Eur J Cardio-thorac Surg (1992) 6 [Suppl l]:S17-S24

‘C Springer-Verlag 1992

Importance of the mitral apparatus for left ventricular function:

an experimental approach
E. Gams ‘, S. Hag1 3, H. Schad ‘, W. Heimisch ‘, N. Mendler I, and F. Sebening ’
1 Department of Cardiovascular Surgery, German Heart Center, Munich, FRG
’ Department of Thoracic and Cardiovascular Surgery, University Hospital, Homburg/Saar, FRG
3 Department of Cardiac Surgery, University Hospital, Heidelberg, FRG

Abstract. In an experimental study of 31 anesthetized dogs the importance of the mitral apparatus for the left ventricular
function was investigated. During extracorporeal circulation bileaflet mitral valve protheses were implanted preserving the
mitral subvalvular apparatus. Flexible wires were slung around the chordae tendineae and exteriorized through the left
ventricular wall to cut the chordae by electrocautery from the outside when the heart was beating again. External and
internal left ventricular dimensions were measured by sonomicrometry, left ventricular stroke volume by electromagnetic
flowmeters around the ascending aorta, left ventricular end-diastolic volume by dye dilution technique, and left ventricular
pressure by catheter tip manometers. Different preload levels were achieved by volume loading with blood transfusion
before and after cutting the chordae tendineae. When the chordae had been divided peak systolic left ventricular pressure
did not change. Heart rate only increased at the lowest left ventricular end-diastolic pressures of 3-4 mmHg, but remained
unchanged at higher preload levels. Cardiac output decreased significantly up to -9% at left ventricular end-diastolic
pressures of 5- 10 mmHg, while left ventricular a’p/dt,,,, showed a consistent reduction of up to - 15% at any preload level.
Significant reductions were also seen in systolic shortening in the left ventricular major axis (by external measurements
- 27%, by internal recording - 43%). Left ventricular end-diastolic dimensions increased in tbe major axis by + 2% when
recorded externally, by + 10% when measured internally. Systolic and diastolic changes in the minor axis were not
consistent and different in the external and internal recordings. Left ventricular end-diastolic volume was increased by
+ 18% at any preload level, revealing an approximately parallel shift of the left ventricular end-diastolic pressure-volume
relation to the right after severance of the chordae tendineae. It can be concluded that the mitral valve apparatus is of
importance for the left ventricular function. Particularly in patients with chronic mitral valve disease and dilated hearts,
preservation of the subvalvular mitral apparatus might be of benefit for the postoperative course and long-term functional
result. [Eur J Cardio-thorac Surg (1992) 6 [Suppl ll:S17-S24]

