1572
Dissociation Between Left Ventricular
Untwisting and Filling
Accentuation by Catecholamines
Frank E. Rademakers, MD; Maurice B. Buchalter, MB;
Walter J. Rogers, MS; Elias A. Zerhouni, MD;
Myron L. Weisfeldt, MD; James L. Weiss, MD; and Edward P. Shapiro, MD
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
Background. Efficient early diastolic filling is essential for normal cardiac function. Diastolic suction, as
evidenced by a decreasing left ventricular pressure during early filling, could result from restoring forces
(the release of potential energy stored during systolic deformation) dependent on myofilament relaxation.
Although these restoring forces have been envisioned within individual myofibers, recent studies suggest
that gross fiber rearrangement involving the connective tissue network occurs early in diastole. This may
lead to the release of potential energy stored during systole and suction-aided filling.
Methods and Results. To establish precisely the timing and extent of restoration of the systolic torsional
deformation of the left ventricle with respect to early filling at baseline and with enhanced relaxation, we
studied untwisting during control conditions and with catecholamine stimulation. Using noninvasive and
nondestructive magnetic resonance tagging, torsional deformation of the left ventricle was measured at
20-msec intervals in 10 open-chest, atrially paced dogs, starting at aortic valve closure. Eight equiangular
tags intersected the epicardium and endocardium in three short-axis imaging planes (base, mid, and
apex). From the intersection points, epicardial and endocardial circumferential chord and arc lengths
were measured and angular twist of mid and apical levels with respect to the base (maximal torsion and
its reversal, untwisting) was calculated. Echo-Doppler provided timing of aortic valve closure and of
mitral valve opening. Zero torsion was defined at end diastole. Torsion at the apical level reversed rapidly
between its maximum and the time immediately after mitral valve opening: from 7.0±5.80 to 3.2±5.40 and
12.0±8.50 to 6.9±7.80 (mean± SD, bothp<0.01) at the epicardium and endocardium, respectively. During
the same period, no significant circumferential segment length changes occurred. As expected, after mitral
valve opening, filling resulted in significant circumferential segment lengthening, whereas further reversal
of torsion was small and nonsignificant. During dobutamine infusion, torsion at end systole was greater
and reversal during isovolumic relaxation was much more rapid and greater in extent (p<0.01). Torsion
reversed from 11.5±4.3' to 5.7±4.80 and 17.4±6.40 to 6.9±7.70 at epicardium and endocardium.
Concluions. Untwisting occurs principally during isovolumic relaxation before filling and is markedly
enhanced in speed and magnitude by catecholamines. This partial return of the left ventricle to its preejection
configuration before mitral valve opening could represent an important mechanism for the release of potential
energy stored in elastic elements during the systolic deformation. These myocardial restoring forces would be
markedly enhanced by physiological changes consequent to catecholamines such as during exercise, offsetting
the concomitant shortening of the filling period. (Circulton 1992;85:1572-1581)
KEY WoRDs * left ventricular filling * dobutamine * regional function * biomechanics
No ormal left ventricular (LV) filling is dependent
on myofilament relaxation, left atrial pressure,
atrial contraction, and passive LV compliance.1-3 When one of these functions is impaired or
when the filling period is shortened and stroke volume is
increased during physiological stress,45 maintaining cardiac output is more and more dependent on early rapid
filling. Ventricular suction, the intrinsic capability of the
From the Division of Cardiology, Johns Hopkins University
School of Medicine and the Francis Scott Key Medical Center,
Baltimore, Md.
Presented in part at the American Heart Association 62nd
Scientific Sessions, November 1989, New Orleans, La.
Supported in part by National Heart, Lung, and Blood Institute
Ischemic Heart Disease Specialized Center of Research grant
HL-17655-16, RO1-HL-43722 (J.L.W.), and American Heart
left ventricle to create an atrioventricular gradient, is an
additional means of achieving efficient early filling.6
Suction is manifested by a decreasing LV pressure
during early filling; its existence has been proven in
experimental models in which mitral valve opening
(MVO) is delayed and negative transmural pressures
are observed.7,8 The underlying mechanism for LV
suction has not been identified. Restoring forces, providing elastic recoil after ventricular contraction, could
Association grant 891230 (E.P.S.). F.E.R. is a research fellow at
the Johns Hopkins University supported by a NATO fellowship.
Address for correspondence: Edward P. Shapiro, MD, Division
of Cardiology, Francis Scott Key Medical Center, 4940 Eastern
Avenue, Baltimore, MD 21224.
Received September 24, 1990; revision accepted December 3,
1991.
Rademakers et al Left Ventricular Untwisting
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
cause ventricular suction during normal conditions and
could be enhanced by catecholamines released during
stress, but their exact nature is unclear.9
Torsional deformation of the left ventricle or twist
during systole is one possible mechanism by which
potential energy could be stored during ejection and
then released during diastole to create suction and fast
early filling. This could be achieved by straining the
intercellular collagen matrix,'0 which on relaxation
would release its stored energy and restore the diastolic
configuration of the ventricle. Although the torsional
deformation of the left ventricle during systole has been
studied extensively with several techniques,10-'9 the
timing and extent of the recoil or untwisting has only
recently been a focus of attention. Beyar et all5 measured rotational deformation throughout the cardiac
cycle using radiopaque markers implanted in the myocardial midwall and found that a substantial release of
torsion occurred during early diastole. In the present
study, we measured the reversal of epicardial and
endocardial torsion of the canine left ventricle and the
changes in circumferential segment length, starting at
aortic valve closure, encompassing isovolumic relaxation, MVO, and early filling using nuclear magnetic
imaging with tags, noninvasive markers of the myocardium.20 Valve opening and closure were timed by echoDoppler measurements. To study the effect of a shortened filling period and increased contractility and
relaxation, the same measurements were also obtained
during dobutamine infusion.
We tested the hypotheses that untwisting occurs
principally before filling, when LV cavity size is constrained and suction could develop, and that its extent
during isovolumic relaxation is enhanced by dobutamine
infusion. Such evidence would support the importance
of untwisting as the release mechanism of restoring
forces and the likelihood that untwisting plays a critical
role in efficient early diastolic filling.
Methods
Nuclear magnetic imaging with tagging was used to
mark sites in the myocardium noninvasively and to
image the left ventricle at different times during diastole.20 These markers or tags provide the means to
calculate torsion and circumferential segment length
from the same image acquisition and to measure their
changes over time during the cardiac cycle.1",2'
Magnetic Resonance Image Myocardial Tagging
The technique of myocardial tagging used in these
experiments has been reported in detail in previous
studies from this laboratory."120 Tags are noninvasive
markers that appear as dark stripes on the white
myocardium. When placed at end diastole and imaged
at other times in the cardiac cycle, they show the
displacement and deformation of the myocardium on
which they were inscribed. Tagging is achieved by
selective radiofrequency saturation of thin planes intersecting the myocardium in a pattern predefined by the
operator. In the tagged regions, tissue protons are in a
different state of magnetization from those in nontagged regions. This difference in magnetization persists
for a time mainly dependent on the longitudinal relaxation time (T1) of the myocardium, which for magnetic
1573
field strengths of 0.3-0.6 T is about 400-500 msec.
