Download Coronary pressure-function and steady-state pressure

821
Coronary Pressure-Function and Steady-State
Pressure-Flow Relations During
Autoregulation in the Unanesthetized Dog
John M. Canty Jr.
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
The present study was intended to define the interrelation among endocardial flow, endocardial
function, and coronary arterial pressure during spontaneous autoregulation in the left ventricle
of chronically instrumented unanesthetized dogs. Steady-state sonomkrometric measurements
of regional function and epicardial coronary artery pressure were used to determine the lower
pressure limit of endocardial autoregulation while global indexes of myocardial demand
remained constant. Transmural wall thickening in the circumflex bed remained unchanged
( ± 5 % of control values) until coronary pressure fell below 39 ± 5 . 6 (SD) mm Hg. Endocardial
segment shortening was similarly constant until coronary pressure fell below 42 ± 7.4 mm Hg.
There was no significant change in endocardial flow as coronary pressure was reduced over the
autoregulatory plateau from 84 to 49 mm Hg (1.05-0.99 ml/min/g, p=NS). Below the critical
pressure limits, small additional reductions in pressure were associated with marked reductions
in both endocardial flow and function. The coronary pressure-function relation was linear as
well as steep hi this range for both wall thickening (r=0.94 ±0.05) and segment shortening
(r=0.96 ±0.03). Although the relation between endocardial flow and function showed more
variability than pressure-function relations at low pressures, wall thickening reductions and
endocardial flow reductions related on a nearly one-to-one basis. The present study establishes
that the coronary pressure-function relation can be used to define the lower limit of endocardial
autoregulation. It also indicates that the lower pressure limit of endocardial autoregulation is
considerably less than hi anesthetized animals (40 vs. 70 mm Hg) and that steady-state flow
above this limit is controlled more tightly. Although these differences may relate to systemic
hemodynamics, it seems likely that general anesthesia and/or acute surgical instrumentation
alter coronary autoregulation under at least some experimental circumstances. (Circulation
Research 1988;63:821-836)
C
oronary blood flow is autoregulated over a
wide pressure range as coronary artery
pressure is reduced.'- 3 Current knowledge
concerning coronary autoregulation is based entirely
on studies conducted in anesthetized animals. While
the precise mechanisms responsible for autoregulation are still controversial,4 several features in the
From the Departments of Medicine and Physiology, State
University of New York at Buffalo and the Erie County Medical
Center, Buffalo, New York.
Preliminary reports of this work were presented at the 58th
and 60th Annual Scientific Sessions of the American Heart
Association, 1985 and 1987.
Supported by grants from the National Heart, Lung, and
Blood Institute (1RO1-HL-37682, 5PO1-HL-15194, and 1KO8HL-01168) and the American Heart Association with funds
contributed from the Western New York Affiliate (83-717 and
86-912).
Address for correspondence: John M. Canty Jr., MD, SUNY/
B Clinical Center, Room CC155, Erie County Medical Center,
462 Grider Street, Buffalo, NY 14215.
Received June 15, 1987; accepted May 17, 1988.
coronary circulation are noteworthy. First, there
are significant transmural variations. Endocardial
flow autoregulation is lost when mean coronary
pressure falls below 70 mm Hg (diastolic pressure,
below 50 mm Hg) in anesthetized animals.3-6 In
contrast, epicardial flow is autoregulated over a
wider pressure range.3-6 Second, there is a close
coupling between endocardial flow and function
below the lower autoregulatory limit.7-10 In this
range, reductions in endocardial function presumably reflect inadequate oxygen delivery. Previous
studies have not examined regional flow, myocardial function, and coronary pressure simultaneously
during autoregulation. Thus, the specific relation
between coronary pressure and endocardial function has not been established.
Recent experimental studies in anesthetized animals have challenged the notion of "perfect" coronary autoregulation (i.e., autoregulation that is
able to match flow to local vasodilator reserve
822
Circulation Research Vol 63, No 4, October 1988
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
under all circumstances). It is now well established
that reductions in flow (and in some studies, myocardial function) occur at reduced coronary artery
pressure in the presence of considerable vasodilator
reserve, which is recruitable pharmacologically.11-14
Other studies have demonstrated that regional oxygen consumption decreases continuously as coronary pressure is reduced despite constant hemodynamic determinants of oxygen consumption. 31316
The potential effects of these changes on endocardial flow or a sensitive index of endocardia] function
have not been evaluated. Each such observation
raises the possibility that factors other than local
vasodilator reserve modulate coronary autoregulation under at least some experimental circumstances. Because acute instrumentation and general
anesthesia have been shown to alter the relation
between endocardia] flow and function,9-10 it seems
plausible that differences in coronary autoregulatory responses could also occur in these two experimental settings.
The present study was performed to define interrelations among coronary pressure, transmural myocardial perfusion, and endocardial function in chronically instrumented unanesthetized dogs. Specific
aims were to 1) determine whether the relation
between coronary pressure and endocardial function can be used to define the lower endocardial
autoregulatory pressure limit; 2) quantify steadystate endocardial autoregulation and the degree to
which regional flow and function (an indirect index
of changes in endocardial oxygen consumption)
may change over the autoregulatory "plateau," and
3) compare the relative sensitivity at comparable
levels of regional ischemia of endocardial flowfunction relations based on segment shortening and
transmural wall thickening. The findings contrast
with previous studies in anesthetized animals indicating both a significantly lower endocardial autoregulatory pressure limit (—40 mm Hg) and maintained oxygen delivery and function until this lower
autoregulatory limit is reached.
thoracic aorta for microsphere injection and reference blood sampling, respectively. Ascending aortic pressure was measured with a Teflon angiocath
(22 gauge) inserted through vessel wall and connected to Tygon microbore tubing (0.02 in. i.d., 0.06
in. o.d., 24 in. long). The frequency response of this
system when connected to a Statham P23dB pressure transducer and flushed with degassed saline
was flat to at least 15 Hz and therefore sufficient to
measure phasic variations in pressure at the heart
rates encountered. Left ventricular pressure was
measured by a Konigsberg Model P6.5 micromanometer placed through a stab wound in the ventricular apex and secured with a purse-string suture.
Pacing leads were sewn onto the left atrial appendage. The left circumflex artery was dissected free
for 1-2 cm proximal to the first marginal branch. A
fluid-filled hydraulic occluder was placed around
the proximal circumflex artery. A Teflon angiocath
was inserted into the circumflex artery distal to the
occluder with the tip facing downstream for coronary arterial pressure measurement, with care taken
to avoid entering any side branches.
Endocardial function was measured in the distal
circumflex and anterior descending region with piezoelectric crystals placed to measure both wall thickness and segment length, as previously described.17
Circumflex crystals were placed in the posterobasal
free wall (above the minor axis), and anterior
descending crystals were placed in the apical free
wall region (below the minor axis), with care taken
to avoid crystal placement in the anterior or posterior papillary muscles. In the majority of animals,
segment length crystals spanned the same region
occupied by the endocardial wall thickness crystal
and, thus, assessed endocardial function in a similar
anatomic region. Segment length crystals were oriented parallel to the estimated endocardial fiber
orientation. For the inner quarter of the heart, this
Materials and Methods
Animal Preparation
Studies were conducted in awake, chronically
instrumented dogs. All experimental procedures
were performed in accordance with institutional
guidelines. A total of 22 adult mongrel dogs (27 ± 2.3
[SD] kg) were studied. Anesthesia was induced by
injection of sodium thiamylal (20 mg/kg i.v.). After
endotracheal intubation, a surgical plane of anesthesia was maintained with a nitrous oxide (~60%),
oxygen (—40%), and halothane (—1-2%) mixture
during mechanical ventilation. Under sterile conditions, a thoracotomy was performed in the fifth left
intercostal space.
The preparation is illustrated in Figure 1. Tygon
catheters (0.0625 in. i.d., 0.125 in. o.d., 24 in. long)
were placed into the left atrium and descending
LAD W»D
Thlcknata
FIGURE 1. Schematic diagram of experimental preparation. LC, left circumflex; LAD, left anterior descending
coronary artery; PLC, left circumflex pressure; PAo, aortic
pressure; PLV, left ventricular pressure; LA, left atrium.
(See text for discussion.)
Canty
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
angle (a) varies significantly with respect to the
circumferential plane (a- + 10° to +50°, Figure 5,
reference 18). Because of these endocardial variations in fiber orientation with depth, the segment
length crystals were inserted at an approximate
angle of +30° to the circumferential plane. The
position of the crystals was carefully examined at
the time of necropsy. Six segment length pairs and
two wall thickness pairs were excluded because
they were not appropriately positioned. The innermost edge of the endocardial wall thickness crystals
spanned 92 ± 8% of the myocardium. Segment length
pairs were located 90 ± 10% of the distance across
the wall.
