Effects of warfarin on blood rheology in navicular disease


Amin, T.M.; Sirs, J.A.; Allen, B.V.; Colles, C.M.

Research in Veterinary Science 40(3): 308-312

1986


A preliminary investigation has been undertaken of blood rheology in horses and ponies, its variation in navicular disease and the changes following treatment with warfarin. Erythrocyte flexibility, measured by a centrifuge packing technique, is higher in horses (30 per cent min-1) than in ponies (23.8 per cent min-1). There are corresponding differences in blood viscosity. The high erythrocyte flexibility in horses is caused by an unknown factor present in plasma. The erythrocyte flexibility in horses with navicular disease is even higher, at 38.5 per cent min-1. Treatment with warfarin reduces the flexibility to just below the normal value. A significant fall in plasma viscosity and erythrocyte flexibility was found after treatment of four ponies with 6 mg warfarin daily for one week.

Research
in
Veterinary
Science
1986,
40,
308-312
Effects
of
warfarin
on
blood
rheology
in
navicular
disease
T.
M.
AMIN,
J.
A.
SIRS,
Department
of
Physiology
and
Biophysics,
St
Mary's
Hospital
Medical
School,
Paddington,
London,
W2
IPG,
B.
V.
ALLEN,
C.
M.
COLLES,
Animal
Health
Trust,
Balaton
Lodge,
Newmarket
CB8
7DW
A
preliminary
investigation
has
been
undertaken
of
blood
rheology
in
horses
and
ponies,
its
variation
in
navicular
disease
and
the
changes
following
treatment
with
warfarin.
Erythrocyte
flexibility,
measured
by
a
centrifuge
packing
technique,
is
higher
in
horses
(30
per
cent
min
-
I)
than
in
ponies
(23.8
per
cent
min
-
').
There
are
corresponding
differences
in
blood
viscosity.
The
high
erythrocyte
flexibility
in
horses
is
caused
by
an
unknown
factor
present
in
plasma.
The
erythrocyte
flexibility
in
horses
with
navicular
disease
is
even
higher,
at
38.5
per
cent
min
-1
.
Treatment
with
warfarin
reduces
the
flexibility
to
just
below
the
normal
value.
A
significant
fall
in
plasma
viscosity
and
erythrocyte
flexibility
was
found
after
treatment
of
four
ponies
with
6
mg
warfarin
daily
for
one
week.
BLOOD
rheology
is
the
study
of
the
flow
properties
of
whole
blood
and
its
various
constituents,
including
the
shape
and
deformability
of
red
blood
cells.
Although
this
discipline
has
become
established
in
human
medicine
over
the
last
30
years,
little
reference
has
been
made
to
it
in
veterinary
medicine
and
only
passing
reference
has
been
made
to
animal
rheology
in
haematological
journals.
Rheological
changes
in
blood
are
well
recognised
as
being
complicating
factors
in
human
diseases,
such
as
sickle
cell
anaemia
and
diabetes
mellitus.
This
paper
represents
the
first
comparable
report
of
significant
rheological
changes,
and
their
correction,
occurring
in
an
equine
disease.
As
this
field
is
unfamiliar
to
many
veterinary
readers,
a
brief
description
of
the
techniques
used,
and
a
definition
of
the
units,
is
given
below.
The
forelimb
lameness
in
horses,
known
as
navicular
disease,
was
probably
first
described
by
Jeremiah
Bridges
(1752).
It
was
at
one
time
given
the
descriptive
name
`navicular
joint
disease'
by
Turner
(Youatt
1836).
Early
in
the
development
of
the
disease
it
becomes
clinically
evident
when
trotting
as
a
shortening
of
the
forelimb
stride,
the
feet
being
placed
on
the
ground
toe
first.
The
condition
is
nearly
always
present
in
both
forelimbs
and
is
slowly
pro-
gressive.
About
one
third
of
all
cases
of
chronic
forelimb
lameness
in
horses
in
England
has
been
attributed
to
navicular
disease
(Colles
1982a).
