Author + information
- Received July 27, 1998
- Revision received September 18, 1998
- Accepted November 18, 1998
- Published online March 1, 1999.
- Sten Lyager Nielsen, MD∗,†,*,
- Hans Nygaard, DMSc∗,†,
- Arnold A. Fontaine, PhD‡,
- J.Michael Hasenkam, MD, DMSc∗,†,
- Shengqui He, MD‡,
- Niels T. Andersen, PhD§ and
- Ajit P. Yoganathan, PhD‡
- ↵*Reprint requests and correspondence: Sten Lyager Nielsen, MD, Department of Cardiothoracic and Vascular Surgery, Skejby Sygehus, Aarhus University Hospital, 8200 Aarhus N,Denmark
The purpose of this study was to investigate the impact of the chordae tendineae force distribution on systolic mitral leaflet geometry and mitral valve competence invitro.
Functional mitral regurgitation is caused by changes in several elements of the valve apparatus. Interaction among these have to comply with the chordal forcedistribution defined by the chordal coapting forces (FC) created by the transmitral pressure difference, which close the leaflets and the chordal tethering forces(FT) pulling the leaflets apart.
Porcine mitral valves (n = 5) were mounted in a left ventricular model where leading edge chordal forces measured by dedicated miniature force transducers werecontrolled by changing left ventricular pressure and papillary muscle position. Chordae geometry and occlusional leaflet area (OLA) needed to cover the leaflet orifice for a givenleaflet configuration were determined by two-dimensional echo and reconstructed three-dimensionally. Occlusional leaflet area was used as expression for incomplete leafletcoaptation. Regurgitant fraction (RF) was measured with an electromagnetic flowmeter.
Mixed procedure statistics revealed a linear correlation between the sum of the chordal net forces, ∑[FC−FT]s, and OLA with regression coefficient (minimum − maximum) beta = −115 to −65 [mm2/N]; p< 0.001 and RF (beta = −0.06 to −0.01 [%/N]; p < 0.001). Increasing FTby papillary muscle malalignment restrictedleaflet mobility, resulting in a tented leaflet configuration due to an apical and posterior shift of the coaptation line. Anterior leaflet coapting forces increased due to mitralleaflet remodeling, which generated a nonuniform regurgitant orifice area.
Altered chordal force distribution caused functional mitral regurgitation based on tented leaflet configuration as observed clinically.
- apical posterolateral
- anterolateral papillary muscle
- DAL, DPL
- distance from the anterior/posterior leaflet tip to the annular plane
- FC, FT
- chordal coapting and tethering force component. Subscripts AA, AP, PA, PP refer to chordae location (chordae from APM to the anterior leaflet, from APM to theposterior leaflet, from PPM to the anterior leaflet and from PPM to the posterior leaflet). Subscripts APM and PPM refer to the chordal forces emanating from APM andPPM. Subscripts AL and PL refer to the chordal forces supplying the anterior and posterior leaflet. Subