseline values of DLCO and FRC were 333.8 +/- 61.9 ml/min/mm of Hg and 21.464 +/- 4.156 L, respectively. Blood instillation resulted in decreases in DLCO and FRC. The paradoxic decrease in DLCO (we were expecting to find an increase owing to blood in the airspaces, as has been reported in people) appears to be associated with the bronchoscopic procedure and with presence of blood in the airways. We concluded that rebreathing DLCO measurements were not effective for detecting blood introduced bronchoscopically into the lungs of horses.

Comparative aspects of the strength of pulmonary capillaries in rabbit, dog, and horse.

In previous studies of rabbit and dog lung, we demonstrated stress failure of pulmonary capillaries at high transmural pressures (Ptm). The Ptm necessary to elicit stress failure was 40 cmH2O higher in dog than rabbit, and the total blood-gas barrier (BGB) thickness was greater in dog than rabbit. This suggests that stress failure may be related to BGB thickness, and is consistent with the Laplace relationship which states that wall stress is proportional to capillary radius but inversely proportional to wall thickness. In the present studies, we compared BGB thickness and an index of capillary radius in lungs from 3 rabbits, 3 dogs, and 2 horses perfusion fixed at a Ptm of approximately 30 cmH2O. Thicknesses of the BGB were measured at right angles to the barrier at random points on the capillary wall determined by test line intersections. Capillary radius was determined from the mean of major and minor axes measured on electron micrographs. Capillary pressure for failure in the horse was taken to be the mean of pulmonary arterial and left atrial pressures observed in galloping thoroughbreds known to develop exercise-induced pulmonary hemorrhage, although the actual pressure required for failure may be less than this. Average capillary radii were 3.6, 3.4, and 3.2 microns for rabbits, dogs, and horses, respectively. We found that the BGB was thinnest in the rabbit, intermediate in the dog, and thickest in the horse. Calculated capillary wall stress values for the median total BGB thickness at a nominal Ptm of 30 cmH2O were 2.5 x 10(4), 1.7 x 10(4), and 1.5 x 10(4) N.m-2 for rabbits, dogs, and horses, respectively. This species ranking fits with the pressures required to cause stress failure which are approximately 50, 90, and 130 cmH2O in rabbit, dog, and horse, respectively. We conclude that the differences in capillary radius of curvature and BGB thickness account for some of the observed differences in Ptm necessary to cause stress failure. However, other factors may also be important in determining the strength of the BGB.

Endoscopic anatomy and map of the equine bronchial tree.

To develop a bronchoscopic map of the equine respiratory tree, the major airways of the lungs of 6 healthy Thoroughbred horses were systematically explored with a flexible fibreoptic endoscope through a tracheostomy while the horses were sedated in stocks. With the carina as the reference point, measurements were made of distances to the branches of the major airways using markers on the shaft of the endoscope. All branches were explored until the narrowing of their diameters prevented further advancement of the endoscope. Positions of origins of branches from the parent bronchus were recorded in relation to a 12 h clock. Branching patterns of the right and left lungs were similar. Seventeen branches of the principal and caudal lobar bronchi of the left lung, and 18 branches of the principal and caudal lobar bronchi of the right lung were identified. Mean explorable distances from the carina to the ends of the right and left caudal lobar bronchi were 34.0 +/- 3.5 (sd) and 34.5 +/- 3.0 cm, respectively. Generally, smaller horses had shorter explorable bronchial lengths. Branching patterns of the parent bronchi were fairly consistent among horses, particularly the branches closest to the carina. After endoscopy and euthanasia, the lungs were removed, and dried with pressurised air flowing through them for 7-10 days. Attempts to explore the airways of the dried lungs endoscopically were relatively unsuccessful, because airways were much smaller in the dried lungs, and many of the branches were distorted when compared with their antemortem appearances. However, having a dried lung specimen as a reference during the bronchoscopic procedure was useful for maintaining orientation in the lungs. Radiographs were used to estimate the location of the origin and destination of each airway branch in relation to the nearest intercostal space. This makes the airway map useful when lesions identified radiographically are to be lavaged.

Pulmonary intravascular macrophages in horses and ponies.

