Transseptal pressure gradient and diastolic ventricular septal motion in patients with mitral stenosis.

Abstract

Previous studies from our laboratory have shown that the position of the ventricular septum relative to the two ventricles at end-diastole is determined by the instantaneous transseptal pressure gradient (TSG) defined as left ventricular minus simultaneous right ventricular pressure. Since patients with mitral stenosis often have exaggerated leftward (paradoxic) motion of the ventricular septum during early diastole, we studied seven patients with mitral stenosis undergoing cardiac catheterization to determine if position (and therefore motion) of the ventricular septum was determined by TSG throughout diastole. M Mode echocardiograms derived from a two-dimensional parasternal short-axis view were recorded with simultaneous micromanometer measurements of left ventricular and right ventricular pressures. Six of seven patients demonstrated abnormal early diastolic leftward motion of the ventricular septum in at least one cardiac cycle. TSG measured at intervals throughout diastole ranged from -2.5 to +20 mm Hg, with abnormal TSG observed in most of the 40 cardiac cycles selected for analysis. The intracardiac position of the ventricular septum, defined as the distance from the right ventricular epicardium (RVEpi) to the left surface of the ventricular septum normalized for total cardiac dimension (RVEpi-VS), was plotted against left ventricular pressure, right ventricular pressure, and TSG. Linear regression of pooled data from all patients (164 observations) demonstrated a highly significant correlation between the instantaneous TSG and the relative intracardiac position of the ventricular septum (RVEpi-VS = 1.52 TSG + 42.7; r = .79, p less than .0001).(ABSTRACT TRUNCATED AT 250 WORDS)

Full text

Page 1

CR Thompson, I Kingma, RP MacDonald, I Belenkie, JV Tyberg and ER Smith
patients with mitral stenosis
Transseptal pressure gradient and diastolic ventricular septal motion in
ISSN: 1524-4539
Copyright © 1987 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online
72514
Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX
1987, 76:974-980
Circulation
http://circ.ahajournals.org/content/76/5/974
located on the World Wide Web at:
The online version of this article, along with updated information and services, is
http://www.lww.com/reprints
Reprints: Information about reprints can be found online at

journalpermissions@lww.com
Fax: 410-528-8550. E-mail:
Kluwer Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050.
Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters

http://circ.ahajournals.org//subscriptions/
Subscriptions: Information about subscribing to Circulation is online at
by guest on July 10, 2011http://circ.ahajournals.org/Downloaded from