At the conclusion of instrumentation, the crystal
wires and catheters were tunneled through individual stab wounds in the 6th-7th left intercostal
space. The chest was closed, and the pneumonothorax evacuated with a chest tube. Animals were
given streptomycin (300 mg i.m.) and procaine
penicillin (300,000 units i.m.) for 3-5 days after
surgery. Catheters were flushed with sterile saline
and filled with heparin at 1-3-day intervals (10,000
Autoregulation in the Unanesthetized Dog
823
units/ml for the circumflex artery catheter and 1,000
units/ml for all other catheters). Enteric coated
aspirin (325 mg p.o.) was begun on the fourth day
after surgery and administered daily thereafter.
Transmural Coronary Autoregulation
Microsphere flow studies were conducted in 16
animals 19±11 days after instrumentation. Most
studies were conducted during light sedation with
Innovar-Vet (fentanyl 0.4 mg/ml and droperidol 20
mg/ml, 1-3 ml i.m.). The use of this sedation
resulted in stable systemic hemodynamics for 2-3
hours with substantially less variability than that
normally encountered in the unsedated state.
Although lightly sedated, the animals were conscious and easily excited by extraneous noise. Measurements were obtained with the animals lying
quietly on their right side. All pressure transducers
(Statham P23dB) were adjusted to the same height
and referenced to the dorsal spine to closely approximate mid heart level. The micromanometer was
calibrated at the beginning of each study by matching the systolic pressure to that measured simulta-
160
PAO
HQ|
(mm
ir
Ml
a
LC Segment
Length
(mm)
LAD Segment
Length
(mm)
ED ES
Mean P|_(;
88mmHg
47mmHg
36mmHg
30mmHg
28mmHg
26mmHg
23mmHg
Occlusion
Control
FIGURE 2. Analog recordings at selected levels of coronary pressure in an individual animal. Each panel represents a
single cardiac cycle during a steady-state level of coronary pressure reduction. Mean coronary pressure corresponding
to each measurement is illustrated below each panel. Solid vertical lines drawn on the recordings of wall thickness and
segment length represent end diastole (ED, onset of positive dP/dt) and end systole (ES, 20 msec before peak negative
dPIdt). During gradual pressure reduction produced by inflating the hydraulic occluder, regional circumflex function
remained constant over a wide coronary pressure range. When mean coronary pressure fell below 35 mm Hg (diastolic
pressure, ~20 mm Hg), further reductions in pressure were associated with large reductions in both segment shortening
and wall thickening over a relatively narrow pressure range. After restoration of coronary pressure (right panel),
circumflex function remained depressed in a fashion similar to stunned myocardium. PAo, aortic pressure; PLV, left
ventricular pressure; PLC, left circumflex pressure; dPIdt, first derivative of left ventricular pressure; LC, left circumflex;
LAD, left anterior descending coronary artery.
824
Circulation Research
Vol 63, No 4, October 1988
1.0 0.8 FLOW
(mlmln"'•(•')
•f
+
o.e 0.4-
• ENDOCARDIAL
O EPICARDIAL
0.2-
20
40
eo
80
MEAN CORONARY PRE8SURE (mm Hg)
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
FIGURE 3. Plot of endocardial (•) versus epicardial (O)
autoregulation. Flows are shown as functions of mean
coronary pressure. Resting flow was higher in the endocardium than in the epicardium. There was no significant
change in flow when mean pressure fell to 49 mm Hg.
Endocardial flow began to fall when mean coronary
pressure reached 37 mm Hg. In contrast, epicardial flow
remained unchanged until mean coronary pressure
reached 25 mm Hg. Values represent mean±SEM.
neously in the ascending aorta and matching left
ventricular end-diastolic pressure to equal the peak
atrial wave on the simultaneously measured left
atrial pressure. Variations in heart rate over the
experimental period were minimized by atrial pacing at a rate slightly more than the spontaneous
heart rate.
After allowing 30 minutes for the animal to adjust
to the laboratory, measurements of hemodynamics
and regional function were begun. Progressive reductions in distal circumflex pressure were produced
by the hydraulic occluder. Before injecting microspheres, each level of pressure reduction was held
constant for at least 5 minutes. Microsphere flow
measurements were performed under control circumstances and after coronary pressure reached
~50 mm Hg with maintained regional circumflex
function. Additional microspheres were injected
during steady-state reductions in coronary pressure, which resulted in measurable reductions in
regional circumflex function below resting values.
Regional perfusion was quantified with the reference withdrawal technique.19 Up to eightflowmeasurements were performed in individual animals
with 15-/im microspheres labeled with the following
gamma-emitting nuclides: l53Gd, "Co, 114In, ll3Sn,
'"Nb (New England Nuclear, Boston, Massachusetts) and 5lCr, "Sr, ^Sc (3M Incorporated). Microspheres were suspended in 10% dextran and 0.01%
Tween 80. The suspensions were placed in an
ultrasonicator for at least 15 minutes and vortex
agitated before injection. Approximately 2-4 x 106
microspheres were injected into the left atrium over
a 10-15-second period and flushed with warm arterial blood. Before microsphere injection, a reference arterial blood sample was started at a constant
rate (10 ml/min) from the descending aortic catheter
and continued for 2 minutes. Aortic, left ventricu-
lar, and coronary artery pressure as well as measurements of regional function were monitored
throughout the withdrawal period. Of 71 total microsphere injections, six were discarded because of
abruptly altered hemodynamic conditions after
microsphere injection. From two to seven injections were acceptable in each animal {n = l, one
dog; /i = 6, one dog; n = 5, three dogs; n = 4, six
dogs; n = 3, three dogs; n = 2, two dogs). After the
experiments were completed, the animals were killed
with potassium chloride overdose during deep barbiturate anesthesia. The hearts were removed and
placed in formalin for several days to facilitate
sectioning.
The left ventricle was sliced into four concentric
circumferential rings and the apex discarded. Each
ring was cut into eight wedges noting the anatomic
location of each wedge. Wedges were then divided
into four transmural layers of approximately equal
thickness. The papillary muscles were removed and
counted separately. The circumflex perfused core
included wedges from the base of the heart surrounding the crystals. Flow in samples adjacent to
this region were analyzed to ensure that all selected
myocardial samples were within the perfused core.
Each sample was weighed and activity quantified
with a Tracor-Northera sodium iodide detector.
Activity of each isotope was determined with a
least-squares radionuclide separation technique.20
Regional myocardial blood flow was calculated as
previously described.19
Normalization of regional flow. Because of both
temporal and spatial heterogeneity of microsphere
flow measurements,21-22 regional flow responses in the
circumflex (LC) region were normalized to flow
responses in the left anterior descending (LAD) or
control region. This LC/LAD flow ratio reflected the
relative reduction in circumflex flow under each experimental condition in a fashion similar to that previously used by other investigators.3-910 Under control
conditions, the ratio of LC/LAD flow in the endocardium varied among animals (mean, 1.01 ± 0.10; range,
0.86 to 1.21). Because this was believed to reflect
sampling variation and spatial heterogeneity, subsequent measurements of the LC/LAD flow ratio
were divided by the control ratio. This latter parameter was called the normalized LC/LAD endocardial flow ratio.
Flow measurements associated with a depression
in wall thickening and/or segment shortening below
90% of control values were grouped by the level of
normalized endocardial flow reduction (71-90%, 5170%, 31-50%, <30%). To avoid bias, each dog was
represented in a given group only once. Multiple
flow measurements that fell within the same group
for a given dog were averaged along with their
corresponding hemodynamic parameters to obtain
single values. In animals in which coronary pressure
was reduced but regional function unchanged (nonischemic points), closed-loop autoregulatory gain
was calculated as previously described.3
Canty
TABLE 1.
Nonischemic
n
C
S
P
71-90%
d = 9)
n
C
S
P
51-70%
d = 7)
n
C
S
P
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
31-50%
d = 5)
n
C
S
P
<30%
d = 5)
825
Microspbere Flow Experiments: Coronary Pressure and Regional Myocardlal Flow (Core Region)
Group
(% Control
endo flow)
(i=10)
Autoregulation in the Unanestbetized Dog
n
C
s
p
Normalized
LC/LAD
endo flow
10
1
1.03±0.15
NS
Endo
LC flow
(ml/min/g)
Epi
Endo/epi
8
84±10
40±7
<0.001
10
1.05 ±0.22
0.99±0.25
NS
10
0.84±0.27
0.77 ±0.23
NS
10
1.30 ±0.22
1.31 ±0.22
NS
7
73±8
22±2
<0.001
7
82±9
27±3
<0.001
9
1.02 ±0.23
0.84±0.20
<0.001
9
0.77 ±0.20
0.76±0.21
NS
9
1.34±0.15
1.14±0.19
<0.005
6
83±9
31±3
<0.001
5
73±7
18±2
<0.001
5
85±6
22±2
<0.001
7
0.94±0.21
0.57 ±0.16
<0.002
7
0.74 ±0.26
0.69±0.12
NS
7
1.30±0.19
0.81±0.14
<0.001
5
92±10
95±7
NS
5
87±14
28±1
<0.001
5
72 ±12
17±4
<0.002
5
85 ±15
20±l
<0.001
5
1.18±0.17
0.48 ±0.09
<0.001
5
0.90 ±0.26
0.79 ±0.11
NS
5
1.36 ±0.22
0.61 ±0.10
<0.001
5
89±13
89±12
NS
5
85±16
25±5
<0.002
5
70 ±13
15±4
<0.002
5
82 ±18
18±3
<0.002
5
1.18±0.19
0.31±0.15
<0.001
5
0.88±0.15
0.60 ±0.22
<0.018
1.35±O.O8
0.52±0.12
<0.001
PLC
PLC
PLC
(mm Hg)
mean
(mm Hg)
end djas
(mm Hg)
mean dias
(mm Hg)
10
87±10
91±7
NS
9
84±10
49±8
<0.001
8
73±9
33±6
<0.001
9
1
0.82 ±0.07
<0.001
8
86±10
90±6
NS
8
82±9
37±5
<0.001
7
1
0.61 ±0.06
<0.001
7
87±10
92±5
NS
5
1
0.43 ±0.05
<0.001
5
1
0.24 ±0.07
<0.001
Values are mean±SD.
