JOURNAL OF APPLIED PHYSIOLOGY Vol. 39, No. 3, September 1975. Printed in U.S.A. Motion of both mitral valve leaflets: a cineroentgenographic study in intact dogs ANASTASIOS G. TSAKIRIS, DOUGLAS A. GORDON, YVES MATHIEU, AND IRVING LIPTON Department of Medicine, University of Sherbrooke Medical School, Sherbrooke, Qulbec, Canada TSAKIRIS, ANASTASIOS G., DOUGLAS A. GORDON,YVES MATHIEU, AND IRVING LIPTON. Motion of both mitral valve leajets: a cineroent- genographic study in intact dogs. J. Appl. Physiol. 39(3): 359-366. 1975.-Motion and position of both mitral leaflets were studied in five normal dogs l-l 1 wk after radiopaque markers were su- tured on the valve cusps and on the mitral annulus. Cinefluoro- grams and cineangiograms (100-l 20 frames/s) of left atrium and left ventricle were used to study cusp motion and intraventricular flow patterns, and to detect mitral regurgitation during sinus rhythm (42-184 beats/min) and during isolated atria1 or ventricu- lar contractions. Time-motion of both leaflets was similar through- out diastole with the exception of delayed posterior cusp opening. Peak opening and closing speeds, opening and closing times, and time of complete closure, measured from the Q wave of the ECG, were not significantly affected by the variations in heart rate. Diastolic leaflet closure began immediately after opening, while the ventricular cavity was small, and was caused by flow eddies behind the cusps. Isolated ventricular contractions closed the valve leaflets completely and symmetric valve closure was ensured by the different rates of leaflet edge approximation. In contrast, atria1 closure was slow, partial, and of very short duration. radiopaque markers; cardiopulmonary bypass; atrioventricular block; mitral leaflet position; mitral valve dimensions; diastolic valve closure; ventricular v’ortex formation; atria1 valve closure; ventricular valve closure MOTION STUDIES OF THE MITRAL VALVE leaflets ZUeimpOI’tant for the understanding of the closing mechanisms of the valve during normal and pathologic conditions. Since 1961 extensive and valuable information has been obtained by echocardiography concerning the motion of the anterior leaflet in man (3-5) and in the open-chest dog (7, 9). Data concerning the motion of both mitral cusps in the open- chest animal have been reported recently (6, 9). Still, there is a lack of systematic data concerning the variations of leaflet motion which might be expected over a wide range of hemodynamic conditions. The intact dog lends itself well to this type of study. The aims of the present investigation were I) to provide precise information about the movements and the spatial positions of the two mitral cusps in the intact dog during normal sinus rhythm over a wide range of heart rates; 2) to compare valve leaflet motion following isolated atria1 or ventricular contractions with that observed during sinus rhythm; and 3) to correlate leaflet motion with angiogra- phy. Measurements were made in five dogs in which small radiopaque markers had been sutured on the atria1 surface of both mitral cusps and on the endocardial surface of the apparent mitral annulus l-l 1 wk before the studies. METHODS Five healthy dogs weighing between 17.3 and 22.2 kg were operated on using a normothermic cardiopulmonary bypass. A left thoracotomy through the fourth intercostal space provided access to the left atrium and the appendage was incised to expose the mitral valve. Seven perforated lead beads, 2.4-2.7 mm in diameter and 70-l 15 mg in wt, were used in each dog. One marker was sutured to each commissure at the annular level. One bead was used to mark the middle of the atria1 basilar attachment of each cusp. One marker was placed in the middle of the free margin of each leaflet (usually between the free margin and the line of closure) in line with the corresponding annular bead. The last marker was placed on the atria1 surface of the anterior leaflet, midway between the annulus and the free margin (Fig. 1, left). At the time of the studies, l-l 1 wk after surgery, the dogs were in good condition and had regained their preoperative wt. Six experiments were carried out, one dog being studied twice. The animals were premeditated with morphine (2.5 mg/kg body wt) and anesthetized with sodium pento- barbital (15 mg/kg body wt). Catheters were placed in the right atrium and right ventricle (6-F bipolar electrode catheters), pulmonary artery (6-F), left ventricle (7-F), left atrium via transeptal puncture (6.5-F), ascending aorta (5-F), and abdominal aorta (6.5-F). Pressures were meas- ured by strain gauges and photographically recording galvanometers, and cardiac output was determined by dye dilution. The dogs were placed in the right decubitus position in a half-body Plexiglas cast and positioned in a single plane 6-in image-intensifier system so that the plane of the mitral valve ring was in a vertical position and parallel to the . central axis of the roentgen beam with the two commissural beads nearly superimposed (Fig. 1, right). It was in this position that the maximal excursions of the leaflet markers were observed. The mitral valve was positioned in the center of the television image to minimize distortion. Frequent injections of small amounts of contrast medium (7 ml) into the left atrium and left ventricle permitted determination of the position of the mitral cusps relative to the ventricular wall, visualization of flow patterns in the 359 by 10.220.33.2 on April 20, 2017 http://jap.physiology.org/ Downloaded from
Transcript
JOURNAL OF APPLIED PHYSIOLOGY
Vol. 39, No. 3, September 1975. Printed in U.S.A.
Motion of both mitral valve leaflets:
a cineroentgenographic study in intact dogs
ANASTASIOS G. TSAKIRIS, DOUGLAS A. GORDON,
YVES MATHIEU, AND IRVING LIPTON
Department of Medicine, University of Sherbrooke Medical School, Sherbrooke, Qulbec, Canada
TSAKIRIS, ANASTASIOS G., DOUGLAS A. GORDON,YVES MATHIEU,
AND IRVING LIPTON. Motion of both mitral valve leajets: a cineroent-
genographic study in intact dogs. J. Appl. Physiol. 39(3): 359-366.
1975.-Motion and position of both mitral leaflets were studied
in five normal dogs l-l 1 wk after radiopaque markers were su-
tured on the valve cusps and on the mitral annulus. Cinefluoro-
grams and cineangiograms (100-l 20 frames/s) of left atrium and
left ventricle were used to study cusp motion and intraventricular
flow patterns, and to detect mitral regurgitation during sinus
rhythm (42-184 beats/min) and during isolated atria1 or ventricu- lar contractions. Time-motion of both leaflets was similar through-
out diastole with the exception of delayed posterior cusp opening. Peak opening and closing speeds, opening and closing times, and
time of complete closure, measured from the Q wave of the ECG,
were not significantly affected by the variations in heart rate.
Diastolic leaflet closure began immediately after opening, while
the ventricular cavity was small, and was caused by flow eddies behind the cusps. Isolated ventricular contractions closed the
valve leaflets completely and symmetric valve closure was ensured by the different rates of leaflet edge approximation. In contrast,
atria1 closure was slow, partial, and of very short duration.
MOTION STUDIES OF THE MITRAL VALVE leaflets ZUeimpOI’tant
for the understanding of the closing mechanisms of the valve during normal and pathologic conditions. Since 1961 extensive and valuable information has been obtained by echocardiography concerning the motion of the anterior leaflet in man (3-5) and in the open-chest dog (7, 9). Data concerning the motion of both mitral cusps in the open- chest animal have been reported recently (6, 9). Still, there is a lack of systematic data concerning the variations of leaflet motion which might be expected over a wide range of hemodynamic conditions. The intact dog lends itself well to this type of study.
The aims of the present investigation were I) to provide precise information about the movements and the spatial positions of the two mitral cusps in the intact dog during normal sinus rhythm over a wide range of heart rates; 2) to compare valve leaflet motion following isolated atria1 or ventricular contractions with that observed during sinus rhythm; and 3) to correlate leaflet motion with angiogra- phy. Measurements were made in five dogs in which small
radiopaque markers had been sutured on the atria1 surface of both mitral cusps and on the endocardial surface of the apparent mitral annulus l-l 1 wk before the studies.
METHODS
Five healthy dogs weighing between 17.3 and 22.2 kg were operated on using a normothermic cardiopulmonary bypass. A left thoracotomy through the fourth intercostal space provided access to the left atrium and the appendage was incised to expose the mitral valve. Seven perforated lead beads, 2.4-2.7 mm in diameter and 70-l 15 mg in wt, were used in each dog. One marker was sutured to each commissure at the annular level. One bead was used to mark the middle of the atria1 basilar attachment of each cusp. One marker was placed in the middle of the free margin of each leaflet (usually between the free margin and the line of closure) in line with the corresponding annular bead. The last marker was placed on the atria1 surface of the anterior leaflet, midway between the annulus and the free margin (Fig. 1, left). At the time of the studies, l-l 1 wk after surgery, the dogs were in good condition and had regained their preoperative wt.