Key words: Mitral valve - Subvalvular apparatus - Left ventricular function

Compared to other cardiac operations, mitral valve re-
placement is associated with greater postoperative deteri-
oration of left ventricular function [2, 181. Despite im-
provements in myocardial preservation, in surgical tech-
niques, and in anesthetic management and postoperative
care, no substantial reduction in postoperative morbidity
and mortality in mitral valve patients has yet been
achieved [8].
A number of clinical reports have been published sug-
gesting that in mitral valve replacement preservation of
the chordae tendineae or parts of the mitral apparatus is
beneficial to postoperative left ventricular performance
[l, 6,7,9, 33, 14,20,25]. In none of these studies, howev-
er, could it be demonstrated that improvement of the
postoperative results of mitral valve replacement was on-
ly due to the modification of the surgical technique, be-
cause in patients with mitral valve disease it is very diffi-
cult to distinguish between valvular and myocardial dys-
functions when relying only on hemodynamic indicators
available in clinical routine use, e.g., left atria1 pressure,
left ventricular end-diastolic pressure, and ejection frac-
tion.
We therefore designed a new experimental model
providing control of preload and afterload conditions in
the same beating heart. With this approach it was possi-
ble to compare left ventricular function following mitral
valve replacement with and without preservation of the
mitral apparatus.
s 18
Material and methods
Thirty-one mongrel dogs of 20-32 kg body weight were anes-
thetized following premeditation with proprionylpromazine (Com-
belen)‘. This was achieved by a single intravenous injection of
sodium pentobarbital (Nembutal) ‘. The animals were intubated
and artificially ventilated by a Servo ventilator’ during intravenous
infusion of pancuronium bromide4. To maintain anesthesia, contin-
uous intravenous infusion of 1.5 mg/kg piritramide (Dipidolor) 5
was begun thereafter. No further barbiturates were administered
during the whole experimental procedure.
The chest was opened on the left side through the fifth inter-
costal space. After incision of the pericardium 500 IU/kg body
weight heparin was given intravenously and the femoral artery and
the right atrium were cannulated. Extracorporeal circulation was
started and the animal cooled to 20 “C core temperature. The aorta
was cross-clamped and 500 ml 4°C cardioplegic solution
(Bretschneider HTK)6 given into the aortic root. The left atrium
was opened and the mitral valve exposed. Highly flexible steel wires
were put around the chordae tendineae of the anterior and the
posterior papillary muscles separately. Via small plastic cannulas
they were brought to the outside through the left ventricular wall
close to the atrioventricular groove. They were kept loose without
any tension and ready to be used for severing the chordae tendineae
later on.
Then a bileaflet valve prosthesis ’ was implanted in such a way
that the leaflets of the native mitral valve were plicated with teflon
felted U-shaped sutures. The subvalvular apparatus was preserved
and the continuity of the mitral annulus, valve leaflets, chordae
tendineae, and papillary muscles kept intact, while at the same time
care was taken that the discs of the prosthesis could open and close
completely freely.
The left atrium was closed by running sutures and the aorta
unclamped. The mean cross-clamp time was 45 f 4 min. After de-
fibrillation of the heart and during reperfusion and rewarming a
catheter-tip manometers, was inserted into the left ventricle and an
electromagnetic flowmeter put around the ascending aorta for mea-
surement of the left ventricular stroke volume. Heart rate was mea-
sured from the electrocardiogram. In one group of animals (n = 8)
left ventricular end-diastolic volumes were determined by dye dilu-
tion technique using a modification of the method described by
Holt [15]. In the same group of experiments (n= 8) internal left
ventricular diameters were measured by sonomicrometry [IO, 121
using ultrasound transducers sutured to the inner surface of the left
ventricular wall before implantation of the mitral valve prothesis,
for the minor axis diameter anteriorly and posteriorly, for the major
axis diameter apically and with a corresponding circular transducer
fixed at the ring of the mitral prosthesis (Fig. 1). In the other group
of animals (n= 23) the external left ventricular dimensions were
determined by sewing patches with ultrasound transducers on the
epicardial surface of the left ventricle, for the minor axis diameter
anteriorly between the left anterior descending coronary artery and
the first diagonal branch and posteriorly between the first and
second marginal branch of the left circumflex coronary artery, for
the major axis diameter close to the aortic root between the pul-
monary artery and the left atrium and on the apex of the heart.