LC, circumflex region; LAD, left anterior descending region; P^, mean aortic pressure; PLC, distal circumflex pressure; end dias, end
diastolic; mean dias, mean diastolic; Endo, endocardial flow; Epi, epicardial flow; Endo/epi, ratio of endocardial to epicardial flow;
normalized LC/LAD endo flow, normalized LC/LAD endocardial flow ratio for total perfused region calculated as described in the text;
i, number of measurements in each group; C, control; S, stenosis; p, statistical significance for two-tailed paired t test; NS, not
significant. Measurements have been grouped into five levels of pressure and flow reduction as described in the text.
Coronary Pressure-Function Relations
Relations between coronary pressure and circumflex function were examined in 16 dogs 15 ±7 days
after instrumentation. Control measurements were
performed after a 30-minute equilibration period,
after which coronary pressure was gradually reduced
in 2-5 mm Hg increments. At least 2 minutes were
allowed for coronary pressure and regional function
to equilibrate before data sampling. When function
began to fall, coronary pressure was reduced over
smaller increments (1-2 mm Hg) until percent wall
thickening approached zero, after which pressure
was restored. With this approach, 10-54 steadystate points were available for analysis in individual
animals.
Normalization of circumflex function. To compare relative reductions in circumflex function among
different animals, circumflex wall thickening (%WT)
and segment shortening (%SS) were normalized.
Preliminary studies indicated that regional dysfunction was never observed at a coronary pressure
more than 50 mm Hg under these hemodynamic
conditions. Therefore, all measurements of %WT
and %SS at mean coronary pressures more than 50
mm Hg were averaged to obtain the control value of
circumflex function for each individual study. Each
individual measurement of %WT and %SS in a
given experiment was then divided by this mean
value and expressed as a percent. Because of the
small variability in control function with repeated
measurements (coefficient of variation, 3-4%), significant reductions in function were defined as those
having both a mean coronary pressure below 50 mm
Hg as well as a value of normalized function less
than 90% of control. Pressure-function data fulfilling these criteria were fit to linear and quadratic
relations. Pressures corresponding to 100%, 50%,
and 0% of control function were calculated with the
coefficients obtained from the least-squares fit and
tabulated. To avoid errors related to extrapolation,
animals were excluded from regression analysis if
the degree of coronary pressure reduction did not
result in function failing below 50% of control
values. In addition, studies in which the coefficient
of variation for systemic hemodynamics and/or control function exceeded 10% were excluded because
hemodynamic instability could alter myocardial metabolic demand during the study period.
Data Analysis
Experimental data were recorded on an eightchannel Gould 2800 W recorder at a paper speed of
826
Circulation Research
Vol 63, No 4, October 1988
100-
*
(% Control)
¥3
60-
«
40-
•l
1
I-*H
r-OH
1 9
80NORMALIZED
ENDOCARDIAL
FLOW
^
*
8t
•
MEAN
O UCAN DIASTOLIC
20-
A END DIASTOLIC
0C>
20
40
1
1
1
80
1
1
SO
CORONARY PRESSURE (mm Hg)
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
FIGURE 4. Plot of relation between normalized endocardial flow and coronary pressure. Endocardial flow
remained constant until mean coronary pressure (•) fell
below 40 mm Hg. Corresponding values of mean diastolic
coronary pressure (O) and end-diastolic coronary pressure (A) were 30 and 25 mm Hg, respectively. Pressure
range over which endocardial flow approached zero
varied from ~20 mm Hg for mean coronary pressure to
~10 mm Hg for end-diastolic coronary pressure. Endocardial autoregulatory relation was shifted to the left
when flow was related to diastolic coronary pressure
indexes as opposed to mean coronary pressure. Values
are mean±SEM.
100 mm/sec. All data were digitized at a sampling
rate of 200 Hz with a Data Translation DT 2801-A
analog-to-digital convertor (Marlborough, Massachusetts) interfaced to an IBM PC AT computer.
Most experiments were digitized and analyzed online; however, in some, analog signals were recorded
on a Hewlett-Packard FM tape recorder (Palo Alto,
California) and digitized off-line. All data represent
averages of a 15-second sampling interval comprising at least 20 cardiac cycles.
Signals from the ultrasonic crystals were processed with a Triton Technology (San Diego, California) sonomicrometer. Left ventricular pressure
was differentiated with a filter cutoff of 100 Hz. The
first derivative of left ventricular pressure (dP/dt)
was used to determine end diastole (ED; onset of
positive dP/dt) and end systole (ES; 20 msec before
peak negative dP/dt). From these measurements,
the systolic excursion for wall thickness (A WT) and
segment length (A SL) were calculated as follows: A
WT = ESWT-EDWT and A SL = EDSL-ESSL.
Percent wall thickening and percent segment shortening were determined as follows: %WT = A WT/
EDWT and %SS = A SL/EDSL.
Constancy of systemic hemodynamics throughout each study was determined by measuring heart
rate, mean aortic pressure, systolic and enddiastolic left ventricular pressure, and the first derivative of left ventricular pressure (peak + dP/dt and
peak -dP/dt). Several indexes of coronary driving
pressure were calculated from the digitized data.
Mean coronary pressure was averaged over the
entire cardiac cycle. End-diastolic coronary pressure was taken at the onset of positive dP/dt. To
calculate mean diastolic coronary pressure, diastole
was defined as occurring between the point when
left ventricular pressure fell below coronary pressure until it exceeded it again during systole for
each cardiac cycle. Coronary pressure during this
period was then averaged.
Statistical Analysis
All values represent the mean ± SD unless otherwise indicated. Data for microsphere flow measurements were analyzed with one-way analysis of
variance. Significant differences between each level
of stenosis and the corresponding control values
were determined with a two-tailed paired t test. A
value of p<0.05 was considered significant.
Regression analyses of flow-function and pressurefunction relations were performed with a leastsquares linear fit. In some instances, second-order
polynomials and exponential least-squares fits similar to previous studies 89 were also determined.
Statistical significance between different regression
lines as well as between linear and quadratic fits
was determined with an analysis of covariance.23
Results
Analog recordings from an individual experiment
are illustrated in Figure 2. Regional circumflex wall
thickening and segment shortening remained con-
• CNDO
O EP1
NORMAUZED
SEGMENT
SHORTENING
(% Control)
A
1
T
NORMALIZED FLOW (% Control)
NORMALIZED
WALL
THICKENING
(% Control)
too
NORMALIZED FLOW ( « Control)
FIGURE 5. Plots of the relation between endocardial function and transmural flow at selected levels of coronary stenosis corresponding to data
obtained in microsphere flow experiments. Control points have been eliminated for clarity. — , Lines of identity
(i.e., a one-to-one relation between
reductions inflow andfunction). Reductions in both segment shortening and
wall thickening were closely related to
reductions in endocardial flow (•). In
contrast, changes in endocardial function were dissociated with epicardial
flow (o). Values are mean±SEM.
Canty
TABLE 2.