Six experiments were carried out, one dog being studied twice. The animals were premeditated with morphine (2.5 mg/kg body wt) and anesthetized with sodium pento- barbital (15 mg/kg body wt). Catheters were placed in the right atrium and right ventricle (6-F bipolar electrode catheters), pulmonary artery (6-F), left ventricle (7-F), left atrium via transeptal puncture (6.5-F), ascending aorta (5-F), and abdominal aorta (6.5-F). Pressures were meas- ured by strain gauges and photographically recording galvanometers, and cardiac output was determined by dye dilution.
The dogs were placed in the right decubitus position in a half-body Plexiglas cast and positioned in a single plane 6-in image-intensifier system so that the plane of the mitral valve ring was in a vertical position and parallel to the . central axis of the roentgen beam with the two commissural beads nearly superimposed (Fig. 1, right). It was in this position that the maximal excursions of the leaflet markers were observed. The mitral valve was positioned in the center of the television image to minimize distortion.
Frequent injections of small amounts of contrast medium (7 ml) into the left atrium and left ventricle permitted determination of the position of the mitral cusps relative to the ventricular wall, visualization of flow patterns in the
left ventricle, and detection of mitral regurgitation. Fluoro- grams of leaflet motion and angiograms were recorded on cinefilm at speeds ranging from 100 to 120 frames/s (35 mm film, 3.1 lines/mm resolution). Pressure tracings and the electrocardiogram were recorded by a multichannel Visicorder oscillograph (Honeywell, model 1912) at paper speeds of 150 mm/s. The cinecamera was synchronized with the oscillographic record by means of an electronic pulse which marked the exposure of each frame. The delay in this circuit and that of the electrocardiographic signal
d’ 8 LV LA
Q..
‘59 MVA
Al: anterior leaflet PL: posterior leaflet ,LV: left ventricle MVA: mitral valve annulus LA: left atrium
FIG 1. Atria1 view of mitral orifice and mitral leaflets demon- strating position of seven markers inserted in each dog, left panel. Markers are numbered and tracings of individual bead motion are identified with same numbers in subsequent figures. Bead numbers 1, 2, 3, and 4 are placed at annular level while 5, 6, and 7 are sutured on valve leaflets (see text). Right panel shows projections of beads on television screen with mitral ring in vertical position and parallel to central axis of X-ray beam.
MAX DIA
OPE
CKl1.C
IAL OLlC
URE
DURATION OF OPEP
DlASlOLlC CLOSURE
(-I ONSET OF CLOSURE / I --I CLOSING TIME
TIME
FE. 2. Motion of marker placed on free edge of anterior leaflet (heavy solid line) and time intervals of leaflet motion measured in this study. Identical time intervals for posterior leaflet were measured. Z indicates individual marker zero reference position (see text). Peak speeds were determined by taking maximum slopes observed in tracings during opening and closure.
CORRECTED
ECG -b
heart rek (beats min) 107
rro. 3. Correction of leaflet motion by continuous subtraction of annular movement. Panels on left show uncorrected excursions of three leaflet markers at heart rates of 67 and 107 beats/min and panels on right demonstrate same cardiac cycles after correction for annular movement. Abscissa represents instantaneous linear distance between a marker and its own reference position (Z). Upstroke indi- cates motion toward ventricle or valve opening and downstroke valve closure. Numbers on right side of every panel identify individual marker motion tracings (solid line is marker 7; interrupted line is marker 5; dotted line is marker 6). All three leaflet markers moved steadily toward ventricle during ventricular contraction due to valve ring displacement. Annulus returned to its original position early in diastole, and during late diastole uncorrected and corrected marker movements were nearly identical.
were found to be less than 0.1 ms. Thus accurate alignment of the cineframes with the electrocardiogram was possible.
The cineframes were projected with an approximate ~6 magnification to reduce the error of measurement. Cross hairs placed on the image intensifier served as a fixed refer- ence point in the positioning of each projected image. An x, y coordinate system was used to determine the spatial position of the center of every bead in each frame.