Page 2

CLINICAL INVESTIGATION
Transseptal pressure gradient and diastolic
ventricular septal motion in patients with mitral
stenosis
CHRISTOPHER R. THOMPSON, M.D. ,* IRIS KINGMA, M.D., PH.D., ROSALIND P. R. MACDONALD, B.SC.,
ISRAEL BELENKIE, M.D., JOHN V. TYBERG, M.D., PH.D., AND ELDON R. SMITH, M.D.
ABSTRACT
septum relative to the two ventricles at end-diastole is determined by the instantaneous transseptal
pressure gradient (TSG) defined as left ventricular minus simultaneous right ventricular pressure. Since
patients with mitral stenosis often have exaggerated leftward (paradoxic) motion of the ventricular
septum during early diastole, we studied seven patients with mitral stenosis undergoing cardiac
catheterization to determine ifposition (and therefore motion) ofthe ventricular septum was determined
by TSG throughout diastole. M Mode echocardiograms derived from a two-dimensional parasternal
short-axis view were recorded with simultaneous micromanometer measurements ofleft ventricular and
right ventricular pressures. Six of seven patients demonstrated abnormal early diastolic leftward motion
of the ventricular septum in at least one cardiac cycle. TSG measured at intervals throughout diastole
ranged from - 2.5 to +20 mm Hg, with abnormal TSG observed in most of the 40 cardiac cycles
selected for analysis. The intracardiac position of the ventricular septum, defined as the distance from
the right ventricular epicardium (RVEpi) to the left surface of the ventricular septum normalized for
total cardiac dimension (RVEpi-VS), was plotted against left ventricular pressure, right ventricular
pressure, and TSG. Linear regression of pooled data from all patients (164 observations) demonstrated
a highly significant correlation between the instantaneous TSG and the relative intracardiac position
of the ventricular septum (RVEpi-VS = 1.-2 TSG + 42.7; r = .79, p < .0001). The position of the
ventricular septum was significantly better coi'z:elated (p < .0001) with instantaneous TSG than with
left ventricular intracavitary pressure (r
We conclude that, in patients with mitr:! stenosis, septal position throughout diastole is determined
by the instantaneous TSG. An abnormal TSG in early diastole, probably reflecting the increased
impedance to left ventricular filling, results in the exaggerated leftward diastolic septal movement.
These data further clarify the complex interplay between active and passive motion of the ventricular
septum throughout the cardiac cycle.
Circulation 76, No. 5, 974-980, 1987.
THE PRESENCE of "paradoxic" or anterior (right-
ward) motion of the ventricular septum during systole
has been recognized in a number of clinical conditions,
including right ventricular volume-8
or pressure
overload,3' 8-12 left bundle branch block,13' 14 Wolff-
Parkinson-White syndrome,15and transposition of the
Previous studies from our laboratory have shown that the position of the ventricular
- .51) or right ventricular intracavitary pressure (r
= .28).
From the Departments of Medicine and Medical Physiology, Uni-
versity of Calgary, and the Cardiovascular Laboratories, Foothills Hos-
pital, Calgary, Alberta.
Supported in part by grants from the Alberta Heart Foundation. Drs.
Thompson and Kingma were research fellows of the Alberta Heritage
Foundation for Medical Research.
Address for correspondence: Eldon R. Smith, M.D., Department of
Medicine, FoothillsHospital, 1403-29 St. NW, Calgary, Canada T2N
2T9.
Received Jan. 27, 1986; revision accepted July 2, 1987.
Presented in part at the 58th Annual Scientific Sessions of the Amer-
ican Heart Association, Washington, D.C., November 1985.
*Current address: Department ofMedicine. St. Paul's Hospital, Van-
couver, British Columbia, Canada.
974
great artenes,'6 as well as after cardiac surgery.17 18
The abnormal septal motion observed in these various
conditions has been ascribed to a number of mecha-
nisms. For example, in patients with previous cardiac
surgery it has been attributed to excessive anterior
motion of the entire heart, presumably related to
pericardiotomy'7 or to fixation of the heart anteriorly
by sternal adhesions.18 However, many studies have
demonstrated that the abnormal systolic motion in right
ventricular volume or pressure overload is associated
with abnormalshape5-11
septum at end-diastole.
Recently we have shown in both open-chest and
conscious dogs that the position (shape) of the septum
between the ventricles at end-diastole is directly related
to the instantaneous transseptal pressure gradient dur-
ing right ventricular volume overload (produced by
19 20orposition3'" of the
CIRCULATION
by guest on July 10, 2011 http://circ.ahajournals.org/Downloaded from