One hour after weaning from cardiopulmonary bypass control
values for heart rate, left ventricular end-diastolic pressure, left
ventricular systolic pressure, maximum rate of rise of left ventricu-
lar pressure (dP/dr,,,), left ventricular diameters in the major and
minor axis, left ventricular stroke volume and left ventricular end-
’ Bayer, Leverkusen, FRG
2 Deutsche Abbott GmbH, Ingelheim, FRG
3 Siemens-Elema, Stockholm, Sweden
4 Organon-Teknika, OberschleiDheim, FRG
5 Janssen, Dusseldorf, FRG
6 Kiihler Chemie, Alsbach/BergstraDe, FRG
’ St. Jude Medical Minneapolis, USA
s Millar Instruments Inc., Houston, USA
Fig. 1. Mitral valve prosthesis with ultrasound transducers fixed at
the sewing ring before implantation. Left:view from the left atrium;
Right: view from the left ventricle with the transducers in position
diastolic volume were determined under normovolemia. Volume
loading was performed subsequently by blood transfusion (20 ml/
kg body weight) over 10 min up to preload left ventricular end-dias-
tolic pressures of 11 - 12 mmHg (in a few cases to 15 mmHg). Dur-
ing volume loading the hemodynamic variables and left ventricular
diameters were monitored continuously. Thereafter blood volume
was taken from the circulation until normovolemia and hemody-
namic stabilization was reached. Then the chordae tendineae of the
anterior and the posterior papillary muscles were cut by the follow-
ing procedure. The flexible steel wires which had been put around
the chordae tendineae of each papillary muscle separately (Fig. 2)
and passed through the left ventricular wall were now tightened
from the outside via the insulating teflon cannulas (Fig. 3). By
passing alternating current through the wires the chordae of the
anterior and the posterior papillary muscles were divided by electro-
cautery while the heart continued beating without any mechanical
and electrical irritation (Fig. 4). Subsequently volume loading was
performed in the same manner as before and left ventricular func-
tion curves were recorded again. At the end of each experiment a
gross examination of the heart was performed to confirm complete
severance of the chordae tendineae of the anterior and the posterior
papillary muscles.
Left ventricular end-diastolic diameters were defined as the dis-
tance between the ultrasound transducers before the rise of left
ventricular pressure during each cardiac cycle. Systolic shortening
of the left ventricular major and minor axis diameters was defined
as the change in length during the ejection phase of the systole, i.e.,
between opening and closure of the aortic valve.
For statistical evaluation mean values and the standard error of
the mean were calculated for each variable. Wilcoxon’s signed rank
test for matched pairs was performed to compare the results before
and after division of the chordae tendineae at the same preload
levels of the left ventricular end-diastolic pressure. The critical val-
ues for this test were taken from McCormack’s tables [19]. A differ-
ence at PcO.05 was taken to be statistically significant.
Results
Changes in diastolic and systolic left ventricular diameters
The changes in the left ventricular diameters following
severance of the chordae tendineae are summarized in
Tables l-4. There is a significant increase in the external
end-diastolic major axis diameter by 2-3% at any left
ventricular end-diastolic pressure of 3 - 8 mmHg, when
the chordae tendineae had been cut (Table 1). This elon-
gation in the major axis of the left ventricle was even
more pronounced when the internal diameters were mea-
sured: the increase following severance of the chordae
Fig. 2. Mitral valve replacement with preservation of the subvalvu- Table 1. Changes in the external end-diastolic diameters of the left
lar apparatus. View from the left ventricle to the implanted prosthe- ventricle after severance of the chordae tendineae following mitral
sis. Steel wires are slung around the chordae tendineae of the ante- valve replacement
rior and posterior papillary muscle separately and brought to the
outside through the left ventricular wall via insulating Teflon cannu- LVedP Major axis Minor axis
las (mmHg)
Before After n Before After n
Fig. 3. Mitral valve replacement with preservation of the subvalvu- severance severance severance severance
lar apparatus. View from the left ventricle to the implanted prothe- (=lOO%; (X) (=lOO%; (%)
sis. The steel wires have been tightened from the outside via the mm) (mm)
insulating Teflon cannulas
3-4 74.5 102.1 20 57.6 99.4 20
Fig. 4. View from the left ventricle to the implanted mitral valve +1.2 *o.i * +1.0 *0.3*
prosthesis following severance of the chordae tendineae by applica- 5-6 102.2 21 99.3 20
tion of electrocautery on the steel wires from the outside, while the +0.1* +0.2*
heart continued beating. The subvalvular apparatus is completely
disconnected from the left ventricular wall 7-8 76.6 102.4 10 61.3 99.4 13
*1.3 *0.2* * 1.0 0.3 *