Nonischemic
(n=10)
Normalized
LC:LAD
endo flow
n
C
S
P
n
C
S
P
n
51-70%
(n = 7)
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
31-50%
(n = 5)
C
S
10
1
1.03 ±0.15
NS
9
1
0.82±0.07
<0.001
7
1
%LCSS
(%)
8
18.4±8.1
17.9±5.4
NS
8
24.9 ±8.0
22.7±8.2
<0.005
6
16.7 + 4.4
12.5±4.7
<0.01
6
27.4±7.4
17.8±7.3
<0.001
4
6
4
6
4
18.5±4.3
7.8±6.1
<0.007
10.35 ±0.65
10.21 ±0.81
NS
13.48 ±3.20
13.52 + 3.27
NS
2.80±0.61
1.78 ±0.63
<0.001
2.49±0.91
1.13±1.09
<0.001
5
9.15±1.29
9.06±1.15
NS
4
15.92±2.9O
16.31 ±3.09
NS
5
2.45 ±0.87
1.08 ±0.42
<0.017
2.04 ±0.79
0.78 ±0.97
<0.001
0.61 ±0.06
<0.001
n
5
5
4
C
S
1
27.0 ±9.1
11.8±4.5
<0.015
17.7±4.1
4.0±6
<0.007
71
c
s
p
0.43 ±0.05
<0.001
5
1
0.24 ±0.07
<0.001
LCED
WT(mm)
%LC WT
(%)
9
28.9±11.4
28.4±11.6
NS
P
P
<30%
d = 5)
827
Mkrosphere Flow Experiments: Regional LC Function
Group
(% Control
endo flow)
71-90%
(« = 9)
Autoregulation in the Unanesthetized Dog
9
9.89±1.0
10.02+1.1
<0.003
8
10.79±1.83
10.92 ±1.85
NS
LC ED SL
(mm)
8
14.59±2.54
14.5O±2.56
<0.003
6
12.56 ±2.73
12.28±3.01
NS
LCA WT
(mm)
LC ASL
(mm)
9
2.81 ±0.99
2.79±1.02
NS
8
2.69±0.94
2.58 ±0.85
NS
8
2.58±0.64
2.36±0.68
<0.01
2.02 ±0.355
1.44±0.309
<0.01
6
4
5
4
5
4
5
4
24.6 ±6.2
7.9±9.9
<0.014
15.3±3.1
-1.6 + 4.6
<0.009
9.23 ±1.30
9.11± 1.27
NS
16.00 ±1.66
16.40±1.90
NS
2.26±0.61
0.79 ±1.02
<0.006
2.47±0.68
-0.31 ±0.75
<0.016
Values are mean±SD.
Normalized LC: LAD endo flow, normalized endocardial flow for total perfused region (see text for details); %LC WT, circumflex wall
thickening (%); % LC SS, circumflex segment shortening (%); ED WT, end-diastolic wall thickness; ED SL, end-<liastolic segment
length; A WT, end-systolic minus end-diastolic wall thickness; A SL, end-diastolic minus end-systolic segment length; n, number of
measurements in each group; C, control; S, stenosis; p, statistical significance for two-tailed paired t test; NS, not significant.
Measurements have been grouped into five levels of pressure and flow reduction as described in the text.
stant over a wide range as coronary artery pressure
was reduced. In this animal, there were no changes
in function until coronary pressure fell below 35 mm
Hg (end-diastolic coronary pressure, 20 mm Hg).
Below this critical pressure, reductions in coronary
pressure were associated with pronounced reductions in circumflex function over a relatively narrow
pressure range. Despite the reductions in circumflex function, there were no major changes in systemic hemodynamics. As noted in the extreme
righthand panel of this figure, circumflex function
remained depressed below control after occlusion
release. In 10 animals, function was compared
before and after determination of the autoregulatory
relation. On restoration of coronary pressure after
an average of 41 ±18 (SD) minutes of ischemia,
circumflex wall thickening fell from 26.8 ±7.0% to
21.7±6.0% (p<0.002) and segment shortening fell
from 18.3 ±4.9% to 12.4±5.1% (/><0.001). There
were no significant changes in anterior descending
function. Thus, despite restoration of coronary pressure and presumably flow after the determination of
the autoregulatory relation, wall thickening was
depressed to 81 ± 13% of control and segment short-
ening to 66 ± 19% of control and recovered over the
subsequent 24-hour period.
Transmural Coronary Autoregulation
Under control conditions for all animals (n = 16),
heart rate averaged 106 ± 14 beats/min. Aortic pressure was 109 ±9 mm Hg systolic and 74 ±7 mm Hg
diastolic. Left ventricular end-diastolic pressure
averaged 5.7 ±2.9 mm Hg. Left ventricular dP/dt^,
was 3,061 ±606 mm Hg/sec and dP/dt^,, was
-2,556 ±351 mm Hg/sec. There were no significant
differences in systemic hemodynamics comparing
control with corresponding stenosis measurements
except for heart rate, which increased from 100 to
107 beats/min when circumflex flow fell to less than
30% of control (p<0.05). Circumflex zone wall
thickening averaged 26.7 ± 10.5%, and segment shortening averaged 17.8 ±4.8%. In the anterior descending or control zone, wall thickening averaged
28.4±9.8%, and segment shortening averaged
24.8 ±4.8%. Arterial blood gases at the time of
study were pH, 7.41 ±0.04; Pc^, 75 ±6 mm Hg; and
Pcch, 33 ±4 mm Hg. Hematocrit averaged 35 ±5%.
Regional autoregulatory relations at six selected
levels of endocardial flow reduction are illustrated
828
Circulation Research
Vol 63, No 4, October 1988
100-,
NORMALIZED
WALL
THICKENING
NORMALIZED
8EQMENT
SHORTENING
(% Control)
( « Control)
50
100
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
NORMALIZED ENDO FLOW (% Control)
in Figure 3. Corresponding hemodynamic and
regional flow measurements are summarized in Table
1. Under control conditions, circumflex endocardia]
flow for all animals averaged 1.06 ±0.22 ml/min/g
and epicardial flow averaged 0.82 ±0.24 ml/min/g
(p<0.001). There were no significant differences in
flow between the circumflex or the anterior descending region under resting conditions. When regional
function was maintained as coronary pressure was
reduced from 84 ± 10 to 49 ± 8 mm Hg, the reduction
in absolute flow was 6% for the endocardium and
8% for the epicardium (p = NS). With the changes
in absolute flow and mean coronary pressure for
these nonischemic points, closed-loop autoregulatory gain was 0.86 for the endocardium and 0.80 for
the epicardium. Endocardial flow began to fall significantly when coronary pressure reached 37 mm
Hg (p<0.001 vs. control). This was associated with
a corresponding fall in the endocardial-to-epicardial
flow ratio from 1.34±0.15 to 1.14±0.19 (p<0.005
vs. control). In contrast to the endocardium, autoregulation of epicardial flow was maintained until
mean coronary pressure fell to 25 mm Hg (p<0.02
vs. corresponding control measurements). There
were no significant changes in anterior descending
endocardial or epicardial flow during circumflex
ischemia.
Normalized circumflex endocardial flow for each
level of stenosis is plotted versus mean, mean
diastolic, and end-diastolic coronary pressure in
Figure 4. The endocardial autoregulatory relation
was qualitatively similar for each index of distal
coronary pressure. The autoregulatory relations
with end-diastolic and mean diastolic coronary pressure were, however, shifted to the left of the
FIGURE 6. Plots of the relation between
reductions in endocardial flow and both
wall thickening reductions on the left (•)
and segment shortening reductions on the
right (O)for individual flow-function points.
Flow values have been obtained from the
myocardium immediately surrounding the
endocardial crystals. Control microsphere
flow points have been eliminated for clarity. Dashed line on the left indicates the
linear relation between wall thickening
reductions (y) and endocardial flow (x):
y=0.96x+3.9,
r=0.90. Relation was
improved by a second-order polynomial
(y=-0.0103x2+2.34x-33.5,
r=0.93, solid
line on the left), but the deviation from the
linear fit was less than 10% during mild to
moderate ischemia (i.e., endocardial flow,
more than 30% of control). Solid line on the
right indicates the linear relation between
segment shortening reductions and endocardialflow: y=1.41x-4S, r=0.92. Slope of
the linear flow-function relation for segment
shortening was significantly different from
wall thickening (p<0.001).
autoregulatory relation for mean coronary pressure.
Furthermore, the range of coronary pressure over
which endocardial flow fell with diastolic pressure
indexes was narrower.
Flow-Function Relation During Autoregulation
Figure 5 contrasts endocardial and epicardial
variations in flow relative to endocardial function.
Table 2 summarizes measurements of regional function. Reductions in both segment shortening and
wall thickening were closely related to reductions in
endocardial flow. In contrast, epicardial flow
remained unchanged until the most severe level of
stenosis, despite large reductions in endocardial
function. These data indicate a close coupling of
endocardial flow with function measured by wall
thickening or segment shortening. Furthermore,
despite the fact that wall thickening is a transmural
measurement, its dissociation with epicardial flow
was similar to that of segment shortening.
Regression relations for both wall thickening and
segment shortening reductions are illustrated in
Figure 6. Reductions in function (percent of control) were correlated with flow (percent of control)
in the endocardia] sample containing the individual
crystals. Flow-function relations with the circumflex core region flows were not statistically different. The relation between wall thickening reductions (y) and endocardial flow (x) as a percent of
control was described by a linear equation:
y = 0.96x + 3.9, n = 60, r=0.90. A quadratic equation significantly improved the fit of the wall thickening data: y = - 0 . 0 1 0 3 J C 2 + 2 . 3 4 X - 3 3 . 5 , r = 0.93,
p<0.001 versus linear fit by analysis of covariance.
An exponential equation failed to fit the data and
Canty
Aatoregulation in the Unanesthetized Dog
829
37.9 ma Hg
100
4
NORMALIZED
SEGMENT
SHORTENING
(% Control)
At?
. 0/
•.if.) - -
100
NORMALIZED
WALL
THICKENMQ
(% Control)
100-
(
80-
iT
«.