Seven horses (4 anesthetized and 3 awake) and 2 ponies (anesthetized) were studied to evaluate the high sensitivity of the pulmonary circulation of the horse to various blood-borne particles, and to establish the presence of intravascular macrophages in the lung. Pulmonary and systemic pressures and cardiac output before and during particle injection were measured in some animals. An anesthetized foal had a large increase in pulmonary arterial pressure (32 and 34 mm of Hg) within 1 minute of IV administration of small test doses of radioactively labeled liposomes (2.5 mumol/kg of body weight) or a 1% suspension of blue pigment (0.3 ml/kg), respectively. Quantitative real-time gamma camera imaging of the foal revealed high retention of the labeled liposomes during the first pass through the lungs; retention persisted throughout the experiment. Postmortem analysis revealed 55 and 47% lung retention of liposomes and blue pigment, respectively. The 2 anesthetized ponies had increased pulmonary artery pressure of 34 +/- 7 mm of Hg, decreased cardiac output, and 42% lung retention after administration of 1% blue pigment (0.2 ml/kg), whereas 3 awake horses had increased pressure of 28 +/- 9 mm of Hg after 1.8 x 10(8) (1.8-microns-diameter) latex microspheres/kg. None of the injected particles caused vascular obstruction, and they do not cause pulmonary vascular reactivity in species that lack pulmonary intravascular macrophages. Finally, 3 horses (1 anesthetized and 2 awake) were infused IV with small doses of the blue pigment, and their lungs were perfusion-fixed to identify specific labeling of the pulmonary intravascular macrophages.(ABSTRACT TRUNCATED AT 250 WORDS)

Stress failure of pulmonary capillaries in racehorses with exercise-induced pulmonary hemorrhage.

Bleeding into the lungs in thoroughbreds is extremely common; there is evidence that it occurs in essentially all horses in training. However, the mechanism is unknown. We tested the hypothesis that exercise-induced pulmonary hemorrhage (EIPH) is caused by stress failure of pulmonary capillaries. Three thoroughbreds with known EIPH were galloped on a treadmill, and after the horses were killed with intravenous barbiturate the lungs were removed, inflated, and fixed for electron microscopy. Ultrastructural studies showed evidence of stress failure of pulmonary capillaries, including disruptions of the capillary endothelial and alveolar epithelial layers, extensive collections of red blood cells in the alveolar wall interstitium, proteinaceous fluid and red blood cells in the alveolar spaces, interstitial edema, and fluid-filled protrusions of the endothelium into the capillary lumen. The appearances were consistent with the ultrastructural changes we have previously described in rabbit lungs at high capillary transmural pressures. Actual breaks in the endothelium and epithelium were rather difficult to find, and they were frequently associated with platelets and leukocytes that appeared to be plugging the breaks. The paucity of breaks was ascribed to their reversibility when the pressure was lowered and to the fact that 60-70 min elapsed between the gallop and the beginning of lung fixation. Capillary wall stress was calculated from pulmonary vascular pressures measured in a companion study (Jones et al. FASEB J. 6: A2020, 1992) and from measurements of the thickness of the blood-gas barrier and the radius of curvature of the capillaries. The value was as high as 8 x 10(5) dyn/cm2 (8 x 10(4) N/m2), which exceeds the breaking stress of most soft tissues. We conclude that stress failure of pulmonary capillaries is the mechanism of EIPH.

Radiographic and microscopic correlation of diffuse interstitial and bronchointerstitial pulmonary patterns in the caudodorsal lung of adult thoroughbred horses in race training.

Complete thoracic radiographic examinations were performed on 7 horses ranging in age from 24 to 60 months, followed by in-situ lung fixation. Radiographs were examined by 3 radiologists for the presence, degree and distribution of generalised pulmonary patterns within a region of interest in the caudodorsal lung. Pulmonary tissue was obtained from 12 sites within a designated volume of interest in the caudodorsal lung, corresponding to the area of interest evaluated radiographically, and examined for the presence, character and severity of microscopic lesions. Radiographic findings within the volume of interest consisted of mild to moderate bronchial, bronchointerstitial, or interstitial pulmonary patterns. Interstitial and bronchointerstitial radiographic findings were related to severity of peribronchiolar mononuclear cell infiltrates, the degree of bronchiolar mucosal plication, and alveolar capillary and peribronchial blood vessel erythrocyte content. The severity of the interstitial radiographic pattern was inversely associated with the perceived diagnostic quality of the radiographic examinations. There was no evidence of spatial variation in the severity of the microscopic changes examined in this limited pulmonary region. Inter-rater reliability between radiologists was good in the assessment of diagnostic quality of the radiographic examinations but poor in assessing severity of the primary generalised pulmonary patterns within the radiographic region of interest.

Ultrastructural mucosal injury after experimental ischemia of the ascending colon in horses.