Left ventricular end-diastolic pressures (LVedP) were obtained dur-

ing volume loading
* P<O.O5

amounted to 6-10% at left ventricular end-diastolic

pressures of 3 - 12 mmHg (Table 2). Only small changes
in the end-diastolic diameters were seen in the internal
and external minor axes of the left ventricle, which more- Table 2. Changes in the internal end-diastolic diameters of the left
over were inconsistent. ventricle after severance of the chordae tendineae following mitral
valve replacement
Not only did the end-diastolic diameter of the left
ventricle change when the subvalvular apparatus had LVedP Major axis Minor axis
been divided, but also the systolic shortening of the left (mmf-fg)
ventricular diameters was affected. The systolic shorten- Before After n Before After n
ing in the major axis diameter was drastically reduced. At severance severance severance severance
left ventricular end-diastolic pressures of 3 - 8 mmHg a (=lOO%; (%) (=lOO%; (%)
decay of the contraction amplitude in the external major mm) mm)
axis of 20-27% occurred (Table 3). Measurement of the 3-4 53.2 106.6 6 26.2 98.2 6
internal diameters showed the systolic shortening to be +1.4 *0.6* k3.6 *1.5
even more reduced: the decrease during systolic contrac- 5-6 106.6 8 100.3 8
tion ranged from 34% to 43% in the major axis of the left _fO.8* k1.4
ventricle (Table 4). In contrast, there were inconsistent
7-8 109.2 6 101.8 7
changes of the contraction amplitude in the minor axis _+0.7* k2.5
diameters: when measured internally there were no signif-
9-10 110.1 6 104.3 6
icant changes (Table 4). In external recordings significant *0.5* +1.0*
increases in the systolic shortening up to 8- 15% were
11-12 56.1 110.1 32.6 102.5 6
seen at left ventricular end-diastolic pressures of 3-
+1.2 +0.7* +3.3 * 1.1*
6 mmHg. At left ventricular end-diastolic pressures of
7-8 mmHg, however, there was no change of systolic Left ventricular end-diastolic pressures (LVedP) were obtained dur-
shortening in the external minor axis diameter of the left ing volume loading
ventricle (Table 3). * P<O.O5
s 20