•
eo
•f
40-
y - S.S2x -- 122
r o 0.S7
20-
-50
• *
0-
NORMALIZED
WALL THICKENING
(% Control)
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
FIGURE 7. Plot of the relation between steady-state
reductions in wall thickening and segment shortening for
animals in which both measurements were available. — ,
Line of identity. Reductions in segment shortening
exceeded wall thickening reductions at all levels of ischemia. Relation between them could be described as
follows: y=136x-40, n=380, r=0.95. Segment akinesis
occurred at a time when wall thickening persisted (29 J%
of initial values).
also failed to define the control point (100% function, 100% flow). Despite the improved quadratic fit
over the entire range of endocardial flow reduction,
the linear fit was within 10% of the quadratic fit at
levels of flow above 30%. Thus, from a practical
standpoint, reductions in endocardial flow in this
range were related to reductions in wall thickening
on a nearly one-to-one basis.
The relation between reductions in segment shortening (y) and endocardial flow (x) was also linear
n = 43, r = 0.92. The fit
(Figure 6): y=l.4lx-42J,
was not improved significantly by a quadratic equation, and an exponential equation did not fit the
data. As with wall thickening, flow-function relations based on core region flows were not statistically different. However, in comparison to wall
thickening, there were greater reductions in segment shortening at any given level of flow reduction
(p<0.001 for the slope of the linear fits).
Figure 7 depicts normalized wall thickening and
segment shortening reductions at comparable levels
of ischemia in dogs in which both pairs of crystals
were operational and appropriately aligned. The
relation between segment shortening (v) and wall
thickening (x) reductions as a percent of control was
described by the following equation: y = 1.36* - 40.1,
n = 380, r = 0.95. Akinesis of segment length (y = 0%)
occurred at a time when wall thickening persisted
(x = 29.5% of control). Thus, these data corroborate
the finding that reductions in segment shortening
exceed wall thickening reductions during steadystate circumflex ischemia.
Coronary Pressure-Function Relations
Coronary pressure-function relations were similar to endocardial autoregulatory relations based on
relative reductions in coronary flow. Figure 8 illustrates typical pressure-function relations obtained
in an individual animal for both wall thickening and
0
20
40
eo
SO
MEAN CORONARY PRES8URE (mm Hg)
38 1 mm Hg
1 ..
100-
• -
•^ • #
80-
NORMALIZED
SEQUENT
8H0RTENIN0
(% Control)
eo40y = e.70x - 156
r - 0.89
200-
0
20
40
eo
B0
MEAN CORONARY PRESSURE (mm Hg)
FIGURE 8. Plots of pressure-function relations for wall
thickening and segment shortening from an individual
animal. Wall thickening reductions occurred when coronary pressure reached 37.5 mm Hg. Segment shortening
reductions began when coronary pressure reached 38.1
mm Hg. Above the lower autoregulatory limit, there was
little variation in control function. Below the lower autoregulatory limit, reductions in function were linearly
related to coronary pressure.
segment shortening. In each animal, regional function remained constant over a wide range of coronary artery pressure. Below a critical pressure,
reflecting the lower limit of endocardial autoregulation, there was a linear fall in function with pressure. The reproducibility of the pressure-function
relation over time is shown in Figure 9. Here,
pressure-function relations were constructed on six
different days over a period of 4 weeks in the same
animal. Heart rate (118 ±5 beats/min) and systolic
pressure (113±5 mm Hg) remained within fairly
narrow limits over this time. The lowest pressure at
which regional function was maintained averaged
44.9±3.3 mm Hg for wall thickening and 45.5±3.2
mm Hg for segment shortening. The slope of the
pressure-function relation below this lower pressure limit averaged 4.81 ± 1.14%/mm Hg
(r=0.98 ± 0.02) for wall thickening and 8.83 ±2.20%/
mm Hg (r = 0.98 ±0.01) for segment shortening.
Thus, there was close correlation of pressure and
function throughout the duration of the study with a
lower autoregulatory limit that was reproducible
within narrow bounds.
Pressure-function regression relations for all
animals are summarized in Table 3, and corresponding pressure-function points are plotted in
830
Circulation Research Vol 63, No 4, October 1988
*.v • q^i
NORMALIZED
WALL
THICKENING
< * CoJiuofl
g0 -
40
DAY FOLLOWING
m8TRUU£NTATtON
'
MEAN CORONARY PRESSURE
•
•
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
*
*
(»» H«>
FIGURE 9. Plots of reproducibility of the pressurefunction relation in a single animal over 4 weeks.
Each line and symbol represents the pressurefunction relation on a given day. (See text for
discussion).
lift
NORMALIZED
SEGMENT
8HORTENINQ
(% Control)
/,
II
/ /
Jf/l
«0
DAY FOLLOWING
INSTRUMENTATION
—•
•
#-— 17
•
11
A
14
•
31
- - • - - J4
40
•«
«O
MEAN CORONARY PRESSURE <~i Ho)
Figure 10. The stability of control hemodynamics
was ascertained with the coefficient of variation
(Table 4). This ranged from 1% to 6% for the
different variables examined indicating that over
the time period of the pressure-function determination, global determinants of myocardial demand
were essentially constant. Among all animals, the
correlation coefficient for the regression of mean
coronary pressure and function averaged
0.94 ±0.05 for wall thickening reductions and
0.96 ±0.03 for segment shortening reductions. A
quadratic relation did not improve the fits and
frequently could not be extrapolated back to 100%
function. Wall thickening began to fall (100%
function) below a mean coronary pressure of
38.9 ±5.6 mm Hg, reaching a point of akinesis (0%
function) at a pressure of 21.1 ±3.5 mm Hg. Segment shortening began to fall at a mean coronary
pressure of 41.7±7.4 mm Hg (p<0.056 vs. wall
thickening) and reached a point of akinesis at a
mean coronary pressure of 26.2±3.8 mm Hg
(/?<0.006 vs. wall thickening). Using diastolic
indexes of coronary pressure, the pressurefunction relations were shifted to the left. Wall
thickening began to fall at a mean diastolic pressure of 29.8±5.6 mm Hg and at an end-diastolic
pressure of 22.0 ±2.9 mm Hg. Segment shortening
began to fall when mean diastolic pressure reached
31.9±8.1 mm Hg (p<0.062 vs. wall thickening)
and end-diastolic pressure reached 24.5 ±3.7 mm
Hg (p<0.062 vs. wall thickening). The slope of the
pressure-function relation increased when mean
diastolic pressure and end-diastolic pressure were
used. This reflected the fact that the pressure
range over which function fell from 100% to 0% of
control became narrower (—10 mm Hg for enddiastolic coronary pressure).
Discussion
This study indicates that the coronary pressurefunction relation can be used to define the lower
limit of endocardia] autoregulation. In contrast to
anesthetized animals, in which endocardial autoregulatory reserve is exhausted at coronary pressures
of 70 mm Hg, endocardial flow and function remain
constant in conscious animals until coronary pressure falls to 40 mm Hg. As outlined below, this
substantial shift in the lower autoregulatory limit is
difficult to explain on the basis of differences in
hemodynamic factors influencing myocardial metabolic demand or vasodilator reserve. This raises the
possibility that other factors may be responsible for
modulating coronary autoregulation in the openchest anesthetized animal.
Canty
TABLE 3.
Autoregulation in the Unanesthetized Dog
831
Pressure-Function Experiments: Coronary Pressure-Function Relations
Abscissa parameter
LC pressure (mm Hg)
Mean
Mean diastolic
End-diastolic
Normalized LC WT (« = 14)
50%
100%
38.9±5.6
29.8 ±5.6
22.0±2.9
3O.0±3.3
22.0±3.7
17.1±2.1
0%
Slope
Intercept
r
21.1 ±3.5
14.1 ±3.7
12.2 + 2.5
6.36±2.35
7.29±2.74
11.39±3.69
-137±6O
-106±51
-145±68
0.94 ±0.05
0.92 ±0.09
0.90±0.08
7.69±3.16
9.23±3.81
12.7±5.27
-203 ±88
-170±71
-199±94
0.96 ±0.03
0.95 ±0.05
0.91 ±0.05
Normalized LC SS (« = 12)
LC pressure (mm Hg)
Mean
Mean diastolic
End-diastolic
41.7±7.4
31.9±8.1
24.5 ±3.7
34.0±4.7*
25.3±5.1*
20.0 ±2.5*
26.2 ±3.8*
18.7 + 3.1*
15.5 + 2.0t
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
Values represent mean±SD.
100%, 50%, and 0% indicate values of coronary pressure (mm Hg) that correspond to three selected levels of regional function
reduction for normalized wall thickening (LC WT) and normalized segment shortening (LC SS). The slope (percent/mm Hg) and
intercept (%) represent the mean linear coefficients from the pressure-function relation for each abscissa parameter and r is the linear
correlation coefficient. A total of 22 pressure-function relations were suitable for analysis. In animals in which multiple pressure-function
relations were obtained, the points corresponding to 100%, 50%, and 0% function were determined for each relation and then averaged.
In addition, the slope and the intercept for each relation were averaged to obtain one value.
*/><0.01 vs. corresponding wall thickening values; tp<0.05 vs. corresponding wall thickening values.