During that time, a standard spin-echo magnetic resonance image acquired orthogonally to the tagging plane
will demonstrate a difference in the signal intensity
between the tagged and nontagged regions.
Experimental Protocol
Ten adult mongrel dogs (20-22 kg) were studied. The
dogs were anesthetized with sodium pentobarbital
25-35 mg/kg and ventilated with a Harvard ventilator.
Additional sodium pentobarbital was administered during the study to obtain a steady anesthesia. All dogs had
jugular venous lines for fluid administration and a
carotid intra-arterial line for pressure monitoring. A left
thoracotomy was performed, and the pericardium was
opened anteriorly. A pacing lead was sutured to the
right atrial appendage, and a Gould heart sound sensor
was positioned near the aortic arch with its wires
extending through the anterior chest wall to a Gould
eight-channel ink recorder system. The heart was replaced in the pericardial sac, which was left open, and
the chest wall was approximated. All dogs were atrially
paced at an RR interval of 400 msec. Aortic pressure
tracings were recorded intermittently between actual
image acquisitions to monitor hemodynamic stability.
The dogs were studied under control conditions and
during increased contractility and intrinsic relaxation
induced by dobutamine infusion. The dose of dobutamine was adjusted to increase mean arterial blood
pressure by 10%. This intervention increased intrinsic
heart rate by 20-30 beats per minute, and atrial pacing
was resumed at least 5-10 beats per minute higher,
depending on the variability of the intrinsic rate. A
stable and regular heart rate, required for high-quality
images, could be obtained only in five of the 10 dogs. In
one of the five excluded dogs, this was due to excessive
bleeding after dobutamine infusion. The five remaining
dogs were paced at heart rates between 190 and 205
beats per minute.
To assure that opening the chest does not alter the
overall measurements and conclusions, five additional
dogs were studied with a closed-chest preparation
using Doppler to time aortic closure and a transvenous
pacemaker (repositioned when necessary to continue
capturing).
Imaging Protocol
Imaging parameters. Images were acquired on a
0.38-T iron-core resistive magnet (Resonex, RX4000)
using a spin-echo sequence modified as above to include
tagging, with time to echo (TE)=30 msec, repetition
time (TR)=RR interval, using 128 phase and frequency
encoding steps and averaging two (or rarely four)
excitations. The image slice thickness was 10 mm in all
cases. The tag width was 3.5 mm.
Location of imaging planes. After positioning the dog
in the magnet, a series of scout images was acquired so
that the orientation of the heart in each animal could be
assessed and short-axis cardiac images could be obtained. From the axial image that traversed the largest
left ventricular cavity area, a new image plane that
passed through the left ventricular apex and was parallel to the septum was selected by using visual estimates.
This defined a left ventricular long axis. Perpendicular
1574
Circulation Vol 85, No 4 April 1992
TAG PLANE
D
OLE
SYSTLE
FIGURE 1. Schematic diagram of tag planes intersecting the
long axis of the left ventricle. The relation of only one image
plane and two tag planes is depicted (left). At end diastole, the
tag planes form a star pattern in each imaging plane; tag
planes are seen as radial lines (center). At end systole, the
myocardial tags twist and bend, reflecting myocardial deformation (right).
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
to this long axis, five parallel short-axis LV planes were
selected and imaged using a multislice sequence. From
these five short-axis images three were chosen, located at
the midventricular level and 15 mm to either side of this
middle image. The orientations of four equiangular tag
planes, centered at the visual center of the LV cavity on
each image plane, were then obtained. One tag plane was
chosen to be equivalent to the echocardiographic apical
four-chamber projection. As the four tag planes were
equiangular, the positions of the other tags were therefore
standardized from day to day. After this series of scouts,
images of the three LV short-axis planes, with 3.5-mmwide tags applied at end diastole (the first negative deflection of the R wave of the electrocardiogram), could then
be acquired at the specified times (Figure 1).
Timing of image acquisition. Gated images were acquired by entering the desired delay after the electrocardiographic QRS. A Hewlett-Packard 77020A echoDoppler system with a 2.5-MHz duplex probe was used
to determine the time of aortic valve closure and MVO
with respect to the QRS by identifying the mitral and
aortic flow velocity profiles and valve clicks with continuous-wave Doppler. Control echo-Doppler measurements were obtained at the beginning of each study.
During control conditions, imaging commenced at the
time of aortic valve closure and was repeated at 20-msec
intervals. For the dobutamine intervention, the echoDoppler was performed at the end of the imaging
sequence because removing the dogs from the magnet
to use the Doppler between control and dobutamine
imaging would have altered the slice orientation. Because the systolic period is shortened by dobutamine,
the time of aortic valve closure with respect to the QRS
was determined by a phonocardiographically determined second heart sound, which can be done inside the
magnet. The value was rounded down to the nearest 10
msec. The exact timing of aortic valve closure and MVO
was determined by echo-Doppler after removal of the
dog from the magnet.
Imaging sequences for obtaining 20-msec time resolution. The total time required for each image frame
including radio frequency pulses and gradient generation, was 56 msec. It was therefore impossible to acquire
images 20 msec apart in one sequence. To obtain the
desired time resolution, three interleaved sequences
separated by a 20-msec delay were used, each sequence
acquiring three images at 60-msec intervals. Specifically, during any one image acquisition the basal, mid,
and apical levels were acquired at times separated by 60
msec. The locations of the three imaging planes were
then rotated for the next two acquisitions. This entire
sequence was then twice repeated, each time adding an
additional 20-msec delay after the QRS. A complete set
of images therefore required nine acquisitions.
Because images were acquired during nine acquisitions and each takes 512 cardiac cycles for a total
duration of about 40 minutes, constancy of heart rate
and hemodynamic stability were obtained by pacing and
monitored by arterial pressure readings.
Analysis
The three image planes used in this study correspond to
three short-axis levels of the left ventricle, i.e., basal, mid,
and apical. The four tag planes, perpendicular to these
image planes, cross at the center of the left ventricular
cavity and form a star pattern on each image level. At a
specific image level, tags (being radial lines) intersect with
both endocardium and epicardium. These intersection
points, eight for both epicardium and endocardium at
each image level, were identified as X, Y coordinates on a
Dell 310 computer-based contouring system (Cardiology
Image Processing System, JHU). Endocardial and epicardial contours were also created with an Akima algorithm,22 which uses the intersection points of the tag lines
with endocardium and epicardium to create smooth contours. These contours were subsequently manually
adjusted to fit the actual epicardial and endocardial surfaces optimaLly. From these data, torsion angles and
circumferential segment lengths were calculated.