Page 3

PATHOPHYSIOLOGY AND NATURAL HISTORY-VALVULAR HEART DISEASE
opening a pulmonary artery-to-right atrial shunt), pres-
sure overload (produced by pulmonary artery constric-
tion), and simulated left bundle branchblock (produced
by right ventricular pacing).21 22 More recently, we
have demonstrated that the same relationship exists in
man; when the septum was shifted markedly leftward
at end-diastole (by the development of an abnormal
transseptal
gradient)
paradoxic
occurred during the subsequent systole.23 Thus, sys-
tolic motion of the septum is a function of the position
of the septum at end-diastole; the position in turn
directly reflects the instantaneous transseptal pressure
gradient.
In contrast to these abnormalities of systolic septal
motion, patients with mitral stenosis often have abnor-
mal leftward motion of the ventricular septum in early
diastole. This has been termed paradoxic diastolic
septal motion, although it actually reflects an exag-
geration ofthe normal slight posterior septal movement
during this phase of the cardiac cycle. Some patients
with mitral stenosis and atrial fibrillation have marked
variability in diastolic filling periods. Since right ven-
tricular filling is unimpeded while the stenotic mitral
valve restricts left ventricular filling, long diastolic
intervals result in marked variation in the diastolic
transseptal pressure gradient. Based on our previous
observations,
we
postulated
throughout diastole would be determined by the instan-
taneous transseptal pressure gradient. We expected,
therefore, that the exaggerated early diastolic septal
motion observed in mitral stenosis would result from
an abnormal early diastolic transseptal pressure gra-
dient. The wide range of transseptal pressure gradients
throughout diastole, which could uniquely be obtained
without intervention in this group ofpatients, prompted
us to explore the relationship between instantane-
ous diastolic transseptal pressure gradient and septal
position.
Methods
Subjects. The study population was selected from patients
scheduled for diagnostic right and left heart catheterization with
a diagnosis of mitral stenosis and a supine echocardiogram of
good quality. The group was comprised of seven such patients,
several ofwhom also had mild or moderate mitral regurgitation.
There were three men and four women ranging in age from 31
to 61 (mean = 47) years. Mitral stenosis was considered to be
mild (mitral valve orifice greater than 1.5 cm2 by the Gorlin
formula) in four patients and moderate (1 to 1.5 cm2) in three
patients. No patient had evidence of significant coronary artery
disease (greater than 70% luminal diameter reduction) and none
had previous cardiac surgery. Chronic atrial fibrillation was
present in four patients.
Protocol. All patients gave written informed consent to par-
ticipate in the study, which had prior approval of the institutional
Ethics Commuittee. Each was studied supine in the postabsorp-
(anterior)
motion
thatseptal
position
Vol. 76, No. 5, November 1987
tive state afterthe administration of2.5 to 10mg diazepamorally
or intravenously. Micromanometer-tip catheters (Models PC-
470, PC-484A, Millar Instruments Inc., Houston, TX) were
positioned in the left ventricle from a femoral artery and in the
right ventricle from a femoral or brachial vein. Accuracy of the
pressures was ensured by comparing the output of each micro-
manometer to that of an equisensitive external manometer (Bell
andHowell 4-327-I, Pasadena, CA) connected to the fluidlumen
of the high-fidelity catheter. This was reassessed immediately
before and after each measurement and the data were discarded
unless the difference between the two systems was less that 1
mm Hg. The pressures were recorded on paper with the use of
a multichannel recorder (Electronics for Medicine VR16, Hon-
eywell Medical Electronics, White Plains, NY). For the assess-
ment of septal position and motion, simultaneous echocardio-
graphy wasperformed with a 2.25 MHz phased-array transducer
and a commercially available imaging system (CV-3400R, Dia-
sonics Cardio/Imaging, SaltLake City, UT). The transducerwas
positioned in the left parastemal area and oriented to obtain a
short-axis image of the left ventricle at the tips of the papillary
muscles. Images were stored onvideotape and a derivedMmode
was recorded on paper. With the auxiliary input channels on this
instrument, left and right ventricular pressures were superim-
posed on the echocardiographic recordings. Pressures and echo-
cardiograms were recorded simultaneously for several minutes
during quiet respiration.
Data analysis. Measurements were obtained from 4 to 8
(average = 5.7) cardiac cycles per patient, selected to produce
the greatest possible range in transseptal pressure gradients (left
ventricular minus simultaneous rightventricular pressure). Mea-
surements were obtained throughout diastole, defined as the
interval from the nadir to the beginning of the rapid upstroke of
left ventricular pressure. The onset of this interval was selected
to ensure nearly complete ventricular relaxation since the stiff-
ness characteristics of the septum would be expected to vary
during contraction and relaxation. Two to six measurements
(average = 4.1) were obtained during diastole, as dictated by
the total duration of diastole. In all diastoles demonstrating
abnormal early septal movement, the first measurement was
made before the peak of the abnormal leftward motion.
To relate the transseptal pressure gradient to the instantaneous
septal position, we analyzed the derived M mode echocardio-
graphic tracing using a method similar to that previously
described by Pearlman et al.3 This technique is illustrated in
figure 1. The right ventricular epicardium at end-diastole was
defined and a line was constructed through this point, parallel
to the transducer face. The distance from this line to the endo-
cardial echo of the left ventricular surface of the ventricular
septum (RVEpi-VS) was expressed as a percentage of the dis-
tance from the right ventricular epicardial line to the posterior
left ventricular epicardium (total cardiac dimension). The
RVEpi-VS values were then plotted against the instantaneous
transseptal pressure gradient, thereby defining a septal diastolic
"compliance" curve for each patient.
Statistical analysis. Correlations between septal position
and instantaneous left and right ventricular pressures and the
transseptal pressure gradientwereanalyzed bylinearregression
using the forced entry method (SPSS Inc., Chicago). Thesig-
nificance ofthe differences between correlation coefficients was
tested by the sample estimate test.
Results
Satisfactory pressure and echocardiographic record-
ings were obtained in all patients. Figures 1 and 2 show
M mode echocardiograms of patients with mitral ste-
nosis. These recordings demonstrate the characteristic
975
by guest on July 10, 2011 http://circ.ahajournals.org/Downloaded from