Coronary Pressure-Function Relation
In view of the close relation between endocardial
flow and function during myocardial ischemia, it is
not surprising that the relation between endocardial
function and coronary pressure is similar to the
endocardial autoregulatory relation between flow
and pressure. This correspondence permits the
steady-state coronary pressure-function relation to
be used to characterize endocardial autoregulation
at any point in time, as well as before and after
specific interventions. Furthermore, under experimental circumstances in which oxygen extraction
increases as coronary pressure is reduced, the
pressure-function relation may allow more precise
interpretation of flow changes in terms of the adequacy of endocardial oxygen delivery.
Both endocardial segment shortening and transmural wall thickening remained constant over a
wide range of coronary pressure within individual
animals (coefficient of variation, 3-4%). Below the
lower autoregulatory limit, function fell in a linear
fashion with further reductions in coronary pressure (although a second-order term occasionally
improved the data fit, the second-order fit deviated
in an unphysiological fashion outside of the data
range). Reductions in function correlated best with
100NORMALIZED
WALL
THICKENING
«H
( I Control)
14 •
W
4 0 t o
ftO
100
100NORMALIZED
SEGMENT
SHORTEN! NO
SO-
to
40
K
M
100
40
00
10
100
100-
•W
jF
so-
•
(«CortroD
-
«
-ao-
•
10
40
00
MEAN
W
100
10
40
M
«0
MEAN CHA3TOUC
CORONARY PRESSURE
100
to
40
to
10
100
END DiASTOUC
( — Ho)
10. Plots of regional function versus mean, mean diastolic, and end-diastolic coronary pressure. Points
represent data from pressure-function relation determinations in 16 animals. Arrows indicate pressures corresponding to
the lower limit of autoregulation (function, 100% of control) based on the pressure-function relations summarized in
Table 3. Both wall thickening (upper panels) and segment shortening (lower panels) remained constant over a wide
pressure range. Below a critical pressure, large changes in function occurred over a narrow pressure range.
Pressure-function relations using diastolic pressure indexes (middle and right panels) were shifted to the left of the mean
coronary pressure-function relation (left panel).
FIGURE
832
Circulation Research
TABLE 4.
Vol 63, No 4, October 1988
Pressure-Function Experiments: Systemic Hemodynamks, Regional Function, and Coefficients of Variation
Mean
Heart rate (beats/min)
Aortic pressure (mm Hg)
Systolic
Diastolic
Mean
LVEDP (mm Hg)
LV dP/dt (mm Hg/sec)
108
113
74
91
6.7
Max
Min
3,095
-2,535
±SD
18
11
8
8
4.0
527
321
Coefficient of variation
(mean±SD)
2.7±2.3%
2.8+1.1%
4.8±2.0%
3.6±1.1%
*
5.9±2.6%
-6.3±1.9%
Regional Function (PLC>50 mm Hg)
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
LC WT % W T
(n=14) ED (mm)
ES (mm)
26.0
9.29
11.66
8.2
1.34
1.78
3.8 ±2.6%
0.9 + 0.5%
0.8±0.5%
LC SL
(n =11)
% SS
ED (mm)
ES(mm)
17.2
14.45
11.88
4.4
2.85
2.02
2.9±3.1%
0.4 ±0.4%
0.8 ±0.8%
LAD WT% WT
(n=14) ED (mm)
ES (mm)
25.5
8.78
11.03
9.8
2.09
2.60
3.3±2.1%
0.7 ±0.6%
0.8±0.5%
LAD SL % SS
(n =10) ED (mm)
ES(mm)
24.8
16.12
12.13
4.8
4.22
3.24
2.8±2.1%
1.0±0.7%
0.9 ±0.6%
Values are mean±SD.
LV, left ventricular; LVEDP, left ventricular end-diastolic pressure; LV dP/dt, first derivative of LV pressure;
LC, left circumflex; LAD, left anterior descending; WT, wall thickness; SL, segment length; %WT, wall thickening
(see text for calculation); %SS, segment shortening (see text for calculation); ED, end diastole; ES, end systole.
*For LVEDP, the standard deviation of LVEDP within an experiment averaged 1.8± 1.3 mm Hg.
mean coronary pressure (r = 0.94 for wall thickness,
r=0.96 for segment length). They also correlated
closely with mean diastolic and end-diastolic coronary pressure (r=0.90-0.95, Table 3). Although
reductions in segment shortening appeared to begin
at slightly higher coronary pressures than reductions in wall thickening (42 vs. 39 mm Hg), the level
of significance of the diflference was borderline
The temporal reproducibility of coronary autoregulation is currently unknown but has potential importance in interpreting studies examining changes in
coronary flow over time. In addition to variations in
heart rate, factors that alter systemic pressure,
oxygen carrying capacity, or neural tone are likely
to alter the overall autoregulatory relation. In the
present study, in which heart rate and systolic
pressure remained within narrow bounds, there was
minimal day-to-day variation in the coronary
pressure-function relation up to 5 weeks after surgical instrumentation. These results suggest fairly
constant characteristics of endocardial autoregulation over this time period.
Reproducibility of the pressure-function relation
in short-term studies is more problematic than on a
day-to-day basis. Pressure-function relations in the
present study were constructed with progressive
reductions in coronary pressure. Although the experimental procedure resulted in regional dysfunction
and ischemia for only short periods, regional function remained depressed below control by 20-30%
on restoration of normal coronary pressure. Regional
function returned to control levels within 24 hours,
and gross evidence of myocardial infarction was not
observed. The depression in function appeared similar to the reversible postischemic dysfunction
reported after brief periods of total coronary
occlusion24 or prolonged partial occlusion.25 This
finding highlights an important methodologic consideration in studying autoregulation and/or the
pressure-function relation on a repeated basis at a
single setting, that is, effects of interventions can
only be evaluated after allowing a sufficient time
period for endocardial function to recover after
preintervention data have been collected. This issue
may be germane to the interpretation of previous
studies examining the role of interventions during
autoregulation in anesthetized animals.1213-26-27 Characteristics of autoregulation in the postischemic or
"stunned" state may differ and contribute to the
general lack of reproducibility of autoregulation
observed in individual anesthetized animals.27
Two additional factors need to be addressed
when using the coronary pressure-function relation
Canty
ENDOCARDIAL
FLOW
(ml -mln*1- g"1)
• PfiESENT STUDY
OQALLAQHER «t | | . i t t O
20
40
90
80
100
60
80
100
30LC WALL
THICKENING
20
10I-QH
0-
0
•BH--
20
40
MEAN CORONARY PRE88URE (mm Hg)
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
FIGURE 11. Plots of comparison ofendocardial autoregulatory and pressure-function relations in pentobarbital
anesthetized dogs (Gallagher et aP) with those in the
present study. To compare the studies, absolute values of
endocardial flow and function have been used. Systemic
hemodynamic parameters were comparable as summarized in the text. Despite lower levels of resting endocardial flow in anesthetized animals (O), reductions in flow
and wall thickening occurred at higher coronary pressures than in unanesthetized animals (•). The shift to the
left of the endocardial autoregulatory relation and the
pressure-function relation in the awake animals suggests
that anesthesia may modulate coronary autoregulatory
responses independently of effects on heart rate, systolic
pressure, and contractility. Values are mean±SEM.
to study endocardial autoregulation. First, stable
systemic hemodynamics are required for constancy
of both myocardial metabolic demand and coronary
pressure at any fixed level of coronary stenosis. In
animals in which coefficients of variation for systemic hemodynamics exceeded 10%, correlation
coefficients between reductions in function and
coronary pressure were relatively poor (i.e., r<0.9).
Second, after changes in the stenosis severity, it is
important to delay observations until function and
pressure have reached a new steady state. In the
present study, function and pressure typically stabilized within 1 minute after an increase in degree of
occlusion. While flow in the epicardial artery was
not measured, a previous study in anesthetized
dogs from our laboratory indicated that the damped
flow oscillations after a sudden large step reduction
in coronary pressure stabilize within 90 seconds."
In response to gradually increasing stenosis and
coronary pressure reduction, these oscillations would
be expected to be attenuated.
Transmural Variations in Autoregulation
The results of this study are consonant with
previous studies in anesthetized animals demonstrating transmural variations in autoregulation and vul-
Autoregulation in the Unanesthetized Dog
833
nerability of the endocardium to ischemia.3-6 They
differ, however, in terms of the degree to which
flow, oxygen delivery, and function remain constant above the lower endocardial autoregulatory
pressure limit and the level of coronary artery
pressure at which autoregulation is exhausted (as
manifest by a decline in flow and function).
Investigators from several laboratories have
attempted to quantify the degree to which flow
remains constant during autoregulation by calculating the closed-loop autoregulatory gain.3 In the
present study, the small variations in absolute flow
(6-8%, p = NS) observed when coronary pressure
was reduced from 84 to 49 mm Hg resulted in values
of closed-loop gain (GJ of 0.86 for the endocardium
and 0.80 for the epicardium. Direct comparison
with results from anesthetized studies is difficult
because transmural variations in Gc have not been
reported. Available measurements based on epicardial artery flow changes vary markedly, depending
on the coronary pressure range examined. 162627
Dole and Nunno15 have also demonstrated that
increases in heart rate from 40 to 120 beats/min
increase Gc significantly. In general, however, most
anesthetized studies report considerably lower values of Gc than the present study over a similar
pressure range (i.e., Gc<0.5).