From
descriptions
of
the
disease
by
Bridges
(1752),
Moor-
croft
(1819)
and
Turner,
it
was
generally
accepted
that
the
lameness
involved
ulcerations
and
erosions
of
the
fibrocartilage
of
the
navicular
bone
and
lesions
of
the
deep
digital
flexor
tendon.
However,
over
the
last
decade
Colles
(1979a)
has
shown
that
cartilage
erosion
is
seldom
the
cause
of
lameness
and
tendon
lesions
are
not
the
decisive
factor.
As
long
ago
as
1885,
Walley
suggested
that
the
disease
might
be
caused
by
a
circulatory
disturbance,
as
more
recently
have
Hickman
(1964)
and
Nemeth
(1972).
In
1977,
Colles
and
Hickman
showed
that
arterial
thrombosis
is
a
common
pathological
change
in
navicular
disease
and
suggested
it
led
to
ischaemic
necrosis.
Recent
investigations
have
established
(Colles
1983)
that
the
clearance
time
of
a
contrast
agent
in
blood
flowing
through
the
feet
is
extended
from
about
12
seconds
in
normal
animals,
to
one
to
two
minutes
in
cases
of
navicular
disease,
and
Svalostoga
and
Smith
(1983)
have
shown
a
raised
sub-
cortical
bone
blood
pressure
in
diseased
cases.
This
suggested
that
navicular
disease
may
in
part
be
due
to
venous
congestion,
producing
a
rise
in
blood
presure,
which
in
turn
initiates
secondary
arterial
disease.
Current
treatment
is
aimed
at
improving
blood
flow
by
correct
shoeing,
regular
exercise
and
anti-
coagulant
therapy
with
warfarin.
Following
the
intro-
duction
of
warfarin
treatment
by
Colles
(1979b),
82
per
cent
of
treated
horses
were
able
to
return
to
work
and
61
per
cent
have
remained
sound
for
four
years
(Colles
1983).
A
number
of
treated
horses
have
been
free
of
the
disease
and
working
for
five
to
eight
years.
Little
is
known
of
the
pharmacokinetics
of
warfarin
in
the
horse.
As
a
number
of
cases
have
responded
satisfactorily
with
low
doses,
in
which
the
one-stage
prothrombin
time
would
not
have
been
considered
significantly
lengthened
in
man,
it
was
decided
to
investigate
the
possibility
that
the
improvement
in
blood
flow
was
due
to
a
modification
of
the
rheo-
logical
properties
of
blood.
Colles
(1982b)
had
noted
a
small,
but
significant,
drop
in
plasma
viscosity
following
treatment
with
warfarin.
Because
of
the
sparsity
of
rheological
data
on
horses
a
preliminary
308
Blood
rheology
in
navicular
disease
309
survey
was
first
undertaken
on
thoroughbreds,
cross-
breds
and
ponies.
Measurements
were
then
made
on
treated
and
untreated
horses
with
navicular
disease.
Finally
data
from
a
preliminary
trial
on
the
effect
of
treating
four
normal
ponies
with
warfarin
are
presented.
Materials
and
methods
Horses
The
following
groups
of
horses
were
used
in
this
study:
(1)
nine
mature
healthy
horses,
comprising
three
thoroughbreds,
three
crossbred
horses
and
three
crossbred
ponies;
(2)
two
crossbred
horses
with
chronic
navicular
disease;
(3)
six
crossbred
horses
under
treatment
with
warfarin
for
navicular
disease;
(4)
four
healthy
crossbred
ponies
which
were
each
given
6
mg
warfarin
by
mouth
daily
for
five
days
to
assess
the
short
term
changes
in
blood
rheology
following
warfarin
administration;
(5)
seventy-seven
horses
of
mixed
breeds,
which
were
used
to
assess
plasma
viscosity
before
and
after
warfarin
treatment.
Blood
sampling
Blood
was
obtained
by
jugular
venepuncture
and
anticoagulated
with
lithium
heparin
at
a
concentra-
tion
of
12.5
iu
ml
-
'
of
blood.
All
rheological
measurements
were
undertaken
within
six
hours
of
blood
collection.