Calculation of Torsion Angle
The technique of analysis used in these experiments
is similar to the technique previously described in
human studies."1 Considering the epicardial and endocardial points separately, the centroid of these points
was calculated and the slopes of the lines connecting
each tag intersection point and the corresponding centroid were calculated and expressed as an angle. This
was done for the epicardial and endocardial intersection
points of the basal, mid, and apical images at each time
point.
To subtract out the effect of rigid body rotation and
whole-heart translation, the image of the basal plane at
each time point was used as a reference. The difference
in the slope of the line connecting a tag point on the
basal plane with its centroid and the line connecting the
corresponding tag point to its centroid on the more
apical plane was calculated, and its arctangent was
called the torsion angle (Figure 2). This angle is the
difference, at each time point, between the position of a
basal tag point and the corresponding tag point on the
more apical image level, expressed as an angle of
rotation. Because there was no difference in the relative
positions of these two points at end diastole (the tags
being inserted as a plane in all slices simultaneously),
this angle represents the rotation of one point with
respect to the other one between end diastole and the
time of the measurement. Zero torsion is the reference
at end diastole, the time of tag insertion. Increasing
angle (positive sign) indicates counterclockwise rotation
of a point on a more apical slice when viewed looking
from the apex toward the base.
Rademakers et al Left Ventricular Untwisting
TORSION ANGLE
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
FIGURE 2. Schematic diagram shows basal and a more
apical image plane during the cardiac cycle. Only one myocardial surface (epicardial or endocardial) is shown for
simplicity. On the basal slice, the line connecting the tag
intersection point with the centroid is shown. On the more
apical slice, the position of the corresponding tag intersection
point, connected to the centroid, is shown by the solid line.
The interrupted line represents the projection of the basal line
on the more apical slice. The torsion angle is defined as the
angle between the tag line of that slice and the projection of the
basal tag line on that slice. h, Distance between the two slices.
Intraobserver variability was tested by comparing the
calculated angles from two analyses by the same observer. The mean differences (±SD) in the angles were
0.77+0.65° and 1.52±+1.21° for the epicardium and
endocardium, respectively.
Cakculation of Segment Length
Regional circumferential shortening was calculated
in two ways. First, using the coordinates of the intersections of the tags with epicardium and endocardium,
epicardial and endocardial chord lengths were calculated. Second, using the epicardial and endocardial
contours, the arc lengths between each adjacent pair of
tag intersection points on the epicardial and endocardial surfaces were calculated (Figure 3). Both sets of
measurements were then expressed as percent change
from the first (end-systolic) time point to every other
time point. Zero lengthening is the reference at end
systole. Positive lengthening values represent increasing
segment lengths with respect to the length at end
systole. No significant differences existed between the
two methods in these normal ventricles. Arc length
changes are reported.
Dissociation of Untwisting and Circumferential
Segment Lengthening From Filling
The dissociation of untwisting from filling is expressed
as the percentage of untwisting that occurred before the
onset of filling, that is, MVO as measured by echoDoppler. MVO usually did not exactly coincide with an
1575
FIGURE 3. Schematic diagram shows position and numbering of tag intersection points on a short-axis image plane;
section of the left and right ventricles perpendicular to the left
ventricular long axis is shown. Epicardial (EPI) and endocardial (ENDO) arc lengths are demonstrated for segment 6-7
by the thicker lines.
image time; therefore, the linearly interpolated value
between the two adjacent imaging time points was used
for the percent calculation. The dissociation of untwisting from filling is defined as the peak torsion (at any
time) minus the torsion at the time of MVO (interpolated linearly between the two imaged times bracketing
MVO) divided by peak torsion. Similarly, the dissociation
of segment lengthening from filling is defined as the
percentage circumferential segment lengthening at the
time of MVO (interpolated linearly between the two
times bracketing MVO) minus the smallest or most
negative percentage segment lengthening (at any time
during isovolumic relaxation). However, actual data
points for torsion and segment length given in the
"Results" section and in the tables refer to the value of
torsion or segment length closest to MVO rather than an
interpolation, because the standard errors for interpolations are not meaningful.
Statistical Analysis
The results from each time point were analyzed
separately. Epicardial results were analyzed separately
from endocardial results. Differences between the time
points were tested using an analysis of variance with
repeated measures. If a significant difference was found,
subgroup analysis using Scheffe and Student-NewmanKeuls techniques was performed. Results are presented
in tabular and graphical form as mean torsion angles
and percent segment lengths at either epicardium or
endocardium at one image level. For control, the results
of 10 dogs are reported. For dobutamine, the results of
five dogs are presented. Student's t test for paired
observations was used to compare results between the
control and dobutamine conditions.
Results
Control Conditions
A consistent dissociation in time was observed between untwisting and filling. Approximately 50% of
untwisting occurred before MVO, whereas circumferential segment lengthening occurred nearly entirely
after MVO, during filling. This is demonstrated in
Figure 4, showing apical epicardial and endocardial
Circulation Vol 85, No 4 April 1992
1576
CONTROL
14-
-40
TORSION EPI
-4P- TORSION ENDO
-|-
12-30=
0-
10-
-El-
LENGTH
EPI
ENDO
L - eENGTH
ro
-
-20
z
uu
Ul
0
I i
UU
z
D
4-
0
'U
-0
2z-
in
0
150
200
l l
AoVC
240
t
280
320
360
TME (MSEC)
MVO
Graph shows apical segment length and torsion
measurements during control conditions. Horizontal axs shows
time afterpeak of QRS. Only the period of interest, ie., diastole
from just before aortic valve closure, shown. Isovolumic
relaxatin
ending with mitral valve opening is shaded.
period
starts
Filling
after mitral valve opening and continues until the
next QPS (400 msec, not shown). Vertical axis shows percent
segment length on the right and torsion in degrees on the left.
Filled symbols refer to torsion and open symbols to segment
length. Squares are epicardial (EPI) measurements; circles are
endocardial (ENDO) measurements. Ten dogs were studied;
each apical slice contained eight segments. AoVC, aortic valve
FiGuRE
4.
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
closure; MVO, mitral valve opening.
torsion angle and segment length changes versus
time for 10 control dogs. Apical torsion decreased
and
to 3.2±5.40
(untwisting) from
12.0±8.50 to 6.9±7.80 (±SD,p<0.01) at epicardium and
mean
7.0+5.8°
TABLE 1.