The ultrastructural injury that develops sequentially in the ascending colon during experimentally induced ischemia was examined in 6 halothane-anesthetized horses. Colonic ischemia was created by 2 types of vascular occlusion 24 cm proximal and distal to the pelvic flexure. In all horses, transmural vascular compression was created. The colonic venous circulation was obstructed in 3 horses, whereas in the other 3 horses, arterial and venous circulation was obstructed. Two additional horses were anesthetized as controls for determination of any morphologic alterations associated with the experimental protocol. Full-thickness colonic biopsy specimens were obtained from the antimesenteric border of the pelvic flexure at 0, 0.25, 0.5, 1, 1.5, 1.75, 2, 2.25, 2.5, 3, 3.5, 4, 4.5, and 5 hours during occlusion, and were studied by light and transmission electron microscopy. Morphologic alterations did not develop in the colon of control horses. Mucosal congestion was observed by light microscopy in the colon of horses with experimentally induced ischemia, but congestion developed early in those with obstructed colonic venous circulation, compared with those having arterial and venous obstruction. Inter- and intracellular vacuolation and loss of staining initially resulted in groups of 3 to 5 superficial luminal epithelial cells. Alterations in the glandular epithelium lagged behind those in the superficial epithelium, but were observed in both groups by 2 hours of obstruction. These changes progressed to 100% sloughing of all epithelium by 4.5 to 5 hours. The initial cellular alterations, which were observed by transmission electron microscopy, developed at 0.25 hour in horses with colonic venous obstruction and was characterized by inter- and intracellular edema.(ABSTRACT TRUNCATED AT 250 WORDS)

Computer-assisted morphometry: point, intersection, and profile counting and three-dimensional reconstruction.

The use of computers in morphometry can involve 1) automated image analysis, semiautomated image analysis and point, intersection, intercept and profile counts of two-dimensional images on tissue sections with mathematical extrapolation to the third dimension, 2) direct measurement of volumes, surfaces, lengths, and curvature using x,y,z coordinates of serial sectioned images, or 3) stereologic techniques and serial sections which is a combination of 1 and 2 above. Automated and semiautomated image analysis are generally restricted to specimens that are characterized by differential contrast such as interalveolar septa in the lung or histochemically stained mucous granules in pulmonary epithelium. Point, intersection, and profile counts using hand-held, notebook PCs, portable PCs, or standard PCs and MS-DOS-based application programs are extremely efficient, precise, affordable, and convenient methods of quantitating average values of a population. When morphometric measurements of individual structures are required, computer-assisted three-dimensional reconstruction using x,y,z coordinates of the surface outline from serial sections is a tedious yet precise method. We describe a computer program that efficiently estimates mean caliper diameter, volume, and surface area with less than five percent error with five sections per structure. We also describe a program that does digital image subtraction on serial sections, superimposes digitally generated test systems on biological images, and accumulates point, intersection, and profile counts using a Macintosh II series computer.

Pulmonary blood flow distribution in anesthetized ponies.

Results of recent investigations in humans and dogs indicate that gravity-independent factors may be important in determining the distribution of pulmonary blood flow. To further evaluate the role of gravity-independent factors, pulmonary blood flow distribution was examined using 15-microns radionuclide-labeled microspheres in five prone ponies over 5 h of pentobarbital sodium anesthesia. The ponies were killed, and the lungs were excised and dried by air inflation (pressure 45 cmH2O). The dry lungs were cut into transverse slices 1-2 cm thick along the dorsal-ventral axis, parallel to gravity. Radioactivity of pieces cut from alternate slices was measured with a gamma well counter. The main finding was a preferential distribution of pulmonary blood flow to dorsal-caudal regions and higher flow in the center of each lung slice when compared with the slice periphery. Flow was lowest in cranial and ventral areas. Differences of +/- 2 SD were observed between core and peripheral blood flow. No medial-lateral differences were found. Pulmonary blood flow distribution did not change over 5 h of anesthesia, and the basic flow pattern was not different in the left vs. right lung. These results suggest that in the intact prone mechanically ventilated pony (inspired O2 fraction greater than or equal to 0.95) factors other than gravity are primary determinants of pulmonary blood flow.

Effects of the sex-linked dwarf gene (dw) on the expression of the muscular dystrophy gene (am) in chicken.