Rheological
techniques
Viscosity
can
be
regarded
as
the
internal
friction
within
a
fluid
that
restrains laminar
flow
and
its
effect
is
present
whenever
relative
motion
exists
between
adjacent
layers
of
a
liquid.
The
coefficient
of
viscosity,
n,
is
the
constant
of
proportionality
between
the
force
applied
to
generate
flow
and
the
ratio
of
the
resultant
difference
in
velocity
of
two
adjacent
fluid
layers,
of
unit
area,
to
their
distance
apart.
The
ratio
of
the
difference
in
velocity
to
distance
apart
is
called
the
'shear
rate'.
If
the
force
is
measured
in
dynes
cm
-2
,
the
velocity
in
cm
sec
I
,
and
the
distance
apart
in
cm,
the
coefficient
of
viscosity
is
given
in
units
of
'poise'
dyne
sec
cm
-2
.
In
Si
units,
with
the
force
per
unit
area
in
Pascals
(kg
m
-
'
sec
-2
)
and
the
velocity
in
m
sec
-
',
the
coefficient
of
viscosity
is
given
in
units
of
Pascals.sec
(Pa.sec).
One
poise
is
equal
to
0.1
Pa.sec.
The
coefficient
of
viscosity
of
water
at
20°C
is
0.01
poise,
or
one
centi-
poise
(cp),
which
is
the
most
common
unit
in
inter-
national
use,
and
adopted
in
this
paper.
With
simple
fluids,
such
as
water,
ri
does
not
vary
with
the
flow
rate,
more
accurately
shear
rate,
and
the
flow
behaviour
is
termed
'Newtonian'.
Complex
fluids
such
as
blood
exhibit
non-Newtonian
behaviour,
the
viscosity,
rl,
increasing
as
the
flow
rate
(shear
rate)
falls.
This
is
believed
to
be
mainly
due
to
additional
Rouleaux
formation
and
aggregation
between
red
cells
at
slow
flow
rates.
There
are
two
current
methods
of
measuring
blood
viscosity;
by
monitoring
the
flow
rate
of
blood
through
a
capillary
tube
when
subject
to
a
constant
driving
pressure
or
by
placing
a
small
volume
of
blood
between
two
surfaces
rotating
relative
to
each
other
and
assessing
the
associated
drag.
A
fluid
flowing
through
a
tube
has
zero
velocity
at
the
vessel
wall
and
maximum
velocity
at
its
centre.
If
the
fluid
is
Newtonian,
the
variation
of
velocity
with
radial
distance
is
parabolic
and
the
flow
obeys
Poiseuille's
law.
The
difference
of
velocity
between
two
adjacent
fluid
layers
thus
varies
across
the
tube,
ie,
the
shear
rate
varies.
Hence
capillary
viscometers
can
only
be
used
to
obtain
accurate
measurements
of
viscosity
with
Newtonian
fluids,
which
only
pertains
with
blood
at
high
shear
rates.
The
majority
of
the
following
results
were
obtained
using
a
Coulter-Harkness
capillary
viscometer
at
37.5°C.
As
designed,
the
pressure
of
the
mercury
column
in
the
Harkness
capillary
viscometer
could
not
be
sufficiently
reduced
to
give
flow
rates
with
a
mean
shear
rate
of
200
to
250
sec
-
',
comparable
to
measurements
with
the
Wells-Brookfield
rotational
visometer
discussed
later.
This
difficulty
was
simply
overcome
by
applying
a
small
back
pressure
to
the
mercury
column
outlet
to
the
air
through
pressure
tubing
connected
to
a
sealed
plastic
bottle
which
could
be
squeezed
by
a
clamp.
The
clamp
was
tightened
just
sufficiently
to
reduce
the
blood
flow
to
a
mean
shear
rate
of
about
230
sec
I
.
At
this
flow
rate
it
was
established
that
the
blood
of
horses
is
effec-
tively
Newtonian
and
hence
its
viscosity,
relative
to
plasma,
could
be
accurately
assessed.