The comparison of the present values of Gc with
previous results is further complicated by the finding
that myocardial oxygen consumption decreases as
coronary pressure is reduced over the autoregulatory plateau in some anesthetized studies. 31316 This
variation was originally described by Gregg28 and
suggests that under some circumstances, coronary
pressure and/or flow appear to determine myocardial oxygen consumption. While oxygen consumption was not measured in the present study, the fact
that both endocardial function and oxygen delivery
remained unchanged over a wide range of distal
coronary pressure argues against such an effect.
Thus, the generally higher values of Gc in comparison to those in anesthetized animals may reflect
both a lower resting coronary venous Pc>2 (and less
reliance on enhanced oxygen extraction) as well as
the absence of the Gregg effect as coronary pressure is reduced.
The lower autoregulatory pressure limit has traditionally been believed to reflect the critical pressure at which endocardial vasodilator reserve is
exhausted. Reductions in the time available for
endocardial perfusion with increased rate and
increases in myocardial metabolic demand both
cause this limit to rise.3 Recent studies have also
demonstrated reductions in flow during autoregulation in the presence of pharmacologically recruitable vasodilator reserve in anesthetized animals.' '-14
Although the presence of pharmacological vasodilator reserve was not examined in the present
study, it is difficult to attribute the lower endocardial autoregulatory pressure limit in the present
study (40 vs. 70 mm Hg5 mean pressure and 25 vs.
834
Circulation Research Vol 63, No 4, October 1988
100-
NORMALIZED
WALL
THICKENING
1OO-
A
50-
50-
/
(% Control)
0-
LINEAR
SO
100
SO
100
-60-I
NORMALIZED ENDO FLOW (% Control)
100-i
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
- LINE OF IDENTITY
-PRE8ENT STUDY
50-
-GALLAGHER et al.. 1984
NORMALIZED
SEGMENT
SHORTENING
VATNER.1980
(% Control)
50
100
FIGURE 12. Plots of comparison of flowfunction relations from the present and
previous studies in unanesthetized animals. All relations have been adjusted to
reflect endocardial flow (x) and function
(y) as a percent of control. Panel A:
Compares the linear relations for wall
thickening from the present study
(y=0.96x+3.9, r=0.92) to that of Gallagher et al"> (y=1.25x-19, r=0.95).
Panel B: Compares the quadratic relations for wall thickening from the same
two studies
(y=-0.0103x2+2J4x-33J,
r=0.93, present study, andy=-0.0075x2
+2.15X-39, T=0.96, Gallagher et al'°).
Panel C: Compares the exponential segment shortening relation from Vatner9
(y=100-161.6e-"M7(IO0-x), r=0.92) to the
linear relation from the present study
(y=1.41x-42.7, r=0.92). (See text for
further discussion.)
- 5 0 -j
NORMALIZED ENDO FLOW (% Control)
50 mm Hg6 diastolic pressure) to differences in
resting flow and/or endocardial vasodilator reserve
between anesthetized and unanesthetized animals
for several reasons. First, despite somewhat higher
heart rates, values of resting endocardial flow in
previous anesthetized studies are similar to those in
the present study (~1 ml/min/g). This may reflect a
reduction in contractility in the anesthetized state
that counteracts the effects of increased rate on
myocardial demand. Second, differences in heart
rate between the present study and previous studies
are modest (—100 vs. —150 beats/min).56 Autoregulatory reductions in flow at pressures as high as 70
mm Hg are difficult to reconcile with measurements
of endocardial vasodilator reserve at a rate of 150
beats/min demonstrating flows of 4.1 ml/min/g at a
pressure of —90 mm Hg.29 In addition, a preliminary
report from our laboratory in awake animals has
demonstrated a lower autoregulatory limit of less
than 70 mm Hg in the face of an even higher heart
rate (200 beats/min) and higher resting flow (1.5 ml/
min/g).30 Finally, in contrast to endocardial flow,
epicardial flow appears to be independent of heart
rate during maximum vasodilation.29 Nevertheless,
the lower limit of epicardial autoregulation is also
higher in the anesthetized animal (-40 mm Hg3) as
opposed to the present study (25 mm Hg).
Studies of autoregulation in other vascular beds
have demonstrated attenuated autoregulation in
response to anesthetic agents.4 It is therefore possible that effects of anesthesia and/or acute surgical
instrumentation alter coronary autoregulatory
responses as they do the relation between endocardial flow and function.910 In this regard, the present
data are of particular interest in comparison to the
study of Gallagher et al7 where coronary pressure
and wall thickening were examined during myocardial ischemia in anesthetized open-chest dogs (Figure
11). The experimental preparation was similar to
the present study. Systolic pressure levels were
similar, but heart rate was slightly higher (130 vs.
100 beats/min). Despite lower control endocardial
flows in the anesthetized animals (0.62 vs. 1.05 ml/
min/g in the present study), reductions in flow
began at significantly higher coronary pressures. An
—20% reduction in endocardial flow occurred at a
mean coronary pressure of 55 versus 37 mm Hg in
the present study. Reductions in endocardial flow in
each study were associated with concomitant reductions in wall thickening, even though control values
of wall thickening were lower in the anesthetized
Canty
animals (19.3% vs. 28.9%). The shift to the left in
both the endocardia! autoregulatory and the pressurefunction relations in the present study supports the
hypothesis that factors other than systemic hemodynamics produce differences in coronary autoregulation during anesthesia. These differences warrant further investigation and may bear importantly
on the extrapolation of data obtained in anesthetized animals to humans.
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
Endocardial Flow-Function Relation
It has been appreciated since the classic study of
Tennant and Wiggers in 193531 that myocardial performance is closely coupled to coronary flow.
Although function falls when flow is reduced to a
critical level, reported myocardial flow-function relations exhibit considerable variability.7-10-32-34 While
some of this variation no doubt relates to anesthesia,
differences also exist between endocardial flowfunction relations obtained in conscious animals when
function is assessed by transmural wall thickening10
as opposed to segment shortening.9 Although beatto-beat variations in wall thickening and segment
shortening during temporary total coronary occlusion have been reported to follow a one-to-one
relation,35 correlative studies during steady-state
ischemia have not previously been reported. Based
on the present results, there appear to be quantitative differences between the magnitude of wall thickening and segment shortening reduction at comparable levels of ischemia. Flow-function relations from
the present study are compared with two previous
studies in conscious dogs in Figure 12.
Wall thickening and endocardial flow were closely
coupled in a fashion similar to that reported by
Gallagher et al.10 In each study, the flow-function
relation was mildly curvilinear and better described
by a second-order polynomial than a linear fit.
Although the linear relations differ modestly in
slope (Figure 12A), the curvilinear relations from
the two studies (Panel B) are quite similar. The
linear difference may reflect the fact that relatively
few measurements were performed in the present
study during severe ischemia. When endocardial
flow was more than 30% of control, deviations of
quadratic from linear fits were small (<10%). Thus,
during mild to moderate ischemia, relative reductions in normalized wall thickening provide a useful
index of the relative reduction in endocardial flow
on a nearly one-to-one basis.
Although segment shortening reductions were
also closely coupled to endocardial flow, the quantitative relation between the two parameters in the
present study differed from that previously reported
in conscious animals by Vatner9 (Figure 12C). In
the present study, the relative magnitude of segment shortening reduction exceeded that of endocardial flow reduction and was linear (Figures 5 and
6). The previous study found the flow-function
relation to be exponential, with a 50% reduction in
flow resulting in only a —20% reduction in segment
Autoregulation in the Unanesthetized Dog
835
shortening. This would seem to indicate a relative
insensitivity of segment shortening reductions during mild to moderate ischemia in contrast to the
present findings. While it is difficult to reconcile the
difference in terms of methodology, regional variations in the endocardial flow-function relation may
be responsible. Segment shortening in the present
study was measured in the posterobasal free wall
(distal circumflex artery) as opposed to the anteroapical free wall (anterior descending artery) in the
earlier study. Observations in anesthetized animals
indicate that relations between epicardial segment
shortening and flow during myocardial ischemia
differ between the two regions.33*36 An additional
difference between the studies may relate to the fact
that resting coronary flow levels were somewhat
higher in the earlier study, despite similar global
determinants of myocardial oxygen demand. Thus,
differences in autoregulatory gain between the two
studies could conceivably have altered the flowfunction relation, particularly if oxygen delivery at
reduced flow were maintained by increasing extraction in the earlier study.
Previous studies examining relations between
regional function and endocardial flow have not
compared simultaneous measurements of wall thickening and endocardial segment shortening during
steady-state myocardial ischemia. The present study
indicates that the magnitude of reduction in segment shortening is more than that of wall thickening
during steady-state circumflex ischemia. Akinesis
(i.e., absence of systolic shortening) occurred at a
time when wall thickening, while significantly
reduced, was still present (Figure 7). In terms of
detecting the onset of endocardial ischemia, segment shortening began to fall at a slightly higher
mean coronary pressure (42 vs. 39 mm Hg), but the
difference was of borderline significance (p<0.056).