The sex-linked dwarf gene (dw) was introduced into companion muscular dystrophic (am) and nondystrophic (Am+) New Hampshire chicken lines to investigate influences of the dwarf gene on breast muscle weights, muscle fiber area, and the histological expression of muscular dystrophy. Dystrophic and nondystrophic chickens within dwarf or nondwarf genotypes were similar in body and carcass weights. Pectoralis and supracoracoideus muscle weights (as a percentage of adjusted carcass weight) were similar in nondystrophic dwarf and nondwarf males and females. In addition, pectoralis weight was similar in dystrophic dwarf males and dystrophic nondwarf males and females. However, pectoralis weight was significantly smaller in dystrophic dwarf females than in dystrophic nondwarf females, whereas supracoracoideus weight was significantly larger in dystrophic dwarf males than in dystrophic nondwarf males. Supracoracoideus weight was similar in dystrophic dwarf males and females and dystrophic nondwarf females. Pectoralis muscle fiber area was influenced by sex and by dwarf and dystrophy genotype. Muscle fiber area was larger in females than in males, smaller in dwarfs than in nondwarfs, and smaller in dystrophic than in nondystrophic muscles. Muscle fiber degeneration and adipose infiltration was more extensive in dystrophic than in nondystrophic females and males, and it was more advanced in dwarfs than in nondwarfs. Excessive acetylcholinesterase staining patterns were characteristic of dystrophic muscle in both dwarf and nondwarf genotypes. Nondystrophic and dystrophic dwarf male and female chickens are comparable substitutes for nondwarfs as biomedical models with respect to pectoralis histology, acetylcholinesterase staining pattern, and pectoralis muscle hypertrophy.

Vascular injury associated with naturally occurring strangulating obstructions of the equine large colon.

Ten horses were euthanatized before, during, or after surgery to correct severe volvulus of the large colon. At surgery, the colonic serosa changed from blue-gray, blue or purple toward a more normal pink in seven horses after the volvulus was corrected. The mucosa consistently remained black or dark red. Results of postmortem colonic microangiography revealed perfusion of the serosa and the circular and longitudinal muscle layers, but mucosal perfusion was limited by thrombosis in the muscularis mucosae and submucosa. There was evidence of thrombosis of the mesenteric colic vessels in six horses. Damage to the colonic vascular system, especially thromboembolism in the submucosa, may be an important limitation to colonic viability after surgical correction of volvulus of the large colon.

Microvascular circulation of the ascending colon in horses.

Microvascular circulation of the ascending colon in healthy horses was studied using microangiography, light microscopy, and scanning electron microscopy. The pelvic flexure with 30 cm of ventral and dorsal colon attached was removed from 14 adult horses immediately after horses were euthanatized. The lumen was flushed with warm water, and this section of the ascending colon was placed in a 37-C bath of isotonic NaCl. In sections from 8 horses, colic vessels were perfused with a radio-opaque medium for microangiography. After angiographic evaluation, tissue sections were prepared for light microscopic observation, using standard histologic methods. In sections from 6 horses, injection replicas were made by perfusing the vessels with 2 types of plastics. The results of microangiography, light microscopy, and scanning electron microscopy of vascular replicas were correlated, providing a comprehensive documentation of the microvasculature of the ascending colon at the pelvic flexure. Arteries branched from mesenteric colic vessels approximately every 2 cm toward the colonic tissue. Immediately after branching, arterial vessels formed an anastomotic plexus, the colonic rete. However, each branch from the colic vessel eventually continued into the colonic tissue. A second set of vessels originated from the colonic tissue. A second set of vessels originated from the colonic rete and supplied the mesenteric lymph nodes. Arterial vessels penetrated the tunica muscularis into the submucosa 3 to 4 cm toward the antimesenteric border forming a submucosal vascular network. From the submucosal arterioles, branching took place at right angles to supply the mucosal capillaries. Capillaries surrounded the colonic glands and anastomosed at the luminal surface, forming a superficial luminal honeycomb-appearing vascular plexus.(ABSTRACT TRUNCATED AT 250 WORDS)

Comparison of daily and seasonal exposures of young monkeys to ozone.

Oxidant air pollution tends to occur in both seasonal and daily cycles of polluted and clean air. To compare the effects of these 2 cycles, we exposed 2 groups of 7-month-old male monkeys to 0.25 ppm (0.49 mg/m3) of ozone (UV photometric standard) 8 h/day either daily or, in the seasonal model, days of alternate months during a total exposure period of 18 months. This is a longer ozone exposure period than any previously reported. A control group breathed only filtered air. Young monkeys were studied as their lungs are similar to those of man and their lungs, like those of man, grow over a period of several years. Monkeys from the seasonal exposure model, but not those exposed daily, had significantly increased total lung collagen content, chest wall compliance, and inspiratory capacity. All monkeys exposed to ozone had respiratory bronchiolitis with significant increases in related morphometric parameters. The only significant difference between seasonal and daily groups was in the volume fraction of macrophages. Even though the seasonally exposed monkeys were exposed to the same concentration of ozone for only half as many days, they had larger biochemical and physiological alterations and equivalent morphometric changes as those exposed daily. Lung growth was not completely normal in either exposed group. Long-term effects of oxidant air pollutants which have a seasonal occurrence may be more dependent upon the sequence of polluted and clean air than on the total number of days of pollution. Estimations of the risks of human exposure to seasonal air pollutants from effects observed in animals exposed daily may underestimate long-term pulmonary damage.