In
principle
rotational
viscometers
have
a
geo-
metric
construction
that
permits
the
fluid
to
be
subject
to
a
uniform
shear
rate.
The
variation
of
the
coefficient
of
viscosity
with
shear
rate
can
then
be
ascertained
by
using
different
speeds
of
rotation.
This
is
important
in
situations
of
slow
flow
and
high
such
as
occurs
during
flow
through
large
veins
in
some
disease
states.
A
Wells-Brookfield
cone-on-
plate
rotational
viscometer
was
used
in
an
attempt
to
obtain
this
information
for
horses'
blood.
The
geo-
metric
construction
of
this
viscometer
is
basically
that
of
an
inverted
cone,
point
downwards,
which
can
be
rotated
just
above
a
flat
plate.
While
this
method
gives
values
with
normal
human
blood,
at
high
shear
rates,
about
5
per
cent
higher
than
the
capillary
technique,
with
horses'
blood
and
plasma
the
results
were
erratic.
On
occasions,
measurements
of
plasma
viscosity
on
the
Wells-
Brook
field
were
more
than
twice
the
value
obtained
310
T.
M.
Amin,
J.
A.
Sirs,
B.
V.
Allen,
C.
M.
CoIles
with
the
Coulter-Harkness
viscometer.
Preliminary
investigations
suggested
that
there
is
an
artefact
when
using
rotational
viscometers
without
a
guard
ring,
owing
to
the
formation
in
horses'
blood
of
a
relatively
viscous
film
at
the
air-plasma
interface.
No
simple
method
of
avoiding
this
effect
in
the
Wells-
Brookfield
viscometer
was
discovered.
The
viscosity
of
whole
blood
varies
with
haematocrit
in
a
semi-
logarithmic
manner
and
the
values
of
viscosity
have
been
standardised
to
a
fixed
haematocrit,
by
the
procedure
given
in
the
results
section.
In
addition
to
the
influence
of
haematocrit
and
plasma
viscosity,
the
viscosity
of
blood
depends
on
the
shape
and
flexibility
of
red
blood
cells.
The
latter
factor
is
a
measure
of
the
viscoelastic
properties
of
the
cells
and
is
ascertained
by
measuring
the
rate
at
which
the
cells
are
deformed
when
subjected
to
stress.
In
the
present
case
this
factor
was
measured
using
the
strobo-
scopic
recording
method
of
Amin
et
al
(1983).
This
technique
has
the
advantages
that
no
additional
preparation
or
washing
of
the
cells
is
involved
and
only
05
ml
of
blood
is
required.
In
principle
it
relies
on
the
fact
that
at
a
haematocrit
of
35
to
42
per
cent
randomly
oriented
red
blood
cells
must
be
in
contact
with
each
other.
During
centrifugation
of
blood,
at
a
haematocrit
of
greater
than
35
per
cent,
the
separa-
tion
of
cells
and
plasma
is
brought
about
by
the
weight
of
cells,
one
on
top
of
the
other,
deforming
and
squashing
the
cells
at
the
bottom
of
the
centrifuge
tube.
The
less
flexible
are
the
cells,
the
slower
is
the
rate
of
packing
and
rigid
cells
cannot
be
made
to
pack
at
all.
The
rate
cells
pack
in
whole
blood,
under
a
constant
acceleration
of
200
g,
is
recorded
photo-
graphically,
using
stroboscopic
illumination.
The
rate
of
packing,
in
per
cent
min
I,
is
ascertained
from
the
recorded
change
of
the
length
of
the
red
cell
column
with
time.
To
allow
for
variations
of
the
individual
haematocrits,
measured
at
13,000
g,
a
calibration
curve
of
the
packing
rate
with
haematocrit
was
obtained
for
each
breed
and
the
quoted
packing
rates
have
been
corrected
using
these
curves
to
a
standard
haematocrit
of
45
per
cent.
The
stress
applied
to
pack
the
cells
by
centrifugation
also
depends
on
the
difference
of
cell
to
plasma
specific
gravities.
