Vol. 86, Issue 6, 2034-2043, June 1999
Effect of furosemide on pulmonary blood flow distribution in
resting and exercising horses
Howard H.
Erickson1,
Susan L.
Bernard4,
Robb W.
Glenny4,5,
M. Roger
Fedde1,
Nayak L.
Polissar6,
Randall J.
Basaraba2,
Sten M.
Walther4,
Earl M.
Gaughan3,
Rose
McMurphy3, and
Michael P.
Hlastala4,5
1 Department of Anatomy and
Physiology, 2 Department of
Diagnostic Medicine/Pathobiology, and
3 Department of Clinical Sciences,
Kansas State University, Manhattan, Kansas 66506-5602;
4 Department of Medicine and
5 Department of Physiology and
Biophysics, University of Washington, Seattle, Washington 98195-6522;
and 6 The Mountain-Whisper-Light,
Statistical Consulting, Seattle, Washington 98112-2913
 |
ABSTRACT |
We determined
the spatial distribution of pulmonary blood flow (PBF) with 15-µm
fluorescent-labeled microspheres during rest and exercise in five
Thoroughbred horses before and 4 h after furosemide administration (0.5 mg/kg iv). The primary finding of this study was that PBF
redistribution occurred from rest to exercise, both with and without
furosemide. However, there was less blood flow to the dorsal portion of
the lung during exercise postfurosemide compared with prefurosemide.
Furosemide did alter the resting perfusion distribution by increasing
the flow to the ventral regions of the lung; however, that increase in
flow was abated with exercise. Other findings included
1) unchanged gas exchange and
cardiac output during rest and exercise after vs. before furosemide,
2) a decrease in pulmonary arterial
pressure after furosemide, 3) an
increase in the slope of the relationship of PBF vs. vertical height up
the lung during exercise, both with and without furosemide, and
4) a decrease in blood flow to the dorsal region of the lung at rest after furosemide. Pulmonary perfusion
variability within the lung may be a function of the anatomy of the
pulmonary vessels that results in a predominantly fixed spatial pattern
of flow distribution.
fluorescent microspheres; cardiac output; pulmonary gas exchange; exercise-induced pulmonary hemorrhage
| |
INTRODUCTION |
DURING MAXIMAL EXERCISE, horses frequently experience
exercise-induced pulmonary hemorrhage (EIPH) (5, 25). Recent findings have shown that high vascular pressures may induce stress-related failure of the pulmonary capillaries, thus causing hemorrhage and edema
in the gas-exchange region of the lung (3, 6, 36, 37). This hemorrhage
has been especially prominent in the dorsal-caudal region of the lungs
(22).
Recent studies indicate that distribution of pulmonary blood flow (PBF)
in standing horses is not dominated by gravity (17). More than 70% of
variation in PBF during rest and varying levels of exercise are
determined by a fixed spatial pattern that is most likely caused by the
structure of pulmonary vessels (2). The greatest redistribution occurs
with minimal exercise (trot), and little change occurs subsequently at
higher levels of exercise. Blood flow increases primarily in the
dorsal-caudal region of the lung during exercise (2). Despite this
redistribution of PBF, heterogeneity of perfusion is not altered by exercise.
Horses with EIPH frequently are treated with furosemide (13, 14, 26,
31-33), which attenuates the exercise-induced increases in right
atrial, pulmonary arterial (Ppa), pulmonary wedge, and pulmonary
capillary pressures (20, 23). The mechanisms responsible for the
effects of furosemide and reduced vascular pressures during exercise
may be associated with a redistribution of PBF. Pelletier et al. (27)
have shown, with in vitro studies, that equine pulmonary arteries in
the dorsal portion of the lung dilate more in response to methacholine
than do vessels located in the ventral portion of the lung. If
furosemide acts preferentially on the vessels located in the dorsal
part of the lung, this might explain why EIPH occurs in the
dorsal-caudal region of the lung. This led us to investigate perfusion
distribution in five horses at rest and at near-maximal exercise,
before and after administration of furosemide.
| |
METHODS |
Animals and animal preparation.
All methods were approved by the University of Washington (Seattle) and
Kansas State University (Manhattan) Animal Care Committees and were in
accordance with regulations of the Association for Assessment and
Accreditation of Laboratory Animal Care.
Five Thoroughbred horses (422-500 kg) were trained for 8-10
wk on a high-speed equine treadmill (Sato, Uppsala, Sweden). Each horse
was trained twice a week on a level treadmill, beginning with a warm-up
at 3-4 m/s for 800 m. After the warm-up, the horses ran at 9 m/s
for 2 min and at 12-13 m/s for 2 min. This was followed by a
cool-down period for 400 m. One week before the study, their maximal O2 consumption
(
O2 max) was
determined. O2 max
was elicited at 11.9-13.9 m/s on a 3° incline. Before the day
of the study, lateral thoracic radiographs were taken of each horse
with a radiopaque rod placed horizontally along the chest as a
reference marker for the later determination of isogravitational planes of the lungs. Horse preparations and measurements were similar to those
used in prior studies (2, 17). On the day of the experiment, catheters
were placed in the left transverse facial artery to withdraw arterial
blood samples and in the left jugular vein for microsphere injection. A
solid-state, catheter-tipped, pressure transducer (model SPC 471A,
Millar Instruments, Houston, TX) and a thermistor catheter (110-cm
thermal-dilution catheter, Columbus Instruments, Columbus, OH) were
advanced into the pulmonary artery via the right jugular vein to
measure Ppa, to obtain mixed venous blood samples, and to measure core
temperatures for correction of blood gases. Electrocardiograph
electrodes for heart rate determination and a safety harness were
placed on each horse.