The’first step in our measurement was to determine in each cardiac cycle the coordinates of the markers early in ventricular systole, when the valve was completely closed (Fig. 2, 2). In the subsequent frames, the position of each bead was determined as a linear distance from its individual (2) reference. The distances thus computed were corrected for projection magnification and X-ray distortion. In addi- tion, the movement of the valve ring, determined from the motion of the four markers placed at the annular level,
TABLE 1. Summary of valve dimension and mitral leafret motion data, and hemodynamic variables during regular sinus rhythm (27 cardiac beats from six experiments, heart rate 42-184 beatslmin)
FIG. 4. Mitral valve leaflet motion during regular sinus rhythm at heart rates ranging from 51 to 108 beats/min in same dog (19.1 kg,
morphine-pentobarbital anesthesia). In all panels diastolic opening was followed by immediate reclosure of both leaflets, characterized
during slow and moderate heart rates (&znels I, 2, and 3) by initial fast movement, a short pause, and resumption of closure; during short
was continuously subtracted from the motion of the valve leaflets throughout the cardiac cycle, since it became evi- dent early in the study that concomitant movements of the
diastolic periods (panels 4 and 5) early diastolic closure was minimal. At slow heart rates (panel I) following early diastolic closure, both leaflets remained practically motionless in semiclosed position until atria1 reopening. Heart rate changes did not affect rate of leaflet open- ing or closure nor their maximal diastolic opening.
heart or of the annulus alone might simulate leaflet motion (Fig. 3). Our measurements determining the position of a marker were repeatable with an accuracy of ~0.3 mm.
Maximum slopes of tracings of the markers placed at the free edges of the leaflets were used to calculate peak open- ing and closing speeds.
Valve leaflet motion, pressures, cardiac output, and angiograms of the left atrium and left ventricle were re- corded in each experiment over a wide range of heart rates during normal sinus rhythm (42-184 beats/min). Slow heart rates, i.e., below 70 beats/min, occurred either spontaneously due to the morphine premeditation used (three experiments), or were produced by vagal stimula- tion (three experiments). First and second degree atrio- ventricular block occurred spontaneously in two experi- ments and vagal stimulation was used to produce transient block of varying degree in three experiments. Isolated ven- tricular extrasystoles were produced by stimulating the right ventricle in four experiments.
At the end of the study two dogs were killed and autopsy was performed. The left ventricular cavity was opened and the heart was fixed in formalin for histologic examination The remaining three animals are in good condition 5-8 mo after surgery.
RESULTS
In the two dogs in which autopsy was performed, the inserted beads were found to be placed at their desired locations. A modest amount of fibrous tissue, confined to the endocardium, was present around the beads and no deformity or fibrosis of leaflet tissue or of the mitral ring was apparent. In addition, during regular ventricular con- tractions, at heart rates studied, no mitral regurgitation was detected by angiography. These findings suggest that the presence of the markers did not compromise valvular motion and function.
Leaflet motion during regular sinus rhythm. Twenty-seven cardiac beats were analyzed in detail (Fig. 2) at heart rates ranging from 42 to 184 beats/min and the data of leaflet motion and valve dimensions are presented in Table 1.
Figure 4 illustrates mitral leaflet motion during different heart rates in the same dog. In diastole both valve leaflets opened and rapidly reached their points of maximal excur- sion. In most beats measured, valve opening was asym- metrical with the posterior leaflet opening 8-40 ms later than the anterior. This delay was not related to heart rate. Heart rate changes did not significantly affect the opening time of both leaflets (Fig. 2), their peak opening speed, or the maximal valve opening. The minimal linear distance, early in diastole, between the marker placed at the free edge of the anterior leaflet and the ventricular wall was found to average 0.8 =t 0.3 cm (0.5-1.4 cm), and between the marker placed at the free edge of the posterior leaflet and the ventricular wall 0.2 zt 0.1 cm (0.1-0.4 cm). In general the distance between valve cusps and the ventricu- lar wall was greater at slow heart rates and larger diastolic ventricular volumes.
The cusps were never observed to pause at their fully open positions but began to reclose immediately. Diastolic closure was characterized during slow and moderate heart rates by a rapid movement of both leaflets towards the atrium, followed by a short pause, and a resumption of closure (Fig. 4, panels 1, 2, 3).