The
TABLE
1:
Rheological
parameters
of
horses
and
ponies
small
variations
of
this
factor
between
different
horses,
and
man
and
horse,
had
no
significant
effect
on
the
packing
rate.
Plasma
fibrinogen,
haematocrit
and
mean
corpuscular
volume
The
fibrinogen
concentration
was
estimated
using
the
thrombin
clot
technique
of
Rampling
and
Gaffney
(1976).
The
haematocrit
was
obtained
by
centrifugation
on
a
Hawksley
microhaematocrit
centrifuge,
at
13,000
g
for
three
minutes.
The
red
cell
count
was
measured
using
a
Coulter
counter.
Values
of
the
mean
corpuscular
volume
(Mcv)
were
calcu-
lated
by
dividing
the
haematocrit
by
the
cell
count.
Results
The
observed
rheological
parameters
for
horses
and
ponies
are
shown
in
Table
1.
The
erythrocyte
flexibility
was
very
high,
the
mean
packing
rate
for
horses
being
30
per
cent
min
-
I
and
for
ponies
23.8
per
cent
min
-
I,
compared
with
about
7
per
cent
min
-
'
for
man.
The
increased
flexibility
in
horses
is
associated
with
an
uknown
factor
in
horse
plasma
(Amin
and
Sirs
1982).
If
human
red
cells
are
suspended
in
horse
plasma
their
packing
rate
is
increased
to
31.5
per
cent
min
-
.
The
factor
can
be
removed
from
red
cells
by
repeated
washing
and
resuspension
in
Ringer
Locke
solution.
The
packing
rate
for
horse
erythrocytes
is
then
4.0
per
cent
min
-
'
and
for
human
cells
1
2
per
cent
min
-
I.
If
horse
plasma
is
defibrinated,
by
incubation
at
56°C,
and
washed
horse
cells
resuspended
in
serum,
there
is
no
effective
change
of
the
packing
rate.
This
suggests
that
the
factor
in
plasma
is
not
fibrinogen.
An
increase
of
erythrocyte
flexibility
should
in
principle
lower
the
whole
blood
viscosity
at
high
shear
rates
(Chien
et
al
1967).
If
allowance
is
made
for
the
different
haematocrits,
by
using
the
relationship:
(r1,-)
14
=
010
1-i
s
where
yi
r
is
the
relative
viscosity
(ie,
blood
viscosity/
Blood
viscosity
Plasma
Relative
ABC
PCV
(H
s
=
mean
viscosity
viscosity
Fibrinogen
..,flexibility
MCV
Animal
n
(litre
litre
-1
)
PCV)
(cp)
(cp)
(H
s
=
0-451
1mg
m1
-1
)
t
f
min
-1
1
Ifl)
,T
SD
Ft
SD
Tc
SD
T(
SD
SD:
SD
SD
Horses
6
0.38
0.06
3.11
0.23
1.22
0.05
3.01
0.15
3.98
0.23
30.0
2.2
44.6
2.7
Crossbred
ponies
3
0.36
0.04
316
0.34
1.34
0.07 3.19
0•28
3.32
0.67
23.8
2.13
45.3
1.2
Man
1
0.46
3.89
1.22
3.10
2.70
7.3
86.0
PCV
Packed
cell
volume
ABC
Red
blood
cell
MCV
Mean
corpuscular
volume
Blood
rheology
in
navicular
disease
311
plasma
viscosity)
at
an
observed
haematocrit
of
H
and
ns
the
relative
viscosity
at
the
standard
haematocrit,
H
s
;
the
relative
viscosity
of
horses'
blood
is
com-
parable
to
that
for
man.
The
relative
viscosity,
at
an
haematocrit
of
45
per
cent,
is
3.1
for
man,
compared
to
3.01
for
the
mean
value
of
the
data
for
horses
shown
in
Table
1.
It
is
believed
that
this
anomaly
is
due
to
the
different
shape
of
horse
and
human
erythrocytes.
By
progressively
increasing
the
back-
pressure
in
the
capillary
viscometer,
and
decreasing
the
flow,
the
viscosity
of
blood
at
shear
rates
down
to
20
seconds
could
be
estimated.