Experimental protocol.
PBF distribution and physiological variables were measured at six
different time points: 1) at rest
when the horses were in the barn (PBF distribution only);
2) while the horses were standing on
the treadmill before exercise; 3)
during near maximal exercise (11.5-13.5 m/s at 3°);
4) at rest 60 min after exercise;
5) at rest 4 h after furosemide (0.5 mg/kg) administration; and 6) during near maximal exercise at the same speed and incline after furosemide was administered. For exercise measurements, the treadmill speed was
set to produce an O2 consumption
(O2) of ~90% of each
horse's predetermined
O2 max. The average
exercise duration at near-maximal exercise, both before and after
furosemide, was >3 min. The horses were allowed to rehydrate by
drinking water at will between the first exercise period and the second
rest period. Thus a sufficient time was allowed for the horses to
correct for any water loss during the brief exercise period. After
furosemide was given, food and water were withheld until after
completion of the study.
Physiological measurements.
O2 and
CO2 production
(CO2) were measured by
using a bias-flow technique (34).
O2 contents from arterial and
mixed venous blood samples were determined with a
Lex-O2-Con (model TL, Hospex, Chestnut Hill, MA). Arterial partial pressures of
O2 and
CO2
(PaO2 and
PaCO2, respectively), mixed venous
partial pressures of O2 and
CO2, and pH were obtained by using
a Nova blood-gas analyzer (Nova Stat Profile 4, Nova Biomedical,
Waltham, MA) that was calibrated with standard gases and
buffers. The values were corrected to the horse's
pulmonary arterial temperature by using temperature-correction factors
for horse blood (9). Cardiac output was calculated by using the Fick
equation. Hematocrit was determined by the standard microhematocrit
method, and lactate concentration was measured in mixed venous blood
(model 23L lactate analyzer, calibrated with factory standards; Yellow
Springs Instruments, Yellow Springs, OH). The electrocardiogram was
recorded from surface electrodes to obtain heart rate. Mean Ppa was
determined with the Millar pressure transducer, which was advanced ~8
cm past the pulmonic valve. The transducer was calibrated before and
immediately after each run with a mercury manometer. No drift was
detected after any of the runs. Detailed methods for obtaining the
physiological measurements have been reported previously (17).
Measurements of relative PBF distribution.
Fluorescent 15-µm-diameter polystyrene microspheres (Molecular
Probes, Eugene, OR) with seven different labels (blue, blue-green, yellow-green, orange, red, crimson, or scarlet) were used to determine relative PBF distribution. A different microsphere color was injected at each time point in randomized order. At rest and exercise, when the
animal had reached steady state (after 2 min of exercise), 40-65
million microspheres were injected into the jugular catheter over ~15 s.
Lung processing.
After completion of the experiment, the horses were walked to the
necropsy laboratory, tranquilized with xylazine (1 mg/kg iv),
heparinized (~200,000 IU/horse iv), and injected with papaverine (3 mg/kg iv) to induce vasodilatation. They were then anesthetized with
xylazine (1 mg/kg), ketamine (2.5 mg/kg), and pentobarbital sodium (16 mg/kg), tracheotomized, and exsanguinated by opening the carotid
arteries. A deep surgical plane of anesthesia was maintained at all times.
When the horse was dead, several right ribs were removed and a large
cannula was inserted through the right ventricle into the pulmonary
artery. A second cannula was placed in the left atrium, and a 0.9%
NaCl solution (~300 mosM, pH 7.4) was flushed through
the pulmonary vessels at 40-50
cmH2O pressure. The lungs were
repeatedly inflated (up to 30 cmH2O) and deflated. A 200-liter reservoir of saline was elevated to maintain a constant perfusion pressure. When the effluent from the lungs was clear, they were removed
from the chest, suspended from the tracheal cannula, inflated with air
at 35-40 cmH2O pressure, and
dried for 14 days or longer. To aid in drying, the lungs were pierced
with a 17-gauge needle in several locations, and warm, dry air was
passed into the trachea, through the bronchi, and out to the atmosphere
through the holes.
With the aid of the lateral thoracic radiographs, the dried lungs were
placed in a plywood box (110 × 80 × 80 cm) in the same orientation as in the horse's thorax and surrounded with urethane foam
to provide a rigid orthogonal reference system. The lungs and foam were
sliced into 2.2-cm-thick coronal sections at isogravitational planes by
using a band saw with a custom stage. Cores of lung tissue were
obtained in a rigid
x-y
grid system, with 3.3 cm between each
x- and
y-point, by using an
8.5-mm-diameter core (core volume = 1.3 cm3). Samples of partial volume
or those containing >25% airway were discarded. Between 400 and
1,087 samples were obtained from each horse, depending on the animal's
size. Orthogonal spatial coordinates (x,
y, and
z) were assigned to the center of
each piece. The x-dimension describes
the lateral position, y describes the
vertical position in the gravitational plane (ventral to dorsal), and
z describes the caudal to cranial
position (Fig. 1). The value of zero
(origin) for each dimension is the leftmost point of the lung for the
x-dimension, the lowest ventral point
for the y-dimension, and the most
cranial point for the z-dimension. We
excluded poorly dried areas of lungs in four of the five horses. Lungs
from horses 1, 2, 3, 4, and 5 had 3, 0, 4, 6, and 20% losses,
respectively, from inadequate drying. These areas were small, patchy,
and randomly scattered in the lung parenchyma. This loss accounts for
some of the variation in number of samples obtained from the different
horses.