Peak closing speed of the anterior leaflet during the
TSAKIRIS, GORDON, MATHIEU, AND LH’TON
d on
0
0 1 1 I I t 0 5’0 100 150 beds/min
heart rate
FIG. 5. Significant inverse relationship (P < 0.005) of duration of diastolic closure of anterior Ieaflet and heart rate during regular sinus rhythm (5 experiments represented by different symbols, r = - 0.597).
second part of diastolic closure averaged 10 cm/s and was identical to its rate of closure following isolated atria1 con- tractions. The duration of diastolic closure varied between 30 and 190 ms (Fig. 5) and showed a significant correlation with heart rate (P < 0.005). At the end of diastolic closure the valve was closed 36 & 23 % of its complete closure ex- cursion In general the valve was closed more completely during slow and moderate heart rates, whereas at- faster rates or when atria1 contraction occurred early in diastole, leaflet approximation was interrupted by atria1 opening (Fig. 4, panels 4, 5; Fig. 6, panel 2).
As a result of atria1 contraction, both leaflets reopened, the ventricular excursions being consistently less than those observed early in diastole (Fig. 4; Fig. 6, panel 1) and started to close again immediately.
The onset of closure, measured from the beginning of the P wave of the electrocardiogram (Fig. Z), averaged 110 + 21 ms and occurred clearly before the QRS complex. Closure appeared to be slower at first and more rapid during the second half with both leaflets reaching their respective closed positions simultaneously. The rate of closure of the anterior of the posterior (Table
leaflet was practically twice the rate 1) thus ensuring simultaneous arrival c
of both cusps at their closed positions. Heart rate changes over the range studied did no t significan .tly affect the closing time, the peak closing speed of both leaflets (Fig. 7), or the time of complete closure, measured from the Q wave of the electrocardiogram (Figs. 2 and 8). This time interval showed remarkably little variation (63 & 11 ms).
Measurements of the distance between markers 1 and 4 (Fig. 1) placed on the valve annulus provided information about annular size during the cardiac cycle. They indicated that presystolic annular narrowing (12, 13) occurs before or concomitantly with atria1 leaflet opening.
Leaflet motion during isolated atria1 contractions. Eleven iso- lated atria1 beats were available for analysis and the data are presented in Table 2. Atria1 contractions were in- variably followed by leaflet opening and immediate re- closure. The onset of closure, measured from the P wave of the electrocardiogram, averaged 118 T+ 15 ms and was practically the same as during sinus rhythm. Valve closure was partial and depending on the strength of atria1 con- traction varied between 40 and 95 % of complete closure
FIG. 6. Mitral leaflet motion during normal (left) and sponta- neously prolonged P-R interval (right) at identical heart rates in same dog (17.6 kg, morphine-pentobarbital anesthesia). Diastolic
0 0 0 0
0 O no m cam * l 8, .A A * a 0 A
a
b 50
I 1 I 1 100 150
1 bmls/min
heart rate
FIG. 7. Relationship of peak closing speed of posterior mitral leaflet and heart rate (regular sinus rhythm, 6 experiments represented by different symbols). Variations in heart rate had no apparent effect on rate of leaflet closure.
(Figs. 9 and 10). It was of very short duration (27 zt 15 ms) and was characterized by slower peak closing and opening speeds than those observed during sinus rhythm. Average peak closing speed of the anterior cusp following isolated atria1 beats and during diastolic closure at similar heart rates were identical (10 cm/s).
Leaflet motion during isolated ventricular contractions. Thirteen isolated ventricular beats (slow idioventricular rhythm and regular slow ventricular pacing during complete atrioven- tricular block and isolated ventricular extrasystoles during regular sinus rhythm) were analyzed. Diastolic leaflet motion during slow ventricular rates was similar to that observed during slow sinus rhythm, showing partial dia- stolic closure. With the exception of early rebounds or small slow oscillations, the cusps remained at their semiclosed positions throughout diastole until ventricular contraction closed the valve (Fig. 11). Following isolated ventricular extrasystoles, induced early in diastole, closure of the cusps due to ventricular filling and closure by ventricular con-
m set
129
0.16
closure of both cusps was diminished when atria1 contraction occurred earlier in diastole.