At
this
slower
flow
it
was
observed
that
the
viscosity
of
horses'
blood
was
nearly
twice
that
of
human
blood,
at
an
haematocrit
of
45
per
cent
and
37.3°C.
In
two
untreated
horses
with
chronic
navicular
disease,
the
erythrocyte
flexibility
was
significantly
elevated
to
the
same
level
of
38.5
per
cent
min
-1
.
The
flexibility
values
for
six
horses
with
navicular
disease,
treated
with
warfarin,
were
23
1,
30.7,
28.8,
26.0,
35.2
and
24.1
per
cent
min
-
'.
The
values
were
signifi-
cantly
higher
than
normal
in
untreated
horses
and
lower
after
treatment.
A
similar
significant
reduction
of
plasma
viscosity
has
been
observed
following
treat-
ment
with
warfarin.
Before
treatment
the
mean
plasma
viscosity
of
77
horses
was
1.62
cp,
with
a
standard
deviation
of
0.10.
Following
four
weeks
treatment
with
warfarin
the
mean
plasma
viscosity
was
1.56
±
0.08.
The
difference
is
statistically
significant
at
a
probability
level
of
less
than
0.1
per
cent.
To
obtain
some
indication
of
how
quickly
and
to
what
extent
rheological
changes
occur,
and
to
establish
if
the
effect
of
warfarin
only
occurred
in
animals
with
navicular
disease,
a
preliminary
trial
was
undertaken
with
four
control
ponies.
Measurements
of
blood
and
plasma
viscosities,
erythrocyte
flexi-
bility,
plasma
fibrinogen
level,
haematocrit
and
mcv
were
made
just
before
treatment.
Each
animal
was
then
given
a
daily
dose
of
6
mg
warfarin
by
mouth.
Measurements
of
the
rheological
parameters
were
made
48
hours
later
and
then
on
each
of
the
following
five
days.
The
values
for
each
parameter
for
the
pony
with
the
most
systematic
variation
is
shown
in
Table
2.
The
other
three
ponies
had
similar
variations
but
with
wider
fluctuations
during
the
first
three
days
of
treatment.
In
particular,
the
flexibility
rose
tran-
siently
in
two
ponies
before
falling
below
its
initial
level.
By
the
sixth
day
of
the
trial
all
the
ponies
were
showing
comparable
changes
in
their
blood
rheology.
The
mean
percentage
changes
of
the
haematocrit,
plasma
viscosity,
blood
viscosity,
relative
viscosity
(In
[rh.1
H
-1
),
fibrinogen
concentration,
erythrocyte
flexibility
and
mcv
were
+0.5,
-11,
-8,
+6,
+4.5,
-
15
and
-
1
per
cent
respectively.
The
change
of
plasma
viscosity
was
statistically
significant
with
P<0.05
and
the
decrease
of
flexibility
at
P<0.01.
TABLE
2:
Effect
of
treatment
with
warfarin
on
blood
theology
in
a
pony
After
Day
Before
2
3
4
5
6
PCV
(litres
litre
-
)
Blood
viscosity
(cp)
Plasma
viscosity
(cp)
0.39
126
1.46
0.37
3.02
1.38
0.43
194
1.42
0.37
3.11
1.29
0.39
3.17
1.36
0.38
2.99
1.29
Relative
viscosity
2.23
2.19
2.78
2.42
2.34
2.32
(In
101
H
-1
Fibrinogen
(mg
m1
-1
)
R
BC
flexibility
1%
min
-1
)
2.08
116
30.0
2.12
2.37
26.0
2.40
4.23
24.7
2.42
4.34
28.5
2.21
4.18
27.5
2.25
129
24.1
MCV
(61
47.5
48.4
48.0
48.8
49.2
48.0
See
Table
1
for
key
There
was
an
associated
increase
of
(ln
[rid)
H
-
',
with
decrease
of
flexibility,
but
with
the
small
number
of
ponies
it
was
not
possible
to
show
that
it
was
statis-
tically
significant.
The
differences
between
the
other
factors
were
not
significant.