Each sample was soaked for 5 days in 1.5 ml Cellosolve acetate, and the
fluorescence of each sample was measured with a luminescence spectrophotometer (model LS-50B fluorimeter; Perkin-Elmer, Norwalk, CT)
equipped with a standard cuvette reader. All fluorescent samples were
measured with excitation- and emission-slit widths of 4 nm and an
emission filter that blocked all light <350 nm (10).
Data analysis and statistics.
Volume-normalized relative PBF (RPBF) to each piece of lung was
calculated by dividing the fluorescence in each lung piece by the mean
fluorescence of all pieces and multiplying by 100. This yielded a mean
relative flow of 100% for each horse during each treatment. The
treatments for each horse and abbreviations for RPBF during each
treatment are shown in Table 1. The RPBF in
the barn (RPBFB) was determined
from (RPBFB1 + RPBFB2)/2, and RPBF at rest
before furosemide injection
(RPBFrest) was determined as
(RPBFB + RPBFTM)/2.
The means ± SD for the physiological measurements, slopes of RPBF,
and centers of flow for the five horses were calculated. The Pearson
correlation coefficient was determined between pairs of treatments. The
paired observations that showed the flow to a piece in each of two
treatments summarized the similarity of flows between the treatments
among all pieces. Pulmonary perfusion heterogeneity was calculated as
the coefficient of variation (CV) of RPBF (100 × SD/mean). The
gradients of flow in each orthogonal direction were calculated for each
horse by using simple linear regression with least squares fitting of
flow vs. distance. The gradient dimensions are expressed as percentage
of mean relative blood flow per centimeter. Thus, for example, a value
for the slope of 
4.0 indicates that the flow decreased by 4% of
the average flow per piece of the lung during that treatment for each
additional centimeter. The equine lung does not sit squarely in the
thorax. A relationship exists between
y- and
z-coordinates, as the more caudal portion of the lung lies in the more dorsal planes. To investigate whether correlation between the
y- and
z-dimensions may have affected the
slopes of flow in these dimensions, we carried out a separate analysis
of slopes in the y-dimension,
controlling for z, and in the
z-dimension, controlling for
y. In the
y-dimension, for example, we
calculated the slope of flow vs. y for
each z-plane and then calculated the
weighted mean of these slopes across all z-planes, weighted for the number of
pieces in each plane. This yielded an adjusted slope of flow vs.
y (designated as
y'). A similar adjusted
z-slope was calculated, controlled for
y.
We used ANOVA to partition the variation in flow within a horse into
separate components. When two treatments are compared (e.g., rest and
exercise), the flow variation between treatments can arise from four
components: 1) flow to the piece
that is constant across the two treatments, the "piece effect";
2) a component that changes with the
treatments, e.g., exercise vs. rest;
3) a random component that varies
over time within each treatment; and
4) a random component for each
observation caused by methodological "noise" (e.g., Poisson
distribution of microspheres, fluorescent measurement error). The
methodology has been described in detail previously (10). Mean values
of slopes and other variables were compared by using the paired,
two-tailed t-test. Statistically significant changes or correlations reported are at
P 
0.05.
We calculated flow-weighted spatial coordinates for the center of flow
of the three orthogonal distances for each treatment. The center of
flow was calculated as the statistically weighted mean of
x, y,
and z (separately), where the
statistical weight was the normalized flow to each piece in the
specific treatment. This yielded the center of PBF for each treatment
(see Tables 3-7). The center of flow is a concept
that identified the balance point for blood flow. The value of the
center of flow alone is not particularly informative. It is the change
in the center of flow, both in magnitude and direction, that is of
interest. The shift in center of flow between any two treatments was
calculated as the Euclidean distance between the center of flow. The
shift in the center of flow is the key statistic here. The center of flow may vary considerably among horses due to size and shape of the
lung. However, the shift in the center between treatments is likely to
be uniform if there is a consistent shift in flow between treatments.
| |
RESULTS |
Physiological responses to exercise.
One horse would not accept the face mask on the day of the experiment,
and the arterial catheter was lost after the first resting measurement.
The remaining catheters and pressure transducers were in place, and
this horse completed the entire protocol with no additional problems.
Although most variables were measured, many physiological measurements
are missing for this animal, as indicated in Table
2 in the column for the number of
observations used to calculate the mean and SD.
Mean heart rate increased from 38 beats/min at rest to 209 beats/min
during exercise, an indication of near maximal effort. The
O2 increased from 5.2 ml · min
1 · kg1
at rest to 140.5 ml · min1 · kg1
during exercise. Cardiac output increased from 46.1 l/min (97.4 ml · min1 · kg1)
at rest to 292.4 l/min (616.9 ml · min1 · kg1)
during exercise. Mean Ppa increased from 27.7 mmHg at rest to 83.2 mmHg
during exercise before furosemide injection. After administration of
furosemide, Ppa decreased from 27.7 ± 2.4 to 24.1 ± 1.5 mmHg at
rest, which was not significant, and from 83.2 ± 11.5 to
77.6 ± 13.4 mmHg during exercise
(P < 0.05). Furosemide caused a
small decrease in cardiac output at rest. Body weight decreased 20 ± 4 kg after furosemide and exercise
(P < 0.001).
PBF distribution: resting in the barn vs. resting on the treadmill.