Table 3. Systolic shortening of the external left ventricular diame- Table 5. Left ventricular systolic pressure (LVP,,,,) and maximum
ters after severance of the chordae tendineae following mitral valve rate of rise of left ventricular pressure (dP/dt,,,) after severance of
replacement the chordae tendineae following mitral valve replacement

LVedP Major axis Minor axis LVedP LVP,,,, dP/dt_

(mmHg) (mmHg)
Before After n Before After n Before After n Before After n
severance severance severance severance severance severance severance severance
(=lOO%; (%) (==lOO%; (%) (==lFg(o; (%)
mm)

3-4 1.72 79.9 16 1.33 115.3 14 3-4 85.2 97.8 27 1922 90.7 21
f0.1 f3.6 * f0.35 *5.5* f3.0 +1.4 +204 *1.5*
5-6 76.4 14 107.6 12 5-6 98.9 31 90.1 26
*3.5* *3.0* fl.O f0.9*
7-8 1.95 73.5 12 2.84 98.9 13 7-8 101.2 21 89.6 17
kO.18 *5.1* + 0.46 f3.1 f1.2 +1.5*
9-10 101.2 11 85.0 9
* P<O.O5
+1.2 f2.5*
II-12 133.3 100.2 6 2708 84.7 6
k4.0 k2.8 f228 +3.5*
Table 4. Systolic shortening of the internal left ventricular diame-
ters after severance of the chordae tendineae following mitral valve * P<O.O5
replacement

LVedP Major axis Minor axis

Table 6. Heart rate and cardiac output after severance of the chor-
(mmHg)
Before After n Before After n dae tendineae following mitral valve replacement
severance severance severance severance
(=lOO%; (%) (==$I%; (%) LVedP Heart rate Cardiac output
mm) (mmHg)
Before After n Before After n
3-4 3.15 57.9 6 4.23 112.5 6 severance severance severance severance
kO.57 +4.0* f0.82 k8.9 i;a;l;%; (%) {,;=I:%; (%)

5-6 61.0 8 108.1 8 min)

*4.9* f7.0
7-8 60.7 7 106.5 7 3-4 156+7 103.4 28 1.62 96.4 21
f 7.9 * f8.6 *1.2* +0.2 f2.2

9-10 65.6 6 99.2 6 5-6 100.2 31 92.5 29

*8.6* f9.2 +1.0 f2.1

11-12 3.87 57.2 6 7.19 98.4 6 7-8 100.9 19 91.3 17

f0.77 f8.3* f1.19 f8.7 kl.2 f2.3*
9-10 98.4 10 92.4 8
* P<O.O5 * 1.9 f3.0*
11-12 157+9 101.9 6 3.61 95.2 6
+4.1 f0.4 f6.7
Changes in left ventricular hemodynamic parameters * P<O.O5

