Bleeding in Racehorses
by Professor Ron Slocombe
January 2000 RIRDC

Executive Summary
Exercise-induced pulmonary haemorrhage (EIPH) is a widespread problem of horses involved in high-speed competitions, particularly Thoroughbred racing, pacing, steeple chase and three day events. It is now recognised that the prevalence of EIPH among Thoroughbred
racehorses approaches 100%. The severity of EIPH within this population varies widely from large volume haemorrhage into the lung which presents a life threatening condition, to minute haemorrhages which appear clinically silent. The affects these extremes of haemorrhage volume have on performance are quite distinct and readily discernible, however the affects of intermediate volumes of pulmonary haemorrhage are less well understood. Recently it has been shown that volumes as small as 200 ml of pulmonary haemorrhage can adversely affect oxygen uptake and exercise tolerance in galloping horses.
This report details the findings of a number of studies focused on determining the responses of the lung to the presence of erythrocytes in the airways. The critical responses studied were: the period of time required for removal of erythrocytes from the airway; the development of inflammation in response to intrapulmonary blood; the role of subclinical
pulmonary inflammation in predisposing the lung to EIPH; and the evaluation of possible techniques for quantifying the volume of pulmonary haemorrhage. The motivation for these studies came from a realisation that despite intensive study of EIPH for the past thirty years, the effects of EIPH and intrapulmonary blood on the lung were largely undetermined,
as were the processes of blood removal from the lung.
To study the processes of erythrocyte removal from the lung and the changes in pulmonary cytology that may result from the presence of blood in the airways, 40 ml volumes of blood were inoculated into several segmental bronchi of the cranioventral regions of the caudal lobe.
Each of these segments was then lavaged once at different times over a period of 21 days and changes in the erythrocyte and leucocyte numbers over this time span were noted.
The results indicate that autologous blood is removed more slowly from the lung than has been previously reported. Erythrocytes from a relatively small simulated episode of EIPH were found to persist and were recoverable in bronchoalveolar lavages obtained 21 days following inoculation. Initial removal of erythrocytes from the lung was dependent upon mechanical processes such as mucociliary clearance, coughing
and flow of material proximally along the trachea when the horse’s head is down for grazing. Mucociliary clearance of blood could not be detected endoscopically after 72 hours, and by this time it was estimated that more than half of the originally inoculated erythrocytes had been removed.
Macrophage erythrophagocytic activity had commenced by 72 hours following inoculation and gradually increased until day 10 when it
appeared to stabilise, although increased macrophage numbers
persisted for the remainder of the 21 day monitoring period and
presumably for some time after this. Haemosiderin was first observed within erythrophages at day 10 and was present in approximately 30%
of macrophages at day 21. This is consistent with previous reports
which indicate that the longevity of haemosiderin laden alveolar
macrophages may by as much as 3 months (Step et al 1991).
The prolonged course of erythrocyte removal has a number of important implications for the current management and racing schedules of Thoroughbred horses. The first of these implications being that horses which race frequently may not be permitted sufficient time to remove all the erythrocytes present from a previous episode of the EIPH before they are expected to race again. This could result in the progressive accumulation of blood in the peripheral lung, to a degree where a
number of successive low volume haemorrhages may result in a volume of intrapulmonary blood which is sufficient to impair racing performance.
Unfortunately at this time bronchoalveolar lavage provides only a semi-quantitative estimate of the severity of pulmonary haemorrhage within selected regions of the lung, however the results of the study
demonstrate that bronchoalveolar cytology may be used to provide a
general indication of the severity of haemorrhage within a region of the lung. Based on the findings of this study and on clinical experience it is suggested by the investigators that horses with lavage samples from regions of pulmonary haemorrhage in which greater than 40% of cells
are erythrocytes are in danger of developing problems associated with EIPH.
These problems include subsequent severe episodes of haemorrhage, loss of performance, and the development of interstitial pulmonary fibrosis.
The second important implication resulting from the prolonged course of erythrophagocytic activity is the persistent activation of macrophages. Macrophage activation is involved in both tissue destructive inflammatory processes and in the reparative processes of fibrosis and angiogenesis. Hence prolonged macrophage activation could potentially result in
damage to the delicate structures of the alveolus and subsequently to
the development of alveolar fibrosis. This alveolar fibrosis could be of
great detriment to the lung because of the potential for regional changes in pulmonary compliance and interdependence to predispose the lung to
EIPH (Robinson and Derksen 1980). If this is the case then the removal
of blood via erythrophagocytic mechanisms over a prolonged period is
not desirable, and methods to minimise this activity could form the basis for new treatments to reduce the increasing severity of EIPH seen in
older racehorses (Burke 1973; Cook 1974; Pascoe et al 1981; Raphel
and Soma 1982; Mason et al 1983; McKane et al 1993).