No statistically significant difference existed in the spatial
distribution of flow when the horses were in the barn or resting on the
treadmill (Fig. 2 and Table
3). The difference in slope between the two
locations in any of the dimensions (x, y,
z) was <3%/cm. Similarly, the difference in CV of
the blood flow was negligible. However, we noted nonspatial evidence
that the flow distribution did change between these two resting states.
When blood flow to each piece was compared between rest in the barn and
rest on the treadmill within each horse using linear regression (which
removes spatial components in the comparison), the mean slope and
intercept from the regression analysis were significantly different
from 1 and 0, respectively. All horses had slopes <1.0 (mean ± SD = 0.71 ± 0.16, P < 0.05); this
indicates an increase in flow to the low-flow regions and a decrease in
flow to the high-flow regions from rest in the barn to rest on the
treadmill (Fig. 3). A slope of <1.0
can be expected due to random variation in flow over time within a
treatment state and due to microsphere noise. However, the observed
mean slope of 0.7 is unusually low, on the basis of previous studies of
time variation and microsphere noise. The low slope suggests
considerable redistribution of flow between the two resting treatments.
A slope of <1.0 suggests an increase in flow to low-flow regions and
a decrease in flow to high-flow regions.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 2.
Distribution of relative pulmonary blood flow (RPBF) in 1 horse when
standing in the barn (RPBFB) and
when standing at rest on the treadmill
(RPBFTM). Ventral-to-dorsal
distribution in y-coordinate is
shown.
|
|
PBF distribution: resting on the treadmill vs. exercise and recovery
(no furosemide).
The slope of flow vs. vertical height up the lung
(y-axis) increased significantly during
exercise, with more flow going to the dorsal regions of the lung (Fig.
4 and Table 4).
The z-adjusted y-slope
(y') also increased. Similarly, the
center of flow increased significantly in the dorsal direction with
exercise (0.7 ± 0.3 cm). Heterogeneity of blood flow in the lungs
was not altered by exercise.
View larger version (13K):
[in this window]
[in a new window]
|
Fig. 4.
Distribution of RPBF in the ventral-to-dorsal
y-coordinate in 1 horse resting on the
treadmill (RPBFTM) and while
running (RPBFE) at near-maximal
O2 consumption.
|
|
PBF distribution: resting on the treadmill before vs. after
furosemide.
The slope of flow vs. the vertical height up the lung
(y-axis) decreased significantly after
furosemide administration (Table 5); this
indicates a decrease in blood flow to the dorsal regions of the lung.
Similarly, the slope of flow vs.
y' also decreased. The center of
flow shifted significantly both ventrally (1.1 ± 0.5 cm) and
cranially (0.8 ± 0.5 cm). Heterogeneity increased 8.4% after
furosemide, but the change was not significant.
PBF distribution: resting on the treadmill vs. exercise after
furosemide.
The largest change in slope of flow vs. the vertical height up the lung
occurred with exercise after furosemide administration (Table
6, Fig. 5). The
slope of both y and
y' increased dramatically and
significantly with strenuous exercise. Similarly, the center of flow
shifted dorsally (y-dimension) by 1.7 ± 0.5 cm with exercise (Table 6). Heterogeneity decreased by 11.2% with
exercise after furosemide (P = 0.01)
(Table 6, Fig. 6), largely because of
abatement of the 8.5% increase in heterogeneity that occurred at rest
after furosemide administration.
View larger version (21K):
[in this window]
[in a new window]
|
Fig. 5.
Impact of exercise and furosemide on slopes of RPBF vs. vertical height
up lung in the 5 horses studied. Each symbol and line represent 1 horse.
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig. 6.
Impact of exercise and furosemide on the coefficient of variation
(%CV) of blood flow in the 5 horses studied. Each symbol and line
represent 1 horse.
|
|
PBF distribution: exercise before vs. after furosemide.
The values for slopes, center of flow, and CV did not differ
significantly when flow distribution before furosemide was compared with flow distribution after furosemide during exercise (Table 7, Fig. 5). However, there was a decrease
in the slope of flow vs. y'
during exercise after furosemide (Table 7).
Correlations between treatments.
The correlations of RPBF between similar treatments (rest before vs.
rest after furosemide, and exercise before vs. exercise after
furosemide) were the strongest. The correlation between exercise
treatments (exercise with and without furosemide, Fig. 7) was r = 0.86 ± 0.05 (means ± SD). The correlation between rest with and
without furosemide was lower (r = 0.68 ± 0.07). The correlations between rest and exercise were the
weakest, both with and without furosemide (rest vs. exercise before
furosemide, r = 0.59 ± 0.10; rest
vs. exercise after furosemide, r = 0.49 ± 0.15; Fig. 8).
View larger version (15K):
[in this window]
[in a new window]
|
Fig. 7.
Correlation of RPBF in each lung sample from 1 horse during exercise
before furosemide (RPBFE) to
RPBF in that sample during exercise after furosemide
(RPBFEF).
|
|
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 8.
Correlation of RPBF in each lung sample from 1 horse during rest on
treadmill before furosemide
(RPBFTM) and during exercise
before furosemide (RPBFE).
|
|
PBF variance.