Thus, while the magnitude of segment shortening
impairment is more than wall thickening at comparable levels of ischemia, there appears to be little
difference between the measurements in terms of
detecting the lower autoregulatory pressure limit
using the coronary pressure-function relation.
Summary
The close relation of both endocardial flow and
function with coronary pressure demonstrated in this
study provides a basis for the use of the coronary
pressure-function relation to characterize endocardial autoregulation on a repeated basis in the unanesthetized animal. The characteristics of transmural
coronary autoregulation in the unanesthetized animal appear to differ from the anesthetized animal in
several respects, with the mean coronary pressure at
which endocardial flow begins to fall being considerably lower (40 vs. 70 mm Hg). Although some of the
disparity between conscious and anesthetized animals may reflect variations in myocardial metabolic
demand and/or vasodilator reserve, factors modulating the coronary autoregulatory mechanism during
836.
Circulation Research
Vol 63, No 4, October 1988
anesthesia seem likely to be operative. Future studies with pressure-function as well as pressure-flow
relations should clarify the functional significance of
factors that can modulate endocardial autoregulation
under different experimental circumstances.
Acknowledgments
I would like to thank Dr. Francis J. Klocke for
his helpful suggestions and guidance throughout
this study. The technical assistance of Kathleen
Harris, Kathleen Weibel, Anne Coe, Joseph Giglia,
Verl Harbison, Amy Johnson, and Richard Kohlmeier and the secretarial assistance of Kyle Nugent
are greatly appreciated.
References
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
1. Mosher P, Ross J, McFate PA, Shaw RF: Control of
coronary blood flow by an autoregulatory mechanism. Circ
Res 1964; 14:250-259
2. Beme RM, Rubio R: Coronary circulation, in Beme RM,
Spcrelakis J, Geiger SR (eds): Handbook of Physiology: The
Cardiovascular System, Volume I. Baltimore, Williams and
Wilkins Co, 1979, pp 873-952
3. Dole, WP: Autoregulation of the coronary circulation. Prog
Cardiovasc Dis 1987^9:293-323
4. Johnson PC: Autoregulation of blood flow. Circ Res 1986;
59:483-495
5. Guyton RA, McClenathan JH, Newman GE, Michaelis LL:
Significance of subendocardial ST segment elevation caused by
coronary stenosis in the dog. Am J CardkA 1977;40:373-380
6. Rouleau J, Boerboom LE, Surjadhana A, Hoffman JIE: The
role of autoregulation and tissue diastolic pressures in the
transmural distribution of left ventricular blood flow in
anesthetized dogs. Circ Res 1979;45:804-815
7. Gallagher KP, KumadaT, Koziol JA, McKown MD, Kemper
WS, Ross J Jr: Significance of regional wall thickening
abnormalities relative to transmural myocardial perfusion in
anesthetized dogs. Circulation 1980;62:1266-1274
8. Weintraub WS, Hattori S, Agarwal JB, Bodenheimer MM,
Banka VS, Helfant RH: The relationship between myocardial
blood flow and contraction by myocardial layer in the canine
left ventricle during ischemia. Circ Res 1981 ;48:430-438
9. Vatner SF: Correlation between acute reductions in myocardial blood flow and function in conscious dogs. Circ Res
1980;47:201-207
10. Gallagher KP, Matsuzaki M, Koziol JA, Kemper WS, Ross
J Jr: Regional myocardial perfusion and wall thickening
during ischemia in conscious dogs. Am J Physiol 1984;247:
H727-H738
11. Canty JM Jr, Klocke FJ: Reduced regional myocardial
perfusion in the presence of pharmacologic vasodilator
reserve. Circulation 1985;71:370-377
12. Aversano T, Becker LC: Persistence of coronary vasodilator
reserve despite functionally significant flow reduction. Am J
Physiol 1985;248:H403-H411
13. Pantely GA, Bristow JD, Swenson LJ, Ladley HD, Johnson
WB, Anselone CG: Incomplete coronary vasodilation during
myocardial ischemia in swine. Am J Physiol 1985;249:
H638-H647
14. Grattan MT, Hanley FL, Stevens MB, Hoffman JIE: Transmural coronary flow reserve patterns in dogs. Am J Physiol
1986;250:H276-H283
15. Dole WP, Nunno DW: Myocardial oxygen tension determines the degree and pressure range of coronary autoregulation. Circ Res 1986^9:202-215
16. YonekuraS,WatanabeN,CaffrcyJL,GauglJF, Downey HF:
Mechanism of attenuated pressure-flow autoregulation in right
coronary circulation in dogs. Circ Res 1987 ;60:133-141
17. Canty JM, Klocke FJ: Reductions in regional myocardial
function at rest in conscious dogs with chronically reduced
regional coronary artery pressure. Circ Res 1987;60(suppl
II):U-107-II-116
18. Streeter DD Jr, Spotnitz HM, Patel DP, Ross J Jr, Sonnenblick EH: Fiber orientation in the canine left ventricle during
diastole and systole. Circ Res 1969^4:339-347
19. Heymann MA, Payne BD, Hoffman JIE, Rudolph AM:
Blood flow measurements with radionuclide-labeled particles. Prog Cardiovasc Dis 1977^0:55-79
20. Baer RW, Payne BD, Verrier ED, Vlahakes GJ, Molodowitch D, Uhlig PN, Hoffman JIE: Increased number of
myocardial blood flow measurements with radionuclidelabeled microspheres. Am J Physiol 1984;246:H418-H434
21. Falsetti HL, Carroll RJ, Marcus ML: Temporal heterogeneity of myocardial blood flow in anesthetized dogs. Circulation 1975 £2:848-853
22. King RB, Bassingthwaighte JB, Hales FRS, Rowell LB:
Stability of heterogeneity of myocardial blood flow in normal
awake baboons. Circ Res 1985^7:285-295
23. Snedecor GW, Cochran WG: Statistical Methods. Ames,
Iowa, Iowa State University Press, 1967
24. Heyndrickx GR, Baig H, Nellens P, Leusen I, Fishbein MC,
Vatner SF: Depression of regional blood flow and wall
thickening after brief coronary occlusions. Am J Physiol
1978;234:H653-H659
25. Matsuzaki M, Gallagher KP, Kemper WS, White F, Ross J
Jr: Sustained regional dysfunction produced by prolonged
coronary stenosis: Gradual recovery after reperfusion. Circulation 1983;68:170-182
26. Dole WP, Yamada N, Bishop VS, Olsson RA: Role of
adenosine in coronary blood flow regulation after reductions
in perfusion pressure. Circ Res 1985^6:517-524
27. Hanley FL, Grattan MT, Stevens MB, Hoffman JIE: Role of
adenosine in coronary autoregulation. Am J Physiol 1986;
250:H558-H566
28. Gregg DE: Effect of coronary perfusion pressure or coronary flow on oxygen usage of the myocardium. Circ Res
1963; 13:497-500
29. Bache RJ, Cobb FR: Effect of maximal coronary vasodilation on transmural myocardial perfusion during tachycardia
in the awake dog. Circ Res 1977;41:648-653
30. Canty JM, Giglia J, Kandath D, Klocke FJ: The effect of
tachycardia on endocardial autoregulation in the unanesthetized dog (abstract). Circulation 1987;76(suppl IV):IV-148
31. Tennant R, Wiggers CJ: The effect of coronary occlusion on
myocardial contraction. Am J Physiol 1935;112:351-361
32. Downey JM: Myocardial contractile force as a function of
coronary blood flow. Am J Physiol 1976^30:1-6
33. Gallagher KP, Stirling MC, Choy M, Szpunar CA, Gerren
RA, Botham MJ, Lemmer JH: Dissociation between epicardial and transmural function during acute myocardial ischemia. Circulation 1985;71:1279-1291
34. Roan PG, Buja LM, Izquierdo C, Hashimi H, Saffer S,
Willerson JT: Interrelationships between regional left ventricular function, coronary blood flow, and myocellular
necrosis during the initial 24 hours and 1 week after experimental coronary occlusion in awake, unsedated dogs. Circ
Res 1981;49:31-40
35. Sasayama S, Franklin D, Ross J Jr, Kemper WS, McKown
D: Dynamic changes in left ventricular wall thickness and
their use in analyzing cardiac function in the conscious dog.
Am J Cardiol 1976,38:870-879
36. Hattori S, Weintraub WS, Agarwal JB, Bodenheimer MM,
Banka VS, Helfant RH: Contrasting ischemic contraction
patterns by zone and layer in canine myocardium. Am J
Physiol 1982;243:H852-H855
KEY WORDS • pressure-function relation • flow-function
relation • coronary autoregulation • transmural flow •
regional myocardial function
Coronary pressure-function and steady-state pressure-flow relations during autoregulation
in the unanesthetized dog.
J M Canty, Jr
Downloaded from http://circres.ahajournals.org/ by guest on August 9, 2017
Circ Res. 1988;63:821-836
doi: 10.1161/01.RES.63.4.821
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1988 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/63/4/821
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/