In
another
experiment
horse
and
human
cells
were
incubated
with
warfarin
at
37
3°C
for
several
hours
with
no
significant
rheo-
logical
changes
relative
to
controls
without
warfarin.
Discussion
It
has
been
known
for
some
time
that
the
blood
rheology
of
horses
is
unusual
in
several
respects
(de
Haan
1918,
Fahraeus
1921).
There
are
significant
dif-
ferences
in
erythrocyte
aggregation,
Rouleaux
formation
and
plasma
protein
composition.
Higher
aggregation
increases
blood
viscosity
at
low
shear
rates.
Rouleaux
formation
requires
a
change
of
cell
-
shape,
the
biconcave
surfaces
of
the
cells
being
flattened
in
the
region
of
contact
(Chien
et
al
1971,
Rowlands
and
Skibo
1972).
The
high
flexibility
of
horse
cells
facilitates
this
shape
change
and
promotes
aggregation,
with
an
associated
increase
of
low
shear
rate
viscosity,
as
observed
in
this
study.
Other
species,
such
as
sheep
and
cattle,
are
known
to
have
relatively
inflexible
cells
and
have
negligible
Rouleaux
formation.
An
increase
of
erythrocyte
flexibility
may
occur
in
man
in
a
number
of
pathological
conditions
with
high
fibrinogen
levels
(eg,
during
the
post-
operative
period,
in
bronchitis
and
some
forms
of
hypertension).
There
is
an
associated
increase
of
venous
thrombosis
in
man
in
these
circumstances.
It
would
appear
that
in
the
horse
this
risk
is
offset
by
its
ability
to
reduce
its
haematocrit
at
rest
to
30
to
35
per
cent,
by
sequestration
of
cells
in
the
spleen.
The
lower
haematocrit
has
a
pronounced
effect
in
reducing
the
degree
of
aggregation
and
there
is
a
corresponding
fall
of
blood
viscosity
at
low
shear
rates.
The
observation
312
T.
M.
Amin,
J.
A.
Sirs,
B.
V.
Allen,
C.
M.
Colles
ti
that
erythrocyte
flexibility
is
increased
in
untreated
horses
with
navicular
disease
suggests
that
the
balance
may
be
critical.
Only
a
relatively
small,
but
signifi-
cant,
increase
of
erythrocyte
flexibility
is
necessary
to
produce
a
clinically
adverse
increase
of
Rouleaux,
aggregation
and
low
shear
rate
viscosity.
This
could
account
at
least
in
part
for
the
venous
congestion
at
the
extremities
of
the forelimbs,
which
would
reduce
blood
flow
through
the
capillaries
and
arteries
and
cause
ischaemia,
as
suggested
by
Colles
and
Hickman
(1977).
This
is
supported
by
the
significant
changes
of
erythrocyte
flexibility
and
plasma
viscosity
that
were
found
in
horses
being
treated
with
warfarin
and
during
the
trial
of
this
drug
in
ponies.
How
warfarin
produces
this
effect
is
not
known.
The
advantage
to
the
horse
of
having
smaller
and
very
flexible
erythrocytes
appears
to
be
that
they
facilitate
a
more
rapid
exchange
of
oxygen.
Decreas-
ing
the
flexibility
of
human
erythrocytes
lowers
the
rate
at
which
oxygen
is
taken
up
by
haemoglobin
in
the
cell
(Sirs
1968).
There
is
an
apparent
contradiction
in
the
rheological
behaviour
of
horses'
blood
in
that,
although
the
red
cells
are
more
flexible,
the
blood
viscosity
at
high
shear
rates
is
comparable,
or
slightly
higher,
than
for
man
at
the
same
haematocrit.
This
is
because
the
shape
of
horse
erythrocytes
is
more
asymmetric
than
human
cells,
which
increases
the
effective
viscosity.
The
viscosity
of
plasma
is
sensitive
to
the
fibrinogen
concentration
for
the
same
reason.