Stability of flow across all treatments measured (rest, exercise, and
furosemide) was demonstrated when the variation in flow was partitioned
into a component that was constant to each piece between treatments and
a component that changed (2). Thus 83% of the variation in measured
flow was constant between rest before furosemide and exercise after
furosemide, and the balance (17%) was attributed to altered flow
between treatments that was caused by changes from rest to exercise,
changes over time, and methodological noise (expected to be a very
small component). Similarly, 82% of the flow variation was constant
between rest and exercise after furosemide administration. The
similarity of the two exercise treatments (before and after furosemide)
was indicated by 93% of the flow variation being due to an unchanging
component and only 7% due to change. Finally, when the two resting
treatments (treadmill before and after furosemide) were compared, 87%
of the flow variation remained constant between treatments and 13% changed because of furosemide, time (which includes an intervening exercise state with recovery), and methodology. These percentages are
similar to those reported previously (2).
| |
DISCUSSION |
The primary finding of this study was that furosemide did not prevent
redistribution of PBF to the dorsal-caudal region of the lung during
exercise. Other findings included 1)
a large decrease (20 kg) in body weight after exercise and furosemide,
2) unchanged gas exchange and
cardiac output during rest and exercise after furosemide compared with
rest and exercise before furosemide, 3) a decrease in Ppa after
furosemide, 4) an increase in the
slope of the relationship of PBF vs. vertical height up the lung during exercise (both with and without furosemide) compared with rest, and
5) a decrease in blood flow to the
dorsal region of the lung after furosemide (both rest and exercise).
The primary source of the pulmonary perfusion variability within the
lung appears to be a fixed spatial pattern of flow distribution. This
pattern apparently is a function of the anatomy of the pulmonary vessels.
Critique of methods.
The fluorescent microsphere method for measurement of regional organ
perfusion has been validated (10). Techniques for measuring PBF
distribution in standing (17) and exercising racehorses (2) also have
been reviewed. In this study, we increased the distance between each
lung sample from 2.3 cm, as used in our previous study (2, 17), to 3.3 cm to reduce cost and labor; however, the size of the sample (8.5-mm
diameter) was the same in both studies. This yielded 400-1,087
samples for each pair of lungs, compared with 1,621-2,503 samples
in the previous study. Subsequent to our first study, we learned that
decreasing the number of samples did not alter the general descriptors
of the lung blood flow distribution, such as slopes or correlations, when the lung was sampled randomly, as it was in this study.
PBF was measured first at rest in the barn, with additional
measurements obtained at rest 4 h later on the treadmill; this allowed
assessment of temporal changes in pulmonary perfusion. No significant
differences in the mean slopes of RPBF vs. spatial location occurred
between measurements taken at the barn and on the treadmill (Table 3).
However, when the blood flow to each lung piece at rest in the barn was
plotted against blood flow to the same piece at rest on the treadmill,
for each horse, the slopes were substantially <1.0, and intercepts
were not equal to 0 (Fig. 3). A slope substantially <1.0 and beyond
the decrease expected by chance indicates that lung pieces that
received high flow in the barn lost some flow when the horses were on
the treadmill, and pieces that received low flow in the barn gained a
small amount of flow when the horses were placed on the treadmill. The
fact that the slopes of relative blood flow as a function of distance along any of the three axes examined were not different suggests that
redistribution may be flow related, regardless of spatial location.
It was important to investigate the dependence of
y on
z because of the correlation between
y and
z within an equine lung (i.e., the
most caudal portion of the lung lies in the more dorsal planes). There
were no differences in the y-adjusted
z-analysis. For each experimental
condition, the magnitude of the
z-adjusted y-slopes was larger than that of the
unadjusted slopes; this made the changes in adjusted slopes between
conditions larger. There was a statistically significant difference in
the slope of flow vs. y' with
exercise, before and after furosemide. This change was small, but
different from the nonadjusted
y-analysis (see Effects of furosemide).
In prior studies (2), we reported the change in slope of PBF vs.
distance from the hilum. There were no significant changes in the
slopes of flow vs. distance from the hilum with any exercise state or
treatment; therefore, these data were not included in the manuscript.
Physiological measurements during exercise.
Results from this study compare well with those from previous studies
of Thoroughbred racehorses in which heart rate, cardiac output, and Ppa
were measured at rest and during exercise (2, 7, 8, 19, 21, 23,
28-30, 34). During exercise, average values of heart rate
increased to 209 beats/min, cardiac output to 292 l/min (617 ml · min1 · kg1),
O2 to 140 ml · min1 · kg1,
and mean Ppa to 83 mmHg before furosemide was given; these values indicate that near-maximal exercise levels were achieved.
Effects of furosemide.
Furosemide, frequently administered to racehorses 4 h before a race,
has a putative prophylactic effect on EIPH (14, 15, 23). Furosemide is
a diuretic that reduces blood and plasma volume in the racehorse (14,
15, 18). It also induces a reduction in right atrial pressure and Ppa
(20, 23) during high-speed sprint exercise. Decreases in these
pressures may be caused by diminished intravascular volume. Using
anesthetized horses, Hinchcliff et al. (14) have shown recently that
diminished intravascular volume may be the cause of the reduction in
pulmonary artery pressure. If the pulmonary capillary threshold
pressure is not exceeded, decreased pressure will decrease pulmonary
hemorrhage during sprint exercise.
Controversy exists in the recent literature about whether furosemide
decreases the incidence and/or severity of EIPH (13). Although
furosemide appears to improve racing performance (33), this result may
not be related to the diminution of pulmonary hemorrhage. The large
reduction in body weight may explain the improvement in racing
performance after furosemide. Hinchcliff et al. (16) showed that
furosemide caused a large decrease in body weight and small decreases
in CO2, plasma
lactate, and the respiratory exchange ratio, but cardiac output and
O2 were unaffected. When
horses receiving furosemide were brought back to their prefurosemide
weight, blood lactate and CO2
did not differ from the nonfurosemide control group.