Global left ventricular hemodynamics were assessed by

measurement of left ventricular systolic pressure, maxi-
mum rate of rise of left ventricular pressure (dP/dt,,,),
heart rate, and cardiac output, as summarized in Tables
5 and 6. After the chordae tendineae had been divided the
left ventricular systolic pressure did not change signifi-
cantly in any animal, while significant changes occurred
in the dP/dt,,, at all preload levels: dP/dt,,, was dimin-
ished by 9 - 15% at comparable left ventricular end-dias-
tolic pressures (Table 5). At the same time, heart rate
increased only at a preload pressure of 3-4 mmHg, but
remained constant at comparable left ventricular end-
diastolic pressures of 5 - 12 mmHg before and after sever-
ance of the chordae tendineae. Cardiac output, calculated
from heart rate and stroke volume as measured by the
electromagnetic flowmeter around the ascending aorta,
showed a significant diminution of 7-9% at preload
pressures of 5- 10 mmHg, reflecting impairment of the
left ventricular systolic function following loss of the
valvulo-annular continuity of the mitral apparatus (Table
6).
Effects on the left ventricular pressure-volume relationship
Considering the changes in left ventricular end-diastolic
diameters in the major and minor axes following the divi-
sion of the chordae tendineae, one would also expect
effects on the left ventricular end-diastolic volume. After
severance of the chordae tendineae the mean left ventric-
ular end-diastolic volume at each left ventricular end-
diastolic pressure was increased by 18 + 3 % . This implies
that after cutting of the chordae tendineae there is an
additional need of volume for the left ventricle. From
Fig. 6 it can be seen that the “resting tension curve” of
the left ventricle is shifted to the right following division
of the chordae tendineae. When the mitral valve appara-
tus is preserved the left ventricle moves on a different
pressure-volume curve to that obtained where the chor-
dae tendineae have been divided. The latter causes an
approximately parallel shift of the pressure-volume curve
to the right, and previous levels of left ventricular end-
diastolic pressure can only be reached when volume is
added to the left ventricular chamber. It can be calculated
that an increase of the filling volume by 0.9- 1.Oml/kg
body weight is necessary for the left ventricle to reach the
same left ventricular end-diastolic pressure.