O-8 b
50 I I 1 I
loo 150 1
boots/min
heart rate
FIG. 8. Relationship of time of complete valve closure (measured
from Q wave of the electrocardiogram) and heart rate (5 experiments represented by different symbols). Time interval was not affected by heart rate variations.
TABLE 2. summary of mitral leaflet motion data following isolated atria1 contractions (II atria1 beats from three experiments, complete atrioventricular block induced by vagal stimulation)
FIG. 9. Comparison of leaflet motion during a regular atrial- ventricular contraction (middle) and during two isolated atria1 con- tractions (17.6 kg dog, morphine-pentobarbital anesthesia, 2 : 1 atrio- ventricular block). Atria1 leaflet closure was slower than during sinus beat, incomplete, and of very short duration.
Spontaneous
traction merged (Fig. 12). The isolated ventricular con- tractions studied closed both valve cusps rapidly, simul- taneously, and competently with peak speeds similar to those observed during sinus rhythm. The onset of closure, measured from the Q wave of the electrocardiogram, aver- aged 33 =t 16 ms. During very slow heart rates and low stroke volumes the maximal opening of the anterior leaflet
cm
’ &)O ’ ’ ‘m set
A FIG. JO. Effect of an isolated atria1 dontraction on mitral cusps
(17.6 kg dog, morphine-pentobarbital anesthesia, complete A-V block). Atria1 contraction caused leaflet opening, followed by partial transient closure and return to semiclosed valve position.
ECG:
heart rate (beats/min ): 20
I Pacing
11 I I I I 1, , 1 500
II
1000 m
FIG. 11. Complete closure of mitral cusps by two isolated ventricu- long diastole until complete closure by ventricular contractions.
lar contractions (19.7kg dog, morphine-pentobarbital anesthesia, Delayed opening and diastolic closure of posterior leaflet (left) was
complete A-V block). Slow idioventricular rhythm (left) and regular presumably caused by a decrease in transmitral flow and vortex
ventricular pacing (right). Following diastolic closure (similar to that asymmetry.
during slow sinus rhythm), mitral cusps remained motionless during
FIG. 12. Leaflet motion and distance between mitral cusp edges during induced ventricular extrasystole followed by regular sinus beat. Following opening, valve leaflets were rapidly and completely closed by ectopic ventricular beat.
was reduced, opening of the posterior leaflet was delayed, and diastolic closure was asymmetrical (Fig. 11, left).
DISCUSSION
The present study was designed to provide information about the motion of both mitral cusps in the intact animal during sinus rhythm, over a wide range of heart rates, and during isolated atria1 and ventricular contractions. This question was explored in intact dogs in which the valve leaflets and the valve annulus were rendered visible by the insertion of small radiopaque markers. This approach per- mitted direct and precise observations of cusp motion.
During moderate heart rates, the pattern of leaflet motion described by echocardiography in man (3-5) and
365
in the open-chest dog (7, 9), namely partial leaflet approxi- mation following diastolic opening, immediate valve re- opening during atria1 contraction, and closure by the com- bined effect of atria1 and ventricular contractions, was reaffirmed by the present experiments which in addition indicated that except for the delayed opening of the pos- terior cusp, time-motion of both leaflets is similar through- out diastole.
Time intervals of leaflet motion (Fig. 2: opening time, closing time, and time of complete valve closure, rneasured from the Q wave of the electrocardiogram) varied little and did not appear to be significantly affected by the wide variations in heart rate. Peak opening speed of the anterior leaflet averaged 29 I+ 7 cm/s and was similar to values obtained by echocardiography in man (5); it was not sig- nificantly affected by heart rate changes and the same was true for the peak closing speed (37 A 16 cm/s).
The simultaneous closure of the mitral cusps by the con- tracting ventricle (either during sinus rhythm or isolated ventricular contractions) is apparently ensured by the dif- ferent rates of leaflet edge approximation. Since at the onset of ventricular contraction the edge of the larger anterior leaflet was seen to be further away from its closed position than that of the posterior, closure at identical rates would result in asymmetry with the posterior cusp arriving at its closed position first, a situation that could result in mitral incompetence.