Page 4

THOMPSON et al.
RV Epi
vs
Total
Cardiac
Dimension
LV Epi
PLV
PRV
P=o
o _ ECG
FIGURE 1. M mode echocardiogram with superimposed pressure tracings demonstrating the method ofmeasurement. The right
ventricular epicardium (RVEpi) was defined at end-diastole and a line was constructed through this point parallel to the transducer
face. The distance from this line to the left surface of the ventricular septum (VS) was measured (RVEpi-VS). This distance
was normalized by dividing by the total cardiac dimension. Note that at the end of the first diastole the septum is positioned
markedly rightward (straight arrow), associated with alarge transseptal pressure gradient, but that earlyin the subsequent diastole,
the septum shifts abruptly leftward (curved arrow) in response to a markedly reduced transseptal gradient.
#'~~~~~~~~~~
FIGURE 2. M mode echocardiogram below the tips of the mitral
leaflets of a patient with mitral stenosis. This echocardiogram demon-
strates the exaggerated leftward motion of the ventricular septum in early
diastole (arrows).
t
84/1 72/03
W
-n
976
exaggerated (paradoxic) motion of the ventricular sep-
tum in early diastole. This abnormal diastolic motion
was present in 21 of 40 cardiac cycles analyzed. Only
one patient did not demonstrate this motion pattern in
at least one cardiac cycle. The variation in transseptal
pressure gradient during two consecutive diastolic
intervals of different duration in a patient with atrial
fibrillation is shown in figure 3. An abnormal diastolic
transseptal pressure gradient was present in some car-
diac cycles in all patients.
In each patient there was a significant correlation
between transseptal pressure gradient and instanta-
neous septal position, with r values ranging from .41
to .91 (table 1). Figure 4 shows the results of this
analysis in each patient. The regression equation forthe
pooled data from all patients (figure 5) was RVEpi-VS
1.52TSG +42.7(r=.79,p<.0001,SEE 4.3).
Although we haveused linearregression to define the
relationship, we knew intuitively that it would become
curvilinear at the extremes of transseptal pressure gra-
dient. This would occur given the obvious physical
limits of the maximum excursion of the ventricular
septum in either direction.
CIRCULATION
.
.-9-
z-.
.......... .......
....;..
JIM-.Rzk-.
by guest on July 10, 2011http://circ.ahajournals.org/Downloaded from