Prior to the development of significant macrophage activity in this study
a brief acute phase neutrophil response was observed 24 and 72 hours after blood inoculation, during which the percentage of neutrophils in the leucocyte population rose from the normal 3.5 + 0.6% to 10.5
+ 2.8%. This is representative of a moderate degree of acute pulmonary inflammation and may play an important role in the health of the lungs of horses involved in racing carnivals where they are required to race twice within a period of 48 - 72 hours. Over the years low grade pulmonary inflammatory disease has often been implicated in the development of EIPH (Cook 1974; O’Callaghan et al 1987; McKane et al 1993). In 1993 McKane et al noted a correlation between EIPH and increased numbers of neutrophils in bronchoalveolar lavages and following the same line of reasoning as the work by O’Callaghan et al (1987), concluded that bronchiolitis predisposed these horses to EIPH. In light of this new evidence it seems that this idea is in need of revision, as it appears that the presence of intrapulmonary blood itself is enough to evoke this inflammatory reaction and possibly also the fibrosis and angiogenesis observed by O’Callaghan et al (1987) in association with EIPH lesions.
To explore the issue of whether low grade inflammation could predispose to EIPH, a sterile lesion was produced that would elicit an acute inflammatory response of the same magnitude as that observed in response to intrapulmonary blood. To do this, 20 ml of dilute acetic acid
was inoculated into the lung, which caused a rise in the neutrophil percentage to approximately 12%, 24 hours after inoculation.
The horses were exercised 24 hours after inoculation and bronchoalveolar lavages obtained from the inoculated segment and corresponding uninoculated segment of the opposite lung. The neutrophil percentage in the acetic acid inoculated segments and the corresponding control segments were 12.3 + 1.1%
and 4.4 + 0.3% respectively. Comparisons revealed that in the control segments 0.29 + 0.29% of cells were erythrocytes, whereas in the inoculated segments 52.3 + 15.0% of cells were erythrocytes. This indicated a much greater propensity for the segments with acute low
grade pulmonary inflammation to haemorrhage during exercise. The results from these two studies indicate that it is possible that the moderate acute inflammatory reaction associated with episodes of EIPH is perhaps both a consequence of and a predisposing factor for further
EIPH.
Furthermore, contrasting the cytology from segments inoculated with known volumes of blood compared to that from typical field cases of
EIPH, there is indirect evidence to suggest that most subclinical
episodes of EIPH involve less than 40 ml of blood per bronchial
segment. This follows from the fact that lavages containing greater than 50% erythrocytes are rare, in survey samples
(McKane et al 1993, Meyer et al 1998). However it is a common
to find haemosiderophages comprising 30 - 40% of leucocytes in lavage fluid, suggesting that most horses with EIPH have multiple small
bleeding episodes leading to gradual haemosiderin accumulation where
40 - 50% of macrophages may eventually become laden with haemosiderin.
If prodromal bouts of EIPH leads only to a loss of a few millilitres of
blood into alveoli, it is not surprising that detection of these sites by scintigraphy is extremely difficult.
The scintigraphy studies were attractive because they potentially could overcome important limitations to all other studies estimating the
severity of haemorrhage with EIPH, namely that the use of routine lavage cytology or fluorochrome -tagged erythrocytes as markers was
totally reliant on adequate samples obtained by lavage, and this
technique would not be.
Haemorrhagic areas theoretically would be more radioactive than background, because blood cells are more densely packed in sites of haemorrhage than in normal lung.
At the Washington State University tri-state imaging facility we
conducted a series of pilot studies to determine the feasibility of a more comprehensive study. When 5 ml of technecium labelled erythrocytes
was deposited in the peripheral lung field of a horse, the
inoculation site was readily detected by thoracic scintigraphy.
For scintigraphy to be clinically useful, the tagged cells need to be injected IV and then located by scintigraphy if they accumulate in the lungs. We found that when 200 mCi of Technetium m99 linked to
either DTPA or Ultratag was used to radiolabel erythrocytes, if 25 ml of autologous radiolabelled blood were injected into the lung after
equilibrating in systemic blood for 30 minutes, the inoculation site was undetectable by scintigraphy. Additional studies where horses (n=8)
were instrumental and injected with labelled erythrocytes during vigorous
treadmill exercise, also failed to show scintographic changes on
thoracic scans, despite the presence of radiolabelled erythrocytes in
lavage fluids, proving that EIPH had occurred during treadmill exercise. Lavage/blood radioactivity ratios ranged from .0079 to .0001.
Substantial background radiation diffusely from the lung fields after exercise interfered with the ability to detect subtle changes in regional radioactivity. A control group (n=4) were treated identically except for treadmill exercise, and in these the BAL/blood ratios were higher than
the exercised group, a result that remains unexplained.
In conclusion, these studies were technically difficult to perform,
in part due to the intense sources of radiation needed, and sites of localised intrapulmonary haemorrhage were not detectible by
scintigraphy. It is our view that further studies using this methodology
are not justifiable.

For full research article click here