Diastolic valve closure is thought to be due to vortex formation within the ventricle with intraventricular pres- sures being higher behind the leaflets than in the area of the valve annulus (2, 11). In the present experiments, it was observed that diastolic closure begins immediately after leaflet opening, suggesting that vortex formation starts during the early phase of ventricular filling. Careful analysis of left atria1 angiograms demonstrated that contrast me- dium entering the ventricle spread behind the leaflets early in ventricular filling while the ventricular cavity was still small. This was coincident with the onset of leaflet move- ment toward the atrium. The ventricle then gained volume rapidly and flow appeared to travel mainly toward the ventricular apex. This period corresponded to the observed short pause or slowing of leaflet closure. Finally as the blood stream diverged and ran parallel to the ventricular walls, toward the atrioventricular annulus, leaflet closure resumed and continued until blood flow into the ventricle appeared to diminish. The effect of vortex activity throughout this period of diastole was symmetrical and both leaflets started and ended partial closure practically at the same time. Diastolic closing rate of the anterior leaflet in these experi- ments averaged 10 A= 3 cm/s (range 5-l 7) and was higher than values obtained by echocardiography in the open- chest dog (7). Whether the concomitant return of the basilar attachment of the valve cusps, i.e., the valve annulus (Fig. 3) to its original position facilitated in any additional way diastolic valve closure could not be determined from our observations.
The duration of diastolic closure varied and showed an inverse correlation with heart rate (Fig. 5). This could support the concept that during slow and moderate heart rates duration of diastolic valve closure and duration of rapid ventricular filling are directly related. This would
not be true during fast heart rates or when atria1 contrac- tion occurs early in diastole so that rapid ventricular filling and flow due to atria1 contraction merge (8).
tricular size, duration of transmitral flow, and degree of leaflet opening.
Since our observations suggest that the first small quan- tities of blood entering the ventricle early in diastole do initiate diastolic leaflet closure in a way similar to the aortic valve cusps during ventricular ejection (1, 14), it is un- reasonable to expect in the intact heart, during normal ranges of transmitral flow, that the size of the mitral valve orifice would be proportional to volume flow. Indeed remarkably little variation was observed in the individual dog both in the maximal excursions of the leaflets (Fig. 4) and in the distance between them early in diastole (Table 1).
During slow heart rates, after termination of diastolic closure, the cusps remained practically motionless (with the exception of early rebounds or small slow oscillations) in their semiclosed position until subsequent atria1 or ven- tricular contraction (Figs. 4 and 1 l), indicating equilibrium between the forces acting on the two aspects of the cusps. Continuation of closure or leaflet reopening was never observed, and in the same animal this floating position of the cusps seemed to be constant. It would appear to us that this semiclosed cusp position is probably due to a rapid decay of ventricular vortex activity rather than to tension exerted by the chordae tendineae as suggested previously (10). The latter explanation appears unlikely since, during long diastole, isolated atria1 contractions did invariably produce leaflet reopening and closure (Fig. 10).
During regular sinus rhythm, the onset of valve closure at the end of diastole is caused by atria1 contraction. This was evidenced by 1) the beginning of atrialward movement of both leaflets, under all heart rates studied, which occurred before the QRS complex of the electrocardiogram; 2) the onset of closure during sinus rhythm and during isolated atria1 contractions, measured from the P wave of the elec- trocardiogram, was practically identical. Our data do not allow us to answer with certainty the old controversy con- cerning the importance of atria1 systole for competent valve closure, since at the end of diastole and in the absence of atria1 contraction the valve was never found to be wide open as might have been expected, and no regurgitation could be detected following valve closure by isolated ven- tricular contractions. Though it seems unlikely then that under normal transmitral flow conditions atria1 contraction is necessary for competent leaflet closure, it may become important when smaller-than-normal quantities of blood enter the ventricle with low velocity, causing a decrease in intraventricular vortex activity and diminished diastolic leaflet approximation.
Valve motion following isolated atria1 contractions was characterized by leaflet opening and immediate slow par- tial closure of varying degree and short duration followed by a slow reopening (Figs. 9 and 10) with the rate of an- terior leaflet closure being similar to its peak diastolic closing speed. These observations would suggest a similar closing mechanism for atria1 and for diastolic valve closure. Analysis of left atria1 angiograms demonstrated the spread- ing of contrast material behind the valve cusps during iso- lated atria1 contractions. The fact that transient atria1 valve closure was at times more complete than during early ventricular filling is possibly related to differences in ven-
The observations reported in this study may be true only for the center of the mitral cusps (the sites of marker attach- ment), and for normal conditions of ventricular diastolic volume, degree of ventricular emptying, and transmitral
flow. Delayed valve opening, reduced leaflet excursions, diminished diastolic closure, and asymmetry of opening
and closure might be expected to occur under pathologic
conditions.