Page 5

PATHOPHYSIOLOGY AND NATURAL HISTORY-VALVULAR HEART DISEASE
FIGURE 3. Pressure tracings from a patient with atrial fibrillation with
high-gain left ventricular pressure (PLV) and right ventricular pressure
(PRV) above and low-gain PLV below. PRV exceeds PLV throughout
the first diastole (A), but during the subsequent, longer, diastole (B),
PRV rapidly plateaus whereas PLV continues to increase until end-
diastole. This results in marked variation in transseptal pressure gradient
throughout diastole.
By the sample estimate test, the correlations between
septal position and left ventricular intracavitary pres-
sure (r = .51) (figure 6) and right ventricular intra-
cavitary pressure (r = .28) (figure 7) were significantly
weaker than that (r = .79) for transseptal pressure
gradient (p < .0001).
Discussion
The results of this study demonstrate that, at least in
patients with mitral stenosis, the position of the septum
throughout diastole is highly correlated with the instan-
taneous transseptal pressure gradient. The accentuated
leftward early diastolic septal movement observed in
such patients therefore reflects an abnormal early dia-
stolic transseptal pressure gradient that in turn almost
certainly occurs because of delayed left ventricular
TABLE 1
Regression analysis of pressures versus septal position
Patient
Pressure
r values
p values
SEE
1-7
1-7
1-7
1
2
3
4
5
6
7
LV
RV
TSG
TSG
TSG
TSG
TSG
TSG
TSG
TSG
0.51
0.28
0.78
0.91
0.88
0.41
0.76
0.81
0.59
0.76
<.0001
<.0005
<.0001
<.0001
<.0001
<.05
<.0001
<.0001
<.005
<.0001
6.0
6.8
4.3
1.9
2.6
1.8
1.8
2.1
2.9
2.5
LV = left ventricular; RV
pressure gradient (PLV-PRV).
=right ventricular; TSG = transseptal
Vol. 76, No. 5, November 1987
filling secondary to the mitral stenosis. With continued
filling of the left ventricle, the normal (left ventricle >
right ventricle) transseptal gradient is reestablished and
the septum moves progressively rightward. In the pres-
ence of atrial fibrillation and long diastolic pauses, the
left ventricular pressure continues to increase after the
right ventricular pressure has reached a plateau. This
may result in a markedly increased transseptal gradient
at end-diastole (complex B, figure 3). In such in-
stances, the septum becomes markedly displaced into
the right ventricle (complex 2, figure 1), thereby
explaining the large range of septal positions evident
in figure 5. Thus, during diastole, the septum acts as
a compliant membrane between the two ventricles; that
is, the position ofthe septum is responsive to even small
alterations in transseptal gradient. Moreover,
results of this study clearly indicate that septal position
is most closely related to transseptal gradient and is
much less strongly correlated with the absolute pres-
sures in either ventricle.
The current data do not directly address the mech-
anism of the early diastolic leftward motion of the
septum normally observed, but rather only the mag-
nitude of the exaggerated motion occurring in patients
with mitral stenosis. The onset of the leftward motion
occurs at a time when pressure in the left ventricle is
still declining, although right ventricular pressure has
usually reached its nadir and begun to increase. Thus,
leftward motion is initiated at a time when the trans-
septal gradient, although decreasing, is still substan-
tially positive (left ventricle > right ventricle). Data
shown in figures 4 and 5 were collected only from the
nadir of the left ventricular pressure to the onset of
ventricular contraction. Septal position before this time
was not linearly correlatedwith the instantaneous trans-
septal gradient; this is not surprising since during this
period, the relaxation process is still incomplete. The
responsiveness of septal position to transseptal gradient
at this time would be expected to be modulated by the
altered stiffness of the contracted (or incompletely
relaxed) myocardium. In all patients, however, the
nadir or the left ventricular diastolic pressure occurred
before the maximum leftward early diastolic septal
motion. Quite clearly therefore, the data in figure 5
demonstrate that the extent of the leftward motion is
directly related to the degree of alteration in transseptal
pressure gradient. Most of the data points at or to the
left of the zero transseptal pressure gradient line rep-
resent transseptal pressure-position coordinates at or
near the peak of leftward septal motion. That is, the
marked degree of early diastolic leftward motion char-
acteristic of patients with mitral stenosis reflects an
the
977
by guest on July 10, 2011http://circ.ahajournals.org/ Downloaded from