Mid
(degrees)
Apex
5.8±5.0+
1.5±3.33(p<0.0.)
(p<0.01)
p<0.01),
p<O.Ol).
As previously reported, in humans,8 torsion of the apex
is larger than torsion of the mid, and endocardial torsion
is larger than epicardial torsion at each level. This is
related to differences in distance from the basal reference
slice (h in Figure 2) and in radius from the cavity centroid.
Dobutamine Infusion
This intervention resulted in a decrease of cycle
length from 400±3 to 300±26 msec and an increase of
Measurements for Torsion and Segment Length in Control Studies
Mean torsion
Control
(p<0.01)
endocardium, respectively. This corresponds to 54%
and 44% untwisting before MVO, respectively. During
the same interval, apical circumferential segment length
remained essentially constant (p=NS) from
-1.7±1.3% to 2.0±1.5% and -0.3±1.1% to 3.0±2.1%
at epicardium and endocardium, a 2.8% and 3.9%
segment
length change before MVO, respectively. The
the endocardium between 200 and 220
lengthreached
change atborderline
msec
significance (p=0.05). After
MVO, torsion at the apex further decreased nonsignifmsec, respecicantly whereas
(p=0.06) to 2.20 and 3.80 at 340
lengthening occurred at a fast
tively,
segment
rate to maximum levels (15.2% and 30.5% at epicardium and endocardium, respectively; p<0.01). Values
for all time points are reported in Table 1.
At the midventricular level, torsion declined from
3.1±2.90 to
and from
to
at epicardium and endocardium,
4.2±5.60
This corresponds to 52% and 31% untwistrespectively.
before MVO. Changes in segment length during the
ing
same time period were nonsignificant (3.0% and 1.7%
at epicardium and endocardium, respectively). After
MVO, segment lengthening was significant for epicarwhereas the
dium and endocardium (both
torsional change was significant only at the endocardium for the first time interval (260 versus 280 msec,
p<0.05) and at 320 msec (versus 300 and 340 msec,
Base
(%)
Mean segment length
Mid
Apex
Time (mse)
Epi Endo Epi Endo Epi Endo Epi Endo Epi0.0 Endo
0.0
0.0
0.0
0.0
0.0
6.2
10.3
5.5
180 AVC
3.1
...
...3.1 ...-1.2
(6.5)
(4.8)
(4.5)
(2.9)
-0.3
-1.7
0.2
0.1
12.0
7.0
200
5.8
2.8
(1.1)
(1.3)
(1.2)
(1.8)
(1.9)
(1.7)
(8.5)
(5.8)
(5.0)
(3.0)
6.2
1.3
2.0
-0.8
3.1
0.0
5.0
9.1
220
2.7
5.8
(3.0)
(1.5)
(1.6)
(2.5)
(1.5)
(9.1)
(5.5)
(5.3)
(4.1)
3.0
2.0
-1.4
1.7
0.1
-0.9
3.2
4.2
6.9
240 MVO
1.5
(2.1)
(1.5)
(3.2)
(1.5)
(2.3)
(2.1)
(7.8)
(5.4)
(5.6)
(3.3)
7.0
0.9
10.2
2.4
4.0
0.7
5.4
260
1.4
3.4
3.0
(1.8)
(4.8)
(1.9)
(2.4)
(3.1)
(8.3)
(3.3)
(4.6)
(6.5)
12.4
3.3
12.4
3.9
9.1
0.6
280
5.7
1.5
2.9
2.1
(3.6)
(2.0)
(4.9)
(2.2)
(2.6) (5.8)
(3.5) (5.3) (6.3) (7.6)
17.6
6.5
10.6
7.7
10.1
4.1
6.4
300
1.2
1.8
2.9
(4.2)
(3.5)
(4.6) (5.6)
(2.6) (6.9)
(3.2) (5.5) (6.3) (8.7)
26.2
11.1
18.5
7.5
18.5
9.1
3.8
0.9
320
2.3
2.8
(5.8)
(3.8)
(6.8)
(5.4)
(6.7)
(2.8)
(2.9) (4.9) (5.9) (8.1)
30.5
15.2
29.7
10.0
10.2 23.2
3.8
340
0.6
0.7
2.2
(7.1)
(6.0)
(6.4)
(2.7)
(5.6)
(3.5)
(7.5)
(7.3)
and mitral valve
(AVC)
closure
valve
aortic
of
occurrence
Time: timing of image acquisition after the peak of QRS. As a reference,
at the mid and apical
(Endo)
endocardium
and
(Epi)
epicardium
the
are shown. Torsion: torsion expressed in degrees ofsegment
(MVO)
opening
length expressed as percent lengthening at each time point with
level, referenced to the basal level. Segment length: circumferential
(1.7)
(3.1)
(4.6)
to the first (end-systolic) time point. Results for epicardium and endocardium at base, mid, and apex.
respect
Values are mean+SD. n=10 dogs.
(6.9)
Rademakers et al Left Ventricular Untwisting
DOBUTAMINE
20.
40
_*J TORSION ENDO
1816-
l
r- -_.. c /
/
30
F£_ LENT EPI
WZo LENGTH ENDO
J
z14-
i
/
012CD
/
10-
/
20
p
w
u,
..
a
0
in
- -
Z
-10 i-
o 8
92
aortic valve closure and from 243 to 195 msec for MVO,
shortening the isovolumic period from 65 to 37 msec.
The dissociation between untwisting and filling observed during control conditions is further accentuated by this intervention. Figure 5 shows the results
for apical torsion and segment length during dobutamine infusion. All values are reported in Table 2. At
the apical level, torsion decreased from 11.5±4.30 to
5.7±4.80 and from 17.4±6.4° to 6.9±7.7° at epicardium
and endocardium, respectively, corresponding to a
58% and 67% untwisting before MVO. The reduction
in torsion for the midventricular level is from 5.0± 1.90
to 1.4±1.50 and from 6.8±2.7° to 2.2±4.1° at epicardium and endocardium, respectively, corresponding to
a 73% and 65% untwisting before MVO. No further
significant change in torsion angle occurred after
MVO. Both the absolute amount of torsion that had
dissipated by the onset of filling and the extent of
untwisting before MVO were significantly increased
after dobutamine stimulation compared with control
(p<0.01 for both parameters). By contrast, during
isovolumic relaxation, segment lengthening was significant only between 170 and 190 msec at the apical
endocardium (2.3% and 11.6% lengthening at epicardium and endocardium, respectively). At the epicardium and endocardium of the mid level, shortening
continued until 190 msec and was immediately followed by significant lengthening (10.5% and 16.8%,
respectively).