Amin
and
Sirs
(1982)
have
reported
that
the
shape
factor,
K,
for
horses
is
3.8,
compared
to
3.3
in
man,
and
this
balances
the
difference
in
erythrocyte
flexi-
bility.
It
can
be
shown
that,
at
high
shear
rates,
the
slope
of
a
plot
of
the
natural
logarithm
of
the
relative
viscosity
against
haematocrit
depends
on
the
product
of
the
shape
factor
and
reciprocal
of
the
erythrocyte
flexibility.
This
is
consistent
with
the
increase
of
in
(r1
r
)H
--1
shown
in
Table
2,
as
the
flexibility
decreases.
The
constancy
of
the
mcv
supports
the
view
that
this
is
not
due
to
a
change
of
shape.
So
over
all,
in
six
days
of
the
trial,
there
would
not
be
time
for
erythropoesis
to
affect
the
cells,
there
was
no
change
of
shape,
the
erythrocyte
flexibility
decreased
and
the
plasma
viscosity
fell.
The
most
likely
explanation
of
the
action
of
warfarin,
consistent
with
these
observa-
tions,
is
that
it
acts
indirectly
on
some
factor
in
the
plasma
that
can
modify
erythrocyte
flexibility.
References
AMIN,
T.
&
SIRS,
J.
A.
(1982)
Microvascular
Research
24,
235
AMIN,
T.
M.,
SIRS,
J.
A.
&
TURNER,
P.
(1983)
Physics
in
Medicine
and
Biology
28,
269-275
BRIDGES,
J.
(1752)
Quoted
in
Reeks,
H.
C.
(1925)
Diseases
of
the
Horse's
Foot.
Chicago,
Alex
Agar
CHIEN,
S.,
LUSE,
S.
A.,
JAN,
K.
M.,
USAMI,
S.,
MILLER,
L.
H.
&
FREEMONT,
H.
(1971)
Proceedings
of
the
Sixth
European
Conference
on
Microcirculation.
Basel,
Karger.
pp
29-34
CHIEN,
S.,
USAMI,
S.,
DELLENBACK,
R.
J.
&
GREGERSEN,
M.
I.
(1967)
Science
157,
827-829
COLLES,
C.
M.
(1979a)
Veterinary
Record
104,
133-137
COLLES,
C.
M.
(1979b)
Equine
Veterinary
Journal
11,
187-190
COLLES,
C.
M.
(I
982a)
Veterinary
Record
Supplement
In
Practice
4,
29-36
COLLES,
C.
M.
(1982b)
PhD
thesis,
University
of
London
COLLES,
C.
M.
(1983)
Proceedings
of
the
American
Association
of
Equine
Practitioners
29,
265-270
COLLES,
C.
M.
&
HICKMAN,
J.
(1977)
Equine
Veterinary
Journal
9,
150-152
FAHRAEUS,
R.
(1921)
Acta
Medica
Scandinavica
55,
1
DE
HAAN,
J.
(1918)
Biochemische
Zeitschrift
86,
298-308
HICKMAN,
J.
(1964)
Veterinary
Orthopaedics.
Edinburgh
and
London,
Oliver
&
Boyd.
pp
104-117
MOORCROFT,
W.
(1819)
Quoted
by
Youatt
(1836)
NEMETH,
F.
(1972)
Netherlands
Journal
of
Veterinary
Science
5,
56-71
RAMPLING,
M.
W.
&
GAFFNEY,
P.
J.
(1976)
Clinica
Chimica
Acta
67,
43-52
ROWLANDS,
S.
&
SKIBO,
L.
(1972)
Thrombosis
Research
1,
47-58
SIRS,
J.
A.
(1968)
Hemorheology.
Proceedings
of
the
First
Inter-
national
Conference.
Ed
A.
L.
Copley.
London,
Pergamon
Press.
pp
505-513
SVALOSTOGA,
E.
L.
&
SMITH,
M.
(1983)
Nordisk
Veterinaer-
medicin
35,
31-37
WALLEY
(1885)
Veterinary
Journal
20-21,
266-271
YOUATT,
W.
(1836)
Veterinarian
9,
362-363
Accepted
October
31,
1985