Age-related morphometric differences in responses of rat lungs to ozone.

The influence of age on morphologic changes in lungs of rats exposed to ozone was studied in female Sprague-Dawley rats, aged 60 and 444 days. Rats of both age groups were exposed continuously for 72 hr to either 0.35 or 0.80 ppm ozone, or to filtered air. Tissues were evaluated using light microscopic morphometry and scanning electron microscopy. The lungs from ozone-exposed 60-day-old rats had larger volume fractions of centriacinar lesions than lungs from exposed 444-day-old rats. Within each age group there was an observed dose response, with rats exposed to 0.80 ppm ozone having larger volume fractions of lesions than those exposed to 0.35 ppm. Only the 444-day-old rats lost body weight during the exposure period. They also had smaller fixed lung volumes than same-aged controls. All 60-day-old rats gained weight during the exposure period, although rats exposed to 0.80 ppm ozone gained less than filtered air controls. Lesions observed in both age groups of female rats were qualitatively similar to those previously described in young adult male rats. We conclude that there are age-related differences in the morphometric responses of rats to ozone exposure. Younger rats had larger proportions of centriacinar lesions and macrophages while older rats had greater body weight and lung volume changes.

Fine-needle aspiration of spleen: diagnosis of a solitary ovarian metastasis.

We report a patient in whom a solitary splenic lesion detected by computed tomography (CT) was successfully diagnosed by fine-needle aspiration biopsy under CT guidance. The ability to distinguish metastatic carcinoma from lymphoma in the spleen assisted the clinician in making therapeutic decisions. Fine-needle aspiration of spleen is a reasonable diagnostic approach that has been underutilized.

Effects of ozone on lung and somatic growth. Pair fed rats after ozone exposure and recovery periods.

Minor differences in lung growth and development during childhood have been considered as a potential cause of rapid decline in pulmonary function in adulthood. Inhalation of ozone commonly causes changes in both body weight and lung volumes, which complicates interpretation of any changes in lung growth. The effects of ozone on lung growth were studied in rats which were pair fed. This technique permitted comparison of ozone-exposed and filtered-air control rats of the same body weight and body size as well as age and sex. Exposure was to filtered air or to 0.64 or 0.96 ppm ozone (UV standard) 8 h/night for 42 nights. A second control group was fed ad libitum and exposed to only filtered air. Half the rats were studied at the end of the 42-night exposures, the rest after a 42-day post-exposure period during which all rats were fed ab libitum and breathed filtered air. Rats examined at the end of the exposure period had larger saline and fixed lung volumes. These larger lungs had greater volumes of parenchyma, alveoli and respiratory bronchioles. Some of these changes persisted throughout a 42-day post-exposure period. Ozone inhalation by young rats alters lung growth and development in ways likely to be detrimental and those changes persist after ozone inhalation stops.

Exercise-induced pulmonary haemorrhage in the horse: results of a detailed clinical, post mortem and imaging study. I. Clinical profile of horses.

Detailed physical and clinical examinations were performed on 26 Thoroughbred racehorses which were used subsequently in a series of studies to investigate the contribution of the pulmonary and bronchial arterial circulations to the pathophysiology of exercise-induced pulmonary haemorrhage (EIPH). Twenty-five of the horses had been retired from race training in Hong Kong during the 1984-85 season, all but four raced that season; one horse had been retired the previous season. The average number of races for the group that season was 4.1 +/- 2 with an average distance of 1502 +/- 216 metres, mean racing speed 15.5 +/- 0.5 metres/sec. Time from last race to necropsy was 177 +/- 155 days, range 12 to 572 days. All but one horse had a known history of either EIPH or epistaxis. Time from last recorded incident of expistaxis (17 horses) to necropsy was 156 +/- 141 days, range 12 to 513 days, with a longer interval since last recorded endoscopic observation of EIPH. Focal abnormal lung sounds were detected in the dorsocaudal lungfields on auscultation during rebreathing in three horses and six had tracheobronchial cytology consistent with previous episodes of pulmonary haemorrhage (haemosiderophages). No other characteristics which might have allowed separation of this group of horses from other Thoroughbred horses recently in race training were identified.