Similar to observations by Hinchcliff et al. (16), our horses lost
~4% of their body weight after furosemide administration; cardiac
output and O2 remained
unchanged. At near-maximal exercise after furosemide, RPBF distribution
did not change from the prefurosemide levels; this suggests that a
horse's improved performance or decreased severity of EIPH is not
caused by blood flow redistribution.
PBF distribution.
Previous work (1, 11, 12) has shown no consistent vertical gradient to
PBF in the dog lung and a considerable degree of perfusion
heterogeneity within a given isogravitational plane. Thus gravity alone
does not appear to be a major determinant of blood flow distribution.
Results from our present study are consistent with this interpretation.
However, after furosemide administration, a significant decrease in the
slope of flow vs. vertical height up the lung
(y-axis) at rest indicated a decrease in blood
flow to the dorsal regions of the lung. A possible explanation for this
redistribution could be a decrease in mean Ppa at rest after furosemide. However, our Ppa data are inconsistent with this
interpretation. All horses at rest showed a significant decrease in
slope of PBF vs. vertical height up the lung after administration of
furosemide, yet not all horses had a decrease in mean Ppa. Furthermore,
the horse in our study with the largest apparent gravity-dependent change in slope of flow vs. vertical height up the lung from rest before to rest after furosemide had essentially no change in resting mean Ppa after furosemide administration (25.5 mmHg after vs. 24.5 mmHg before).
A second explanation for this redistribution would be preferential
vasorelaxation of pulmonary arteries in the ventral lung because of
regional differences in vascular reactivity. Pelletier et al. (27) have
shown in vitro that equine pulmonary arteries located in the ventral
portion of the lung dilate less in response to methacholine than do
vessels located in the more dorsal portion of the lung. If furosemide
causes ventral vessels to dilate preferentially at rest, pulmonary
perfusion would redistribute vertically as observed in our horses.
We calculated the center of flow to the lung to better understand the
magnitude of the shift in blood flow from rest to exercise. Overall,
changes in the center of flow were small compared with overall
dimensions of the lung, yet they were statistically highly significant;
this confirms that directional changes in blood redistribution are
similar to changes in the slope of flow.
Because of the large change in slope of flow vs. vertical height up the
lung (y-axis) at rest after furosemide
administration, the center of flow shifted 1.7 cm dorsally during
exercise after furosemide (compared with a 0.7-cm dorsal shift during
exercise before furosemide). The center of flow was not different
between exercise before and after furosemide. However, there was a
difference in the slope of flow vs.
y' during exercise before and
after furosemide. This suggests less blood flow to the dorsal portion
of the lung during exercise postfurosemide when compared with
prefurosemide. This change was small, and the effect of furosemide
cannot be separated from a second run effect. However, if this change
were solely due to furosemide, it would fit well with the hypothesis that furosemide alters regional vascular reactivity. The change in
slopes of flow vs. both y and
y' from rest to exercise (both pre- and postfurosemide) was significant; therefore, furosemide did not
prevent redistribution but may have attenuated it.
In a previous study, we determined the spatial distribution of PBF
during increased levels of exercise (2). Of the blood flow variation
during rest and high-exercise states, >70% could be accounted for by
a fixed spatial pattern, most likely determined by the geometry of the
pulmonary vascular tree. However, ~30% of the variation in PBF
during rest and exercise was due to redistribution, the majority
stemming from an increased flow to the dorsal region of the lung during
exercise. This is the region where most lesions of EIPH occur. The
percent variation in flow during rest and exercise before and after the
administration of furosemide in the present study compares well with
data from our previous study (2, 17) and supports the hypothesis that
geometry of the pulmonary vascular tree is the primary determinant of
pulmonary perfusion distribution.
Flow heterogeneity was unaltered during exercise both before and after
furosemide. It also did not change from rest to exercise before
furosemide, but it decreased by 11% from rest to exercise after
furosemide. This might be attributed to a marked increase in resting
heterogeneity after furosemide administration, an effect that was
abated with exercise.
In our study, mean Ppa during exercise decreased slightly after
furosemide from 83 to 78 mmHg. This small change may not alter pulmonary perfusion distribution in a maximally recruited lung. However, it may indicate a significant reduction in capillary pressure
and, therefore, a reduction in capillary rupture. Recent studies that
used bronchoalveolar lavage (19) have shown that, when Ppa in
exercising horses exceeds 90 mmHg, a significantly increased number of
red blood cells are present in the bronchoalveolar lavage. It is widely
accepted that most, if not all, horses bleed to some degree during
strenuous exercise (5). Pascoe (25) recently speculated that physical
disruption of alveolar tissues and/or large increases in alveolar
pressures "due to nonsynchronous ventilation of a region of lung
wherein vascular transmural pressure exceeded the strength of the
vessel wall" may be responsible for the hemorrhage. Horses that are
known to be reproducible bleeders may have weakened areas of lung
tissue that make them prone to repeated hemorrhage. A "critical
capillary-breaking pressure" also may exist that causes vascular
transmural pressure to exceed the strength of the vessel wall. A direct
measurement of regional differences in vascular pressures would greatly
improve our ability to determine the capillary pressures at which
horses bleed. If rupture of the alveolar capillaries is caused solely
by high Ppa, increased hemorrhage in the ventral aspects of the lungs
might be predicted because the hydrostatic gradient increases
intravascular pressures. However, this is not where the bleeding
usually occurs.
Summary.