The authors are grateful for the technical assistance of Mrs. Anita Casey, Mrs. Brigitte Reid, Mrs. Lise Rousseau, Miss Suzanne Du- quette, Mr. Rex-& Rattelade, Mr. Fransois Bourassa, Mr. Andre
Tousignant, and Mr. Jacques Langelier. This investigation was supported in part by Medical Research
Council of Canada Grant MA-4159, and by a grant from the Joseph
C. Edwards Foundation.
Received for publication 7 November 1974.
REFERENCES
1.
9.
2.
3.
4.
5.
BELLHOUSE, B., AND F. BELLHOUSE. Fluid mechanics of model normal and stenosed aortic valves. Circulation Res. 25 : 693-704,
1969. BELLHOUSE, B. J. Fluid mechanics of a model mitral valve and left ventricle. Cardiovascular Res. 6 : 199-2 10, 1972. EDLER, I. Atrioventricular valve mobility in the living human heart recorded by ultrasound. Acta A4ed. &and. Suppl. 370: 85,
1961. FEIGENBAUM, H. Echocardiography. Philadelphia, Pa. : Lea & Febiger, 1972, p. 43-72.
. I ,
t *
KONECKE, L. L., H. FEIGENBAUM, S. CHANG, B. C. CORYA, AND
J. C. FISCHER. Abnormal mitral valve motion in patients with elevated left ventricular diastolic pressures. Circulation 47: 989-
996, 1973. 6. LANIADO, S., E. L. YELLIN, H. MILLER, AND R. W. M. FRATJIR.
Temporal relation of the first heart sound to closure of the mitral valve. Circulation 47 : 1006-1014, 1973.
7.
8.
LANIADO, S., E. YELLIN, M. KOTTLER, L. LEVY, J. STADLER, AND
R. TERDIMAN. A study of the dynamic relations between the mitral valve echogram and phasic mitral flow. Circulation 51: 104-l 13, 1975. NOLAN, S. P., S. H. DIXON, R. D. FISHER, AND A. G. MORROW.
The influence of atrial contraction and mitral valve mechanics
10.
11.
12.
13.
14.
on ventricular filling. A study of instantaneous mitral valve flow in vivo. Am. Heart J. 77 : 784-791, 1969. POHOST, G. M., R. E. DINSMORE, J. J. Rubenstein, D. D. O’KEEFE, R. N. G- ~NTHAM, H. E. SCULLY, E. A. BEIERHOLM,
J. W. FREDERIKSEN ~4. L. WEISFELDT, AND W. M. DAGGETT. The echocardiogram c the anterior leaflet of the mitral valve: corre- lation with hemodynamic and cineroentgenographic studies in dogs. Circulation 51: 88-97, 1975. RUSHMER, R. F., B. L. FINLAYSON, AND A. A. NASH. Movements of the mitral valve. Circulation Res. 4: 337-342, 1956. TAYLOR, D. E. M., AND J. D. WADE. Pattern of blood flow within the heart : a stable system. Cardiovascular Res. 7 : 14-2 1, 1973. TSAKIRIS, A. G., G. VON BERNUTH, G. C. RASTELLI, M. J. BOUR-
GEOIS, J. L. TITUS, AND E. H. WOOD. Size and motion of the mitral
valve annulus in anesthetized intact dogs. J. A/@. Physiol. 30:
611-618, 1971.
TSAKIRIS, A. G., D. D. MAIR, S. SEKI, J. L. TITUS, AND E. H.
WOOD. The motion of the tricuspid valve annulus in anesthetized
intact dogs. Circulation Res. 36 : 43-48, 1975.
VON BERNUTH, G., A. G. TSAKIRIS, AND E. H. WOOD. Effects of
variations in the strength of left ventricular contraction on aortic
valve closure in the dog. Circulation Res. 28 : 705-7 16, 197 1.