Discussion

Long-term results following mitral valve replacement are

still unsatisfactory [8, 161. It has been suggested that this
may be due to the surgical technique widely used in mitral
valve replacement: in conventional mitral valve replace-
ment valve leaflets, chordae tendineae, and parts of the
papillary muscles are excised. As early as in 1963 and
1964, Dahlblck and Schiiller [4] and Lillehei et al. [17]
speculated on improvement of the postoperative results
in mitral valve replacement when the mitral valve appara-
tus is preserved, at least in part. This idea was proposed
again in 1981 by David et al. [5] and in 1983 by Hetzer
et al. [14]. The reasoning behind the suggestion not to
excise the mitral valve in toto, including the subvalvular
apparatus, is based upon the changes in left ventricular
dimensions following the loss of chordal integrity. The
difference between the complete discontinuity of the mi-
tral valve apparatus and the intact architecture of the left
ventricle with the mitral valve complex can be seen in
Figs. 4 and 5.
Experimental investigations performed in dogs by
Rastelli et al. [21] and Cohn et al. [3] on the one hand and
by David et al. [5] on the other gave contradictory results.
More recent studies by Salter et al. [22] did not reveal any
negative effects on left ventricular function following de-
tachment of the chordae tendineae. Severe impairment of
the left ventricular performance, however, was shown in
the isovolumic heart preparation used by Hansen et al.
[ll], Sarris et al. [23], and Spence et al. [24] following
disconnection of the chordae tendineae. In order to avoid
the limitations of the isolated heart preparation (e.g.,
interference of the intraventricular balloon with the chor-
dae tendineae and the papillary muscles, which causes
deterioration of the left ventricular performance) we
chose a new experimental approach. Since it was neces-
sary to prove the importance of the mitral valve appara-
tus for the left ventricular function, we performed animal
experiments and used the same closed beating heart,
pumping physiologically in an intact circulatory system
under comparable preload and afterload conditions.
Without reopening the heart or any reintervention in the
Fig. 5. View from the left ventricle to the intact mitral valve appa-
ratus. The complexity of the left ventricular architecture with the
anchoring points of the valve leaflets, chordae tendineae, and pap-
illary muscles in the ventricular wall can be seen in contrast to
Fig. 14, where the subvalvular mitral apparatus has been com-
pletely destroyed
mitral valve, it was possible to sever the chordae
tendineae by external use of electrocautery. Left ventricu-
lar function could be studied under similar hemodynamic
conditions with and without preservation of the mitral
valve apparatus.
The results obtained showed that following loss of the
functioning subvalvular mitral apparatus there is a
marked change in shape and size of the left ventricle. At
comparable preload (i.e., left ventricular end-diastolic
pressure) and afterload levels (i.e., systolic left ventricular
= aortic pressure) left ventricular end-diastolic major
axis diameters and left ventricular end-diastolic volumes
are increased. At the same time systolic contraction is
impaired. There was a substantial decay of the extent of
systolic shortening in the major axis diameter. The fact
that this reduction was more pronounced in the internal
than in the external measurements might be due to the
different changes in left ventricular wall thickness, which
could be shown in an additional series of experiments
where left ventricular wall stress was determined. In the
minor axis diameter the increased systolic shortening at
left ventricular end-diastolic pressures of 3 - 6 mmHg
might possibly compensate for the diminution of the con-
traction amplitude in the major axis diameters. At higher
s 22
LVedP (mmHg)
20 -
15 - before
10 -
5 -
r shift of the end-diastolic pressure-volume
0-
I I I I
1.5 2.0 2.5 3.0
LV end-diastolic Volume (ml/kg)
left ventricular filling pressures, however, the systolic
shortening in the minor axis diameter did not differ from
“precut” control values; they rather seemed to have a
tendency to further deterioration as they occurred in the
major axis diameter. Salter et al. [22] described similar
findings from their experiments: they saw no difference in
left ventricular minor diameter between ventricles with
attached and detached subvalvular apparatus, and no
changes in stroke work were calculated.
In our study, however, a marked reduction in the left
ventricular systolic contraction became evident along the
major axis resulting in a decrease of cardiac output -
obviously due to reduction of the stroke volume, since
heart rate did not change at these preload levels. In con-
nection with the simultaneous decrease of dP/dt,,, this
must be interpreted as impairment of the overall left ven-
tricular function when the subvalvular apparatus has
been destroyed. Continuity between the mitral valve and
the ventricular wall which is maintained by the subvalvu-
lar apparatus obviously plays in important role during
cardiac contraction. With the removal of the mitral valve
apparatus, previous levels of stroke volume and cardiac
output can only be reached at higher left ventricular end-
diastolic volumes and with a higher left ventricular wall
tension. As was demonstrated in the left ventricular pres-
sure-volume diagram (Fig. 6), there was an almost paral-
lel shift of the “resting-tension curve” to the right after
the chordae tendineae were cut. By definition, this leads
to unfavorable conditions in terms of cardiac muscle me-
chanics, because of restrictions on the additional volume
load needed during physical exercise.
It can be concluded that preservation of the mitral
complex in mitral valve replacement is beneficial to left
ventricular function. In particular in patients with prein-
jured and dilated hearts, integrity of the subvalvular ap-
paratus could possibly improve the functional results fol-
lowing mitral valve replacement.
Severance of
Chordae
after Fig. 6. Left ventricular end-diastolic pres-
sure-volume relationship before and after
severance of the chordae tendineae following
mitral valve replacement. After severance of
the chordae tendineae there is a narallel
curve to the right, indicating a need for ad-
ditional left ventricular volume to reach pre-
vious end-diastolic pressures. LVedP, left
I I ventricular end-diastolic pressure
3.5 4.0
Acknowledgements. The excellent technical assistance of Mrs A.
Bernhard-Abt, Mrs. S. Dorn-Biermeier, MS Ch. Schtilgen, MS U.
Ettner, Mr H. Erk, and Mr A. Schroll, M. D., is gratefully acknowl-
edged.
References
1. Asano K, Furuse A (1987) Techniques of modified mitral valve
replacement with preservation of the posterior leaflet and chor-
dae tendinae. Thorac Cardiovasc Surg 35:206
2. Cheitlin MD (1987) The timing of surgery in mitral and aortic
disease. Curr Probl Cardiol 12:69
3. Cohn LH, Reis RL, Morrow AG (1968) Left ventricular func-
tion after mitral valve replacement. J Thorac Cardiovasc Surg
56:ll
4. Dahlblck 0, Schtiller H (1963) Open correction of mitral insuf-
ficiency. A modification of the Starr-Edwards technique. Acta
Chir Stand 126:300
5. David TE, Strauss HD, Meher E, Anderson MJ, Macdonald
IL, Buda AI (1981) Is it important to preserve the chordae
tendineae and papillary muscles during mitral valve replace-
ment? Can J Surg 24:326
6. David TE, Dunin-Bell 0, Orr S, Nguyen C (1983) Cardiac valve
surgery in patients with poor left ventricular function. Can J
Surg 26: 157
I. David TE, Burns RJ, Bacchus CM, Druck MN (1984) Mitral
valve replacement for mitral regurgitation with and without
preservation of chordae tendineae. J Thorac Cardiovasc Surg
88:718
8. Ferrazi P, McGifIin DC, Kirklin JW, Blackstone EH, Bourge
RC (1986) Have the results of mitral valve replacement im-
proved? J Thorac Cardiovasc Surg 92: 186
9. Goor DA, Mohr R, Lavee J, Serraf A, Smolinsky A (1988)
Preservation of the posterior leaflet during mechanical valve
replacement for ischemic mitral regurgitation and complete my-
ocardial revascularization. J Thorac Cardiovasc Surg 96: 253
10. Hag1 S, Heimisch W, Meisner H, Erben R, Franklin D, Seben-
ing F (1975) Ultraschallverfahren zur direkten Erfassung der
regionalen Myokardfimktion. Thoraxchirurgie 23:291
11. Hansen DE, Cahill PD, Declampi WM, Harrison DC, Derby
GC, Mitchell RS, Miller DC (1986) Valvular-ventricular inter-