The absolute values of torsion were significantly higher
during dobutamine infusion compared with control: Maximal torsion at the apical epicardium and endocardium,
dobutamine versus control, was 11.5 ±4.3 versus 7.0±5.8°
and 17.4±6.4° versus 12.0+8.50 (±SD), respectively. As
expected, circumferential lengthening increased with dobutamine (p<0.01, paired results); shortening of the base
increased more than that of the midventricle or apex
during the dobutamine intervention.
-m- TORSIONEI
C
z
W
6-
4.
-0
W.
2a
U
140B
g -1U
-
220
180
TMUE
AoYC
260
(USEC)
MVO
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
FIGURE 5. Graph shows results for apical segment length
and torsion measurements during dobutamine infusion. Horizontal axis shows time after peak of QRS. Only the period of
interest, i.e., diastole from just before aortic valve closure, is
shown. Isovolumic relaxation period ending with mitral valve
opening is shaded. Filling starts after mitral valve opening and
continues until the next QRS (300 msec, not shown). Vertical
axis shows percent segment length on the right and torsion in
degrees on the left. Filled symbols refer to torsion and open
symbols to segment length. Squares are epicardial (EPI)
measurements; circles are endocardial (ENDO) measurements. Five dogs were studied; each apical slice contained
eight segments. AoVC, aortic valve closure; MVO, mitral
valve opening.
blood pressure from 142±17 to 160±21 mm Hg
(both p<0.001). Because of the increased heart rate,
only seven time points were available. The first time
point was a mean of 8 msec before aortic valve closure,
as measured by Doppler. Mean time of occurrence after
the peak of the QRS shifted from 178 to 158 msec for
mean
TABLE 2. Measurements for Torsion and
Mean torsion
Mid
Endo
Time (msec)
Epi
6.8
5.0
150 AVC
170
190 MVO
210
230
250
270
1577
Segment Length in Dobutamine Studies
(degrees)
Apex
Endo
Epi
17.4
11.5
Base
Epi
0.0
Endo
0.0
Mean segment length (%)
Mid
Endo
Epi
0.0
0.0
Apex
Epi
0.0
Endo
0.0
(1.9)
(2.7)
(4.3)
(6.4)
...
...
...
...
...
...
1.3
1.9
6.4
13.6
3.1
4.8
-9.8
0.7
(2.6)
(4.0)
(6.7)
(11.7)
(2.7)
(3.1)
1.4
(2.3)
9.0
1.4
2.2
5.7
6.9
5.9
(2.8)
11.8
-0.9
-10.2
(4.1)
5.9
(1.5)
(4.1)
(4.8)
(7.7)
(3.9)
(5.2)
(2.5)
(3.5)
(2.4)
(3.7)
1.3
2.8
2.2
2.0
15.6
23.7
5.4
8.6
7.8
19.5
(4.0)
(7.2)
(5.3)
(9.4)
(5.2)
(9.9)
(3.8)
(4.1)
(4.6)
(5.5)
2.2
4.1
3.4
4.3
17.5
29.0
7.2
15.4
12.1
32.2
(5.7)
(9.5)
(8.0)
(12.4)
(8.9)
(11.9)
(4.2)
(6.7)
(4.8)
(11.8)
2.1
2.0
3.8
2.2
22.4
34.8
11.1
28.3
12.2
29.5
(2.6)
(4.3)
(6.2)
(8.7)
(11.3)
(13.4)
(5.6)
(10.8)
(5.9)
(12.5)
(1.8)
-1.3
2.0
1.9
3.4
3.0
28.8
45.0
13.9
29.9
16.8
37.5
(3.9)
(5.5)
(6.7)
(6.3)
(12.8)
(16.7)
(6.8)
(11.5)
(7.6)
(15.4)
Time: timing of image acquisition after peak of QRS. As a reference, occurrence of aortic valve closure (AVC) and mitral valve opening
(MVO) are shown. Torsion: torsion expressed in degrees of the epicardium (Epi) and endocardium (Endo) at the mid and apical level,
referenced to the basal level. Segment length: circumferential segment length expressed as percent lengthening at each time point with
respect to the first (end-systolic) time point. Results for epicardium and endocardium at base, mid, and apex.
Values are mean+SD. n=10 dogs.
1578
Circulation Vol 85, No 4 April 1992
TABLE 3. Results for Regional Torsion During Control and Dobutamine
Control
At MVO
At peak
Untwisting
At peak
at MVO (%)
(degrees)
(degrees)
(degrees)
Mid epi
Septum
3.8
2.3
39
3.8
(1.0)
Inf/post
Anterior
5.0
2.4
(1.3)
(1.2)
1.1
(0.9)
Mid endo
Septum
Inf/post
Anterior
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
Apex epi
Septum
Inf/post
Anterior
Apex endo
Septum
Inf/post
Anterior
(1.1)
-0.1
100
(1.0)
7.6
6.2
(1.4)
(1.5)
7.7
5.3
(1.6)
(1.3)
2.1
1.0
(1.2)
(0.8)
8.9
4.8
(1.8)
(1.6)
10.5
(2.0)
1.9
(1.5)
-1.2
(1.3)
(1.7)
14.8
(2.3)
14.2
(1.3)
10.2
(2.1)
(1.5)
5.7
7.8
18
31
52
45
46
100
47
28
Untwisting
(degrees)
at MVO (%)
22
4.5
2.9
(1.1)
1.1
(1.2)
(1.2)
6.5
(1.1)
(0.8)
(0.9)
52
Dobutamine
At MVO
0.1
7.3
6.0
(1.3)
(1.4)
7.6
1.4
(1.3)
(1.2)
5.8
-0.9
(1.2)
(1.0)
9.9
6.7
(1.4)
(1.4)
15.9
(2.2)
10.1
(1.5)
(1.7)
(1.0)
8.3
3.1
19.0
9.9
(2.3)
(1.5)
17.4
7.3
(2.0)
(1.3)
76
98
19
82
100
33
48
69
48
58
2.1
72
3.6
77
15.8
(0.9)
(1.9)
(1.1)
Torsion: torsion expressed in degrees of the epicardium (Epi) and endocardium (Endo) at the mid and apical level,
referenced to the basal level. Peak: peak mean torsion; MVO, mitral valve opening; percent untwisting at mitral valve
opening: compared to peak. Inf/post, inferoposterior region.
Values are mean±SD.
7.6
(1.5)
Regional Untwisting
Table 3 shows the extent of early untwisting in
individual LV regions. The segments labeled in Figure 3
as 1-3 are considered septal, 4-5 inferoposterior, and
6-8 anterior. Torsion is reported from the mid and
apical epicardium and endocardium. Only two time
points are given: at the time of peak mean torsion and at
the time point closest to MVO (240 and 190 msec during
control and dobutamine, respectively).