Comparisons before and after furosemide administration showed that it
did not prevent the redistribution of PBF to the dorsal-caudal region
of the lung during exercise. Furosemide did alter the perfusion distribution by increasing the flow to the ventral regions of the lung.
| |
ACKNOWLEDGEMENTS |
We are indebted to Pam Davis, Tammi Meyer, LeAnn Rall, Elizabeth
Raub, Eddie Weigle, and Dr. Ingrid Langsetmo in Manhattan, KS; to Dr.
Nicolas Pelletier in East Lansing, MI; and to Dowon An, Myron Chornuk,
and Pam Campbell in Seattle, WA, for excellent technical assistance.
| |
FOOTNOTES |
This research was supported, in part, by National Heart, Lung, and
Blood Institute Grants HL-12174, HL-24163, and HL-02625 and by Grants
from the American Quarter Horse Association and the Kansas Racing
Commission (contribution no. 98-324-J from the Kansas Agricultural
Experiment Station).
Present address of R. J. Basaraba: Dept. of Pathology, Colorado State
Univ., Fort Collins, CO 80523-1601.
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests and other correspondence: H. H. Erickson,
Dept. of Anatomy and Physiology, Kansas State Univ., Manhattan, KS
66502-5602 (E-mail: erickson{at}vet.ksu.edu).
Received 27 March 1998; accepted in final form 22 February 1999.
| |
REFERENCES |
1.
Beck, K. C.,
and
K. Rehder.
Differences in regional vascular conductances in isolated dog lungs.
J. Appl. Physiol.
61:
530-538,
1996.
2.
Bernard, S. L.,
R. W. Glenny,
H. H. Erickson,
M. R. Fedde,
N. Polissar,
R. J. Basaraba,
and
M. P. Hlastala.
Minimal redistribution of pulmonary blood flow with exercise in racehorses.
J. Appl. Physiol.
81:
1062-1070,
1996[Abstract/Free Full Text].
3.
Birks, E. K.,
O. Mathieu-Costello,
Z. Fu,
W. S. Tyler,
and
J. B. West.
Very high pressures are required to cause stress failure of pulmonary capillaries in Thoroughbred racehorses.
J. Appl. Physiol.
82:
1584-1592,
1997[Abstract/Free Full Text].
4.
Caruthers, S. D.,
and
T. R. Harris.
Effects of pulmonary blood flow on the fractal nature of flow heterogeneity in sheep lungs.
J. Appl. Physiol.
77:
1474-1479,
1994[Abstract/Free Full Text].
5.
Erickson, H. H.
Exercise-induced pulmonary hemorrhage.
Equine Vet. J., Suppl.
18:
476-478,
1994.
6.
Erickson, H. H.,
J. L. McAvoy,
and
J. A. Westfall.
Exercise-induced changes in the lung of Shetland ponies: ultrastructure and morphometry.
J. Submicrosc. Cytol. Pathol.
29:
1-6,
1997[Medline].
7.
Evans, D. L.,
and
R. J. Rose.
Cardiovascular and respiratory responses to exercise in Thoroughbred horses.
J. Exp. Biol.
134:
397-408,
1988[Abstract/Free Full Text].
8.
Evans, D. L.,
and
R. J. Rose.
Cardiovascular and respiratory responses to submaximal exercise training in the Thoroughbred horse.
Pflügers Arch.
411:
316-321,
1988[Medline].
9.
Fedde, M. R.
Blood gas analyses on equine blood: required correction factors.
Equine Vet. J.
23:
410-412,
1991[Medline].
10.
Glenny, R. W.,
S. Bernard,
and
M. Brinkley.
Validation of fluorescent-labeled microspheres for measurement of regional organ perfusion.
J. Appl. Physiol.
74:
2585-2597,
1993[Abstract/Free Full Text].
11.
Glenny, R. W.,
W. J. Lamm,
R. K. Albert,
and
H. T. Robertson.
Gravity is a minor determinant of pulmonary blood flow distribution.
J. Appl. Physiol.
71:
620-629,
1991[Abstract/Free Full Text].
12.
Hakim, T. S.,
G. W. Dean,
and
R. Lisbona.
Quantification of spatial blood flow distribution in isolated canine lung.
Invest. Radiol.
23:
498-504,
1988[Medline].
13.
Hinchliff, K. W.,
and
K. H. McKeever.
Frusemide.
Equine Vet. J. Suppl.
18:
256-258,
1995.
14.
Hinchcliff, K. W.,
J. A. E. Hubbell,
D. Grosenbough,
L. A. Mitten,
and
W. L. Beard.
Hemodynamic effects of furosemide are dependent on diuresis.
Am. Assoc. Equine Pract.
42:
229-230,
1996.
15.
Hinchcliff, K. W.,
K. H. McKeever,
and
W. W. Muir.
Furosemide-induced changes in plasma and blood volume in horses.
J. Vet. Pharmacol. Ther.
14:
411-417,
1991[Medline].
16.
Hinchcliff, K. W.,
K. H. McKeever,
W. W. Muir,
and
R. A. Sams.
Effect of furosemide and weight carriage on energetic responses of horses to incremental exertion.
Am. J. Vet. Res.
54:
1500-1504,
1993[Medline].
17.
Hlastala, M. P.,
S. L. Bernard,
H. H. Erickson,
M. R. Fedde,
E. M. Gaughan,
R. McMurphy,
M. J. Emery,
N. Polissar,
and
R. W. Glenny.
Pulmonary blood flow distribution in standing horses is not dominated by gravity.
J. Appl. Physiol.
81:
1051-1061,
1996[Abstract/Free Full Text].
18.