action: importance of the mitral apparatus in canine left ven-
tricular systolic performance. Circulation 73: 1310
12. Heimisch W, Hag1 S, Meisner H, Franklin D, Kemper WS
(1975) Aufzeichnung lokaler Myokardfunktionen nach dem
Ultraschall-Laufzeit-Prinzip. Biomed Tech 20 [Suppl]: 135
13. Hennein HA, Swain JA, McIntosh CL, Bonow RO, Stone CD,
Clark RE (1990) Comparative assessment of chordal preserva-
tion versus chordal resection during mitral valve replacement.
J Thorac Cardiovasc Surg 99: 828
14. Hetzer R, Bougioukas G, Franz M, Borst HG (1983) Mitral
valve replacement with preservation of papillary muscles and
chordae tendineae - revival of a seemingly forgotten concept.
Thorac Cardiovasc Surg 31:291
15. Holt JP (1956) Estimation of residual volume of the ventricle of
the dog’s heart by two indicator dilution technics. Circ Res
4:187
16. Kraus F, Dacian S, Rudolph W (1982) Belastungsuntersuchun-
gen bei valvulirer Herzerkrankung und Herzklappenersatz.
Herz 7: 144
17. Lillehei CW, Levy MJ, Bonnabeau RC (1964) Mitral valve re-
placement with preservation of papillary muscles and chordae
tendineae. J Thorac Cardiovasc Surg 47: 532
18. Marshall WG, Kouchoukos NT, Karp RB, Williams JB (1983)
Late results after mitral valve replacement with the Bjiirk-Shi-
ley and porcine prostheses. J Thorac Cardiovasc Surg 85:902
19. McCormack RL (1965) Extended tables of the Wilcoxon
matched pair signed rank statistic. J Am Stat Assoc 60:864
20. Miki S, Kusuhara K, Ueda Y, Komeda M, Ohkita Y, Tahata T
(1988) Mitral valve replacement with preservation of chordae
tendineae and papillary muscles. Ann Thorac Surg 45: 28
S 23
21. Rastelli GC, Tsakiris AG, Frye RL, Kirklin JW (1967) Exercise
tolerance and hemodynamic studies after replacement of canine
mitral valve with and without preservation of chordae tendineae.
Circulation 35 and 36 [Suppl I]:34
22. Salter DR, Pellom GL, Murphy CE, Brunsting LA, Goldstein
JP, Morris JM, Wechsler AS (1986) Papillary-annular continu-
ity and left ventricular systolic function after mitral valve re-
placement. Circulation 74 [Suppl I]: 121
23. Sarris GE, Cahill PD, Hansen DE, Derby GC, Miller DC
(1988) Restoration of left ventricular systolic performance after
reattachment of the mitral chordae tendineae: the importance
of valvular-ventricular interaction. J Thorac Cardiovasc Surg
95: 969
24. Spence PA, Peniston CM, Mihic N, Jabr AK, Narini P, Salerno
TA (1986) Toward a better understanding of the etiology of left
ventricular dysfunction after mitral valve replacement: an ex-
perimental study with possible clinical implications. Ann Tho-
rat Surg 42: 363
25. Yagyu K, Matsumoto H, Asano K (1987) Importance of the
mitral complex in left ventricular contraction - an analysis of
the results of mitral valve replacement with preservation of the
posterior mitral complex. Thorac Cardiovasc Surg 35: 166
Prof. Dr. Emmeran Gams
Klinik fur Thorax-, Herz- und Gefisschirurgie
Universitatskliniken des Saarlandes
W-6650 Homburg/Saar
Federal Republik of Germany
S 24
Discussion
Dr. E E. E. Vermeulen (Nieuwegein, The Netherlands). Dr. Duran
discussed functional ischemia and regurgitation, or incompetence of
the mitral valve. I believe this concept is rather new. Have you any
specific data on this type of indication?
Dr. C. M. G. Duran (Riyadh, Saudi Arabia). We examined 169
patients who were revascularized and had mitral valve disease, and
the results were very confusing. When we tried to classify them, we
found a very diverse group of patients. These were patients at
Scripps Clinic, and there are obviously many more coronary cases
there than in Saudi Arabia. In the literature, again the results are
very variable, and, in general, poor. One group of patients had
unrelated mitral valve disease and coronary disease. The patients
had old rheumatic disease and perhaps a calcified valve, and then
coronary disease. In those, choice of repair or replacement depends
entirely on what the surgeon finds in the valve. The results were
reasonably good, though not as good as isolated vascularization or
isolated valve replacement or repair. Then there were patients with
real ischemic mitral disease in whom mortality was high, around
20%. We tried to divide those cases, and we found that again there
were two groups. One contained the acute cases with ruptured
papillary muscle. In those, the surgeon opens the left atrium and
finds pathology. Chords, usually papillary muscles, are elongated.
The mortality was extremely high with repair, but lower with re-
placement. In the second group, the surgeon opens the left atrium,
finds a normal mitral valve, and wonders what to do. Sometimes he
hopes that revascularization will solve the problem, which it very
seldom does in our opinion; other times he replaces the valve in spite
of its normal appearance. In a randomized study, we concluded that
we prefer to replace the valve in patients with mitral valve dysfunc-
tion. At least the surgeon knows he is solving the problem. Repair
may not work, and if the physician does not want a long ischemic
time, in addition to the bypasses, replacement is preferred for that
group. Before going on bypass, using transesophageal or epicardial
echo, we overload the patient and produce a vasoconstriction so
that the afterload is at least as high as 180 mmHg systolic. Then we
look at the echo, and if the regurgitation is between 3 and 4, we put
in a ring. If the regurgitation is not increased under these stress
conditions, we leave it alone.
Dr. C. W. Akins (Boston, USA). Some years ago we studied mitral
valve replacement in the MGH and reported a series of patients. We
found that if the ejection fraction was less than 30%, and mitral
valve replacement was performed, the mortality rate in hospital was
quite high. Even if the patient survived hospitalization, the early
attrition rate was extremely high. Therefore, we felt that if the
ejection fraction was less than 30% with isolated mitral regurgita-
tion, results were poor. Does preservation of the mitral apparatus
or mitral reconstruction affect the group of patients with very poor
ejection fractions and significant mitral insufficiency?
Dr. Cams: Ejection fraction is not a good parameter to measure left
ventricular function in mitral regurgitation, because it is very diffi-
cult to distinguish between the myocardial part of the damage, or
the decrease of ventricular function, and the valvular part. A sur-
geon may have either great success or very poor results in mitral
valve replacement. Preservation of the mitral apparatus must be of
benefit to these patients, especially to those with large left ventricles.
If the resting tension curve has already been moved to the right by
excision of the subvalvular apparatus, the patient receives no fur-
ther benefit from increased volume loading.
Dr. Akins: Do you have any evidence of that? Do you have clinical
experience where you have actually operated on patients with ejec-
tion fractions of less than 30% and severe mitral regurgitation?
Dr. Cams: There are no clinical studies showing this clearly. The
question is how to select patients for preservation of the mitral valve
apparatus and reconstruction, or for replacement of the valve and
preservation of the mitral valve apparatus?
Dr. M. J. Antunes (Coimbra, Portugal). In our early experience,
young patients with rheumatic regurgitation came late to surgery
and had very low ejection fractions. We tended not to do repair
because we were concerned that longer ischemic or cardioplegic
times would impair the left ventricular function even further. The
results were dismal with valve replacement. We began repairing
these patients, and to our surprise they did much better than those
with replacement. Whether this was due to the fact that the chordal
apparatus was preserved, or whether it was due to the fact that,
almost invariably, a degree of regurgitation was present after
surgery, I do not know.
Dr. Duran: We had a very similar experience. In the first 250 patients
with mitral valve repair, compared with a similar age group
matched for an isolated Hancock valve mitral repair, there was no
difference except for hospital mortality. We thought we had selected
the good cases for repair and the bad ones for replacement. HOW-
ever, when we compared by functional class and ejection fraction,
the differences were much more significant the worse the patient, so
the benefit of repair was much higher in the bad patients than in the
good patients. We did not consider chordal sparing, and assumed
the benefit was because of repair. It was probably due to mainte-
nance of the chordae. The worse the patient, the more important it
is to maintain subvalvular apparatus.