During control conditions, the amount of maximal
torsion was smallest in the anterior region. However,
the percent untwisting at the time of MVO was significantly greater in the anterior region compared with
septum and inferoposterior (p<0.01). At the mid and
apical anterior epicardium, negative torsion was observed. (In these cases, untwisting is expressed as 100%
in Table 3.)
During dobutamine stimulation, maximal torsion became more homogeneous, whereas regional differences
in untwisting became accentuated. The extent of untwisting in the anterior wall was greater than that in the
inferoposterior wall, which was greater than that in the
septum (p<0.01), except at the apical endocardium, in
which only anterior was different from septum and
inferoposterior.
Comparing control and dobutamine, the extent of
septal untwisting at MVO remained unchanged at the
endocardium and decreased at the epicardium, whereas
inferoposterior early untwisting increased significantly
(p<O.Ol) at the mid epicardium and endocardium and
the apical endocardium. Anterior untwisting increased
significantly at the mid endocardium.
Regional differences in the extent of early untwisting were present under control conditions and became
accentuated with dobutamine stimulation; the regions
showing the largest increase of torsion with dobutamine also showed the most complete untwisting by
the time of MVO.
Effect of Open-Chest Preparation on Untwisting
Five additional dogs were studied with a closed-chest
preparation using Doppler to time aortic closure and a
transvenous pacemaker. Apical values of peak torsion and
torsion at the time of MVO were 7.7±5.20 and 4.25.10
for epicardium and 14.2i7.8° and 8.4+8.10 (+SD) for
endocardium. This corresponds to 45% and 41% untwisting at the time of MVO, not significantly different from
the open-chest preparation reported above.
Discussion
This study establishes that in the open-chest canine
model, untwisting and filling are dissociated in time and
Rademakers et al Left Ventricular Untwisting
that this dissociation is further accentuated by dobutamine
stimulation. Ventricular torsional defonnation was examined using noninvasive magnetic resonance imaging and
tagging to prevent any interference of the measuring
technique with myocardial deformation. Both torsion and
circumferential segment length were measured using the
same tag intersection points and related to Dopplerderived timing of aortic valve closure and MVO. Untwisting occurred mainly before MVO, whereas circumferential segment lengthening was present mainly during filling.
Dobutamine enhanced the extent of untwisting before
MVO and further accentuated the dissociation between
untwisting and filling. Regional inhomogeneity of untwisting was also further accentuated by dobutamine.
Relevance to Filling
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
We speculate that the occurrence of untwisting during isovolumic relaxation, when cavity volume is restrained, promotes the development of suction, inducing explosive LV filling after MVO.
The torsion that occurs during ejection is counterclockwise, that is, in the direction of the epicardial
fibers and perpendicular to the endocardial fibers.23
Ingels et al13 have postulated that epicardial contraction
dominates the direction of torsion because these fibers
are at larger radii and therefore produce greater torque
than do those in the inner layers. The process can be
viewed as a spring between the epicardium and endocardium that is stretched during systolic thickening as
the epicardium twists and moves inward but the endocardium moves further inward. As epicardial fibers
relax and begin to lengthen along their fiber direction
during isovolumic relaxation,24 they dominate the direction of untwisting, and the epicardium starts to move
outward as it unwinds. The spring is stretched further at
its epicardial side and part of the force that stretched it
at its endocardial side is released; therefore a pulling
force on the endocardium develops but cannot displace
it because the intracavitary volume is constant during
isovolumic relaxation. This could, however, contribute to a
faster pressure decay in the cavity and therefore decrease
the time constant of pressure fall. The more torsion
reversal that has occurred by the time of MVO, the more
rapid the release of potential energy will be in the form of
a negative transmural pressure creating an atrioventricular
gradient for filling. Although almost no circumferential
segment lengthening was observed in this study during
isovolumic relaxation, this does not preclude fiber lengthening because fibers run at an approximate 450 angle,23
and a change in fiber length without a change in circumferential length can occur with a shear deformation only.
Elastic recoil and restoring forces have recently received attention as possible factors in early diastolic
phenomena.6,9'2526 As summarized by Robinson et al,9
restoring forces can be subdivided into cardiac and
extracardiac forces. The latter are the result of the
stretching of the great elastic vessels and the connective
tissue, holding the heart in place. The former can be
subdivided into internal and external restoring forces.
Internal restoring forces are dependent on bending or
double overlap of filaments within each cardiac myocyte;
external restoring forces are related to the elastic collagen matrix surrounding the myocytes.27-29 Thickening of
the individual muscle fibers can stretch the net ham-
1579
mock-like endomysial fibers, and slippage of fiber layers
secondary to torsional deformation during systole could
deform the coiled perimysial fibers, which might release
their energy during early diastole.
Dobutamine
During some circumstances, such as exercise, the
same or a larger amount of blood must enter the LV
cavity during a shorter filling period. An enhancement
of the mechanism proposed here could provide the
increased atrioventricular gradient needed to accomplish this. Because the amount of untwisting that has
occurred by the time of MVO could be one of the
determinants of the degree of suction, the results of this
study, showing a further dissociation in time between
untwisting and filling during dobutamine infusion, support this theory. The mechanism for this further dissociation may be the increased intrinsic relaxation rate
during catecholamine stimulation. When the epicardial
fibers relax faster, untwisting proceeds faster and the
spring effect on the endocardium is more pronounced.
The higher paced heart rate, present in this study during
dobutamine, could also contribute to faster fiber relaxation or changed relations in the sequence of epicardial
to endocardial activation and inactivation through a
positive inotropic effect.
Using transient mitral orifice occlusion, Hori et a18
have shown that the pressure in the left ventricle does
indeed fall below zero early in diastole and that infusion
of calcium increased this suction. Suga et al7 used a
completely different model: the excised, immersed dog
heart in which the mitral valve was replaced by a
wide-open ring with a flow probe. They showed a
substantial flow across this ring toward the ventricular
cavity during diastole. This suction flow increased with
inotropic stimulation. The mechanism for this dependence of suction on inotropic state remained unclear,
although a relation with some restoring forces was
invoked. We suggest that restoring forces, generated by
the increased deformation of the collagen matrix of the
left ventricle during systole and released by nearly
complete untwisting occurring during isovolumic relaxation, are responsible for the observed enhancement in
filling characteristics.