Hopper, M. K.,
R. L. Pieschl,
N. G. Pelletier,
and
H. H. Erickson.
Cardiopulmonary effects of acute blood volume alteration prior to exercise.
Equine Exer. Physiol.
3:
9-16,
1991.
19.
Langsetmo, I.,
T. S. Meyer,
M. R. Fedde,
and
H. H. Erickson.
Relationship between pulmonary vascular pressure and pulmonary hemorrhage in horses: is there a threshold value?
Am. Assoc. Eq. Pract.
42:
227-228,
1996.
20.
Manohar, M.,
E. Hutchens,
and
E. Coney.
Frusemide attenuates the exercise induced rise in pulmonary capillary blood pressure in horses.
Equine Vet. J.
26:
51-54,
1994[Medline].
21.
Manohar, M.,
E. Hutchens,
and
E. Coney.
Pulmonary haemodynamics in the exercising horse and their relationship to exercise-induced pulmonary haemorrhage.
Br. Vet. J.
149:
419-428,
1993[Medline].
22.
O'Callaghan, M. W.,
J. R. Pascoe,
W. S. Tyler,
and
D. K. Mason.
Exercise-induced pulmonary haemorrhage in the horse: results of a detailed clinical, post-mortem and imaging study. VIII. Conclusions and implications.
Equine Vet. J.
19:
428-434,
1987[Medline].
23.
Olsen, S. C.,
C. P. Coyne,
B. S. Lowe,
N. Pelletier,
E. M. Raub,
and
H. H. Erickson.
Influence of furosemide on hemodynamic responses during exercise in horses.
Am. J. Vet. Res.
53:
742-746,
1992[Medline].
24.
Parker, J. C.,
J. L. Ardell,
C. R. Hamm,
S. A. Barman,
and
P. J. Coker.
Regional pulmonary blood flow during rest, tilt, and exercise in unanesthetized dogs.
J. Appl. Physiol.
78:
838-846,
1995[Abstract/Free Full Text].
25.
Pascoe, J. R.
Exercise-induced pulmonary hemorrhage: a unifying concept.
Am. Assoc. Eq. Pract.
42:
220-226,
1996.
26.
Pascoe, J. R.,
A. E. McCabe,
C. E. Franti,
and
R. M. Arthur.
Efficacy of furosemide in the treatment of exercise-induced pulmonary hemorrhage in Thoroughbred racehorses.
Am. J. Vet. Res.
46:
2000-2003,
1985[Medline].
27.
Pelletier, N.,
N. E. Robinson,
L. Kaiser,
and
F. J. Derksen.
Regional differences in pulmonary artery endothelial function: a role in the site of EIPH?
Equine Vet. J. Suppl.
18:
99-103,
1995.
28.
Rose, R. J.,
D. K. Hendrickson,
and
P. K. Knight.
Clinical exercise testing in the normal Thoroughbred racehorse.
Aust. Vet. J.
67:
345-348,
1990[Medline].
29.
Seeherman, H. J.,
and
E. A. Morris.
Comparison of yearling, two-year-old and adult Thoroughbreds using a standardised exercise test.
Equine Vet. J.
23:
175-184,
1991[Medline].
30.
Seeherman, H. J.,
and
E. A. Morris.
Methodology and repeatability of a standardised treadmill exercise test for clinical evaluation of fitness in horses.
Equine Vet. J. Suppl.
9:
20-25,
1990.
31.
Sweeney, C. R.,
and
L. R. Soma.
Exercise-induced pulmonary hemorrhage in Thoroughbred horses: response to furosemide or hesperidin-citrus bioflavinoids.
J. Am. Vet. Med. Assoc.
185:
195-197,
1984[Medline].
32.
Sweeney, C. R.,
L. R. Soma,
C. A. Bucan,
and
S. G. Ray.
Exercise-induced pulmonary hemorrhage in exercising Thoroughbreds: preliminary results with preexercise medication.
Cornell Vet.
74:
263-268,
1984[Medline].
33.
Sweeney, C. R.,
L. R. Soma,
A. D. Maxxon,
J. E. Thompson,
S. J. Holcombe,
and
P. A. Spencer.
Effects of furosemide on the racing times of Thoroughbreds.
Am. J. Vet. Res.
51:
772-778,
1990[Medline].
34.
Wagner, P. D.,
J. R. Gillespie,
G. L. Landgren,
M. R. Fedde,
B. W. Jones,
R. M. DeBowes,
R. L. Pieschl,
and
H. H. Erickson.
Mechanism of exercise-induced hypoxemia in horses.
J. Appl. Physiol.
66:
1227-1233,
1989[Abstract/Free Full Text].
35.
West, J. B.
Regional differences in the lung.
Chest
74:
426-437,
1978[Free Full Text].
36.
West, J. B.,
O. Mathieu-Costello,
J. H. Jones,
E. K. Birks,
R. B. Logemann,
J. R. Pascoe,
and
W. S. Tyler.
Stress failure of pulmonary capillaries in racehorses with exercise-induced pulmonary hemorrhage.
J. Appl. Physiol.
75:
1097-1109,
1993[Abstract/Free Full Text].
37.
West, J. B.,
K. Tsukimoto,
O. Mathieu-Costello,
and
R. Prediletto.
Stress failure in pulmonary capillaries.
J. Appl. Physiol.
70:
1731-1742,
1991[Abstract/Free Full Text].
J APPL PHYSIOL 86(6):2034-2043
8570-7587/99 $5.00
Copyright © 1999 the American Physiological Society
TOP
ABSTRACT
INTRODUCTION