Regional Untwisting
Although untwisting starts before MVO in all LV
locations, significant regional differences in untwisting
exist. The anterior wall, showing the least torsion in
control conditions and the largest increase in torsion
during dobutamine, invariably displays the most untwisting at the time of MVO. As torsion at any site is the
end result of opposing forces of the different fiber layers
interacting through the cardiac wall,13 a similar interaction may govern untwisting as well. Because inactivation
progresses from epicardium to endocardium, untwisting
initiated by inactivation and lengthening of counterclockwise epicardial fibers could be promoted by continued shortening of clockwise endocardial fibers.30 The
regional variability in untwisting may reflect local differences in the fiber orientation of the layers of the LV
wall or in the relative strength of these layers. For
example,
the fibers
in
the
anterior
wall
endocardium
may be more numerous or oriented in a way that more
1580
Circulation Vol 85, No 4 April 1992
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
strongly opposes the torsional direction imposed by the
epicardium, resulting in less overall torsion but faster
untwisting during early diastole. Differences in wall
curvature could also be responsible for regional differences in the amount of twist and untwisting. The
importance of these regional differences for global
diastolic function during normal conditions and enhanced relaxation during catecholamine stimulation remains to be established.
It is important in this respect that Gibson et a131 have
shown that wall thinning can be dissociated from filling
and that its timing and extent are important determinants of normal diastolic function. Regional wall thinning variability could be related to timing differences of
untwisting, as areas that are still contracting and moving
inward permit endocardial outward displacement to accompany torsion recoil in other regions during isovolumic
relaxation. It is conceivable that segmental ischemia and
other abnormal conditions might further increase the
regional inhomogeneity of untwisting; these effects await
further study. The phenomenon of intracardiac flow
toward the apex during isovolumic relaxation in hypertrophied ventricles32 could represent another expression
of regional inhomogeneity of untwisting.
Previous Studies
Systolic shear deformation was studied by Prinzen et
al17 using epicardial suction coils and by Osakada et al'8
and Feigl and Fry.'9 Using implanted radiopaque beads
and spirals, Ingels et al13 and Waldman et al10"6 have
studied torsional deformation in humans and dogs.
Those investigations have established the importance of
torsion for anisotropic shortening and proposed its role
in augmenting systolic LV wall thickening. Although the
early diastolic phase is not commented on in most of
these studies, the figures contain information about the
entire cardiac cycle. In most cases, rapid recoil of
torsional deformation is seen in early diastole. Using
radiopaque markers, Beyar et al15 have shown the
dependence of twist-radial shortening relations on the
phase of the cardiac cycle. Although twist and shortening were related during systole, torsion reversal occurred during early diastole without a concomitant
change in radial length. Yun et al33 recently measured
rapid untwisting in early diastole in human transplant
recipients by using implanted beads and noted a reduction during transplant rejection.
The results of our study define the timing of untwisting
with respect to isovolumic relaxation and its relation to
circumferential segment lengthening by using a noninvasive technique, but more importantly, the relevance of this
phenomenon for LV filling dynamics is emphasized by its
augmentation with catecholamine stimulation.
Potential Limitations
Magnetic resonance images were reconstructed from
signals obtained from repeated tissue excitation and
relaxation gated to the electrocardiogram and therefore
represent an average over multiple cardiac cycles. In
addition, information from nine acquisitions was used to
derive the required temporal resolution, spanning a
period of 30 minutes. During this time, a steady state
was desired, so steps were taken to assure this insofar as
possible. The RR interval was held constant by atrial
pacing. Additional doses of anesthesia were not given
during a set of acquisitions. After the appropriate dose
of dobutamine was established, a constant infusion was
maintained, and imaging was then delayed for 30 minutes to allow equilibration. Arterial blood pressure was
monitored after every acquisition (about every 5 minutes) and never varied more than 5 mm Hg during the
scan or 10 mm Hg during the whole series of nine scans.
Inconstant hemodynamic state is therefore unlikely to
have confounded our results.
The width of the tags (3.5 mm) could introduce
variability in measurement of their position, which is
taken as the center of the tag. However, the variability
in determination of tag position" is much smaller than
the changes observed in this study.
Shortening or lengthening along the LV long axis
during isovolumic relaxation would change the level of
the myocardium that is imaged at the different time
points because the image planes are fixed in absolute
space. In this study, long-axis measurements were not
performed, but data available from other studies from
this laboratory show that total long-axis shortening of
the canine heart during systole is 5%; during isovolumic
relaxation, the long axis shortens only 1-2%. This small
long-axis dimension change would not materially affect
the measured torsion and would have only a minor
effect on circumferential segment length.
Although magnetic resonance is a noninvasive
method that does not require the use of an open-chest
preparation, thoracotomy with pericardiotomy were
performed in this protocol to allow epicardial atrial
pacing (to assure constancy of the RR interval) and
intrathoracic phonocardiography (to assure accurate
timing of aortic closure after dobutamine administration). To ascertain that opening the chest does not alter
the overall measurements and conclusions, five additional dogs were studied with a closed-chest preparation
using Doppler to time aortic closure and a transvenous
pacemaker (repositioned when necessary to continue
capturing). The results for peak torsion and torsion at
the time of MVO were essentially the same as in the
open-chest preparation, suggesting that the dissociation
of untwisting from filling was not influenced by the
thoracotomies performed in this study.
Conclusions
The recoil of the systolic torsional deformation of the
left ventricle is dissociated from filling in the canine
model. This dissociation is further accentuated by inotropic stimulation with dobutamine infusion. Regional
inhomogeneity of untwisting is present in control conditions and is accentuated by catecholamines. This
supports the hypothesis that torsion reversal is an
important mechanism for rapid early filling and could
form the physiological basis of external restoring forces,
generating suction at normal ventricular volumes and
promoting both pressure decay and filling. Also, the
increase in extent of torsion reversal during dobutamine
infusion offers a possible explanation for the needed
enhancement of diastolic function during conditions
with shortened filling periods.
Acknowledgments
We are
grateful
to Stephanie Bosley and Susan Benac for
their invaluable assistance
during the course of this study.
Rademakers et al Left Ventricular Untwisting
References
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
1. Nonogi H, Hess OM, Ritter M, Krayenbuehl HP: Diastolic properties of the normal left ventricle during exercise. Br Heart J
1988;60:30-38
2. Ishida Y, Meisner JS, Tsujioka K, Gallo JI, Yoran C, Frater RWM,
Yellin EL: Left ventricular filling dynamics: Influence of left
ventricular relaxation and left atrial pressure. Circulation 1986;74:
187-196
3. Arora RR, Machac J, Goldman ME, Butler RN, Gorlin R,
Horowitz SF: Atrial kinetics and left ventricular diastolic filling in
the healthy elderly. JAm Coll Cardiol 1987;9:1255-1260
4. Lavine SJ, Arends D: Importance of the left ventricular filling
pressure on diastolic filling in idiopathic dilated cardiomyopathy.
Am J Cardiol 1989;64:61-65
5. Turkevich D, Micco A, Reeves JT: Noninvasive measurement of
the decrease in left ventricular filling time during maximal exercise
in normal subjects. Am J Cardiol 1988;62:650-652
6. Sonnenblick EH: The structural basis and importance of restoring
forces and elastic recoil for the filling of the heart. Eur Heart J
1980;1:107-110
7. Suga H, Goto Y, Igarashi Y, Yamada 0, Nozawa T, Yasumura Y:
Ventricular suction under zero source pressure for filling. Am J
Physiol 1986;251:H47-H55
8. Hori M, Yellin EL, Sonnenblick EH: Left ventricular diastolic
suction as a mechanism of ventricular filling. Jpn Circ J 1982;46:
124-129
9. Robinson TF, Factor SM, Sonnenblick EH: The heart as a suction
pump. Sci Am 1986;254:84-91
10. Waldman LK, Nosan D, Villarreal F, Covell JW: Relation between
transmural deformation and local myofiber direction in canine left
ventricle. Circ Res 1988;63:550-562
11. Buchalter MB, Weiss JL, Rogers WJ, Zerhouni EA, Weisfeldt ML,
Beyar R, Shapiro EP: Noninvasive quantification of left ventricular
rotational deformation in normal humans using magnetic resonance imaging myocardial tagging. Circulation 1990;81:1236-1244
12. Hansen DE, Daughters GT, Alderman EL, Stinson EB, Baldwin
JC, Miller DC: Effect of acute human allograft rejection on left
ventricular systolic torsion and diastolic recoil measured by intramyocardial markers. Circulation 1987;76:998-1008
13. Ingels NB, Hansen DE, Daughters GT, Stinson EB, Alderman EL,
Miller DC: Relation between longitudinal, circumferential, and
oblique shortening and torsional deformation in the left ventricle
of the transplanted human heart. Circ Res 1989;64:915-927
14. Beyar R, Sideman S: A computer study of the left ventricular
performance based on fiber structure, sarcomere dynamics, and
transmural electrical propagation velocity. Circ Res 1984;55:
358-375
15. Beyar R, Yin FCP, Hausknecht M, Weisfeldt ML, Kass DA:
Dependence of left ventricular twist-radial shortening relations on
cardiac cycle phase. Am J Physiol 1989;257:H1119-H1126
16. Waldman LK, Fung YC, Covell JW: Transmural myocardial
deformation in the canine left ventricle. Circ Res 1985;57:152-163
17. Prinzen FW, Arts T, Van Der Vusse GJ, Coumans WA, Reneman
RS: Gradients in fiber shortening and metabolism across ischemic
left ventricular wall. Am J Physiol 1986;250:H255-H264
1581
18. Osakada G, Sasayama S, Kawai C, Hirakawa A, Kemper WS,
Franklin D, Ross J: The analysis of left ventricular wall thickness
and shear by an ultrasonic triangulation technique in the dog. Circ
Res 1980;47:173-181
19. Feigl EO, Fry DL: Intramural myocardial shear during the cardiac
cycle. Circ Res 1964;15:536-540
20. Zerhouni EA, Parish DM, Rogers WJ, Yang A, Shapiro EP:
Human heart: Tagging with MR imaging: A method for noninvasive assessment of myocardial motion. Radiology 1988;169:
59-63
21. Lima JAC, Weiss JL, Bouton S, Zerhouni E, Jeremy R, Meils C,
Guier WH, Graves W, Levin H, Shapiro EP: Endo- versus epicardial shortening by MRI tissue tagging: Correlation with systolic
thickening during ischemia (abstract). Circulation 1989;80(suppl
II):II-152
22. Akima H: Algorithm 433: Interpolation and smooth surface fitting
based on local procedures. Commun Assoc Computing Machinery
1972;15:914-918
23. Streeter DD: Gross morphology and fiber geometry of the heart, in
Berne RM, Sperelakis N (eds): Handbook ofPhysiology. Section 2:
The Heart, Volume 1. American Physiological Society, Baltimore,
Md, Williams & Wilkins Co, 1979, pp 61-112
24. Hunter WC, Rodriguez EK, Royce MJ, Leppo MK, Weisman HF:
Myocardial sarcomere length changes are more strongly coupled to
ventricular volume during ejection than filling (abstract). Circulation 1989;80(suppl II):II-178
25. Caillet D, Crozatier B: Role of myocardial restoring forces in the
determination of early diastolic peak velocity of fibre lengthening
in the conscious dog. Cardiovasc Res 1982;16:107-112
26. Kernberg ME, Daughters GT, Alderman EL, Ingels NB Jr:
Measurement of three-dimensional systolic torsional deformation
and diastolic recoil rates in the human left ventricle (abstract).
Circulation 1980;62(suppl III):III-204
27. Weber KT: Cardiac interstitium in health and disease: The fibrillar
collagen network. JAm Coll Cardiol 1989;13:1637-1652
28. Robinson TF, Geraci MA, Sonnenblick EH, Factor SM: Coiled
perimysial fibers of papillary muscle in rat heart: Morphology,
distribution, and changes in configuration. Circ Res 1988;63:
577-592
29. Covell JW: Factors influencing diastolic function: Possible role of the
extracellular matrix. Circulation 1990;81(suppl III):III-155-III-158
30. Bertha BG, Folts JD: Phasic mitral blood flow and regional left
ventricular dimensions: Possible mechanism of active assist to
ventricular filling. Circulation 1986;74:901-911
31. Gibson DG, Greenbaum R, Marier DL, Brown DJ: Clinical
significance of early diastolic changes in left ventricular wall
thickness. Eur Heart J 1980;1(suppl A):157-163
32. Sasson Z, Hatle L, Appleton CP, Jewett M, Alderman EL, Popp
RL: Intraventricular flow during isovolumic relaxation: Description and characterization by Doppler echocardiography. JAm Coll
Cardiol 1987;10:539-546
33. Yun KL, Niczyporuk MA, Daughters GT, Ingels NB, Stinson EB,
Alderman EL, Hansen DE, Miller DC: Alterations in left ventricular diastolic twist mechanics during acute human cardiac allograft
rejection. Circulation 1991;83:962-973
Dissociation between left ventricular untwisting and filling. Accentuation by
catecholamines.
F E Rademakers, M B Buchalter, W J Rogers, E A Zerhouni, M L Weisfeldt, J L Weiss and E
P Shapiro
Downloaded from http://circ.ahajournals.org/ by guest on July 31, 2017
Circulation. 1992;85:1572-1581
doi: 10.1161/01.CIR.85.4.1572
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1992 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is located on
the World Wide Web at:
http://circ.ahajournals.org/content/85/4/1572
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally
published in Circulation 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 is online at:
http://circ.ahajournals.org//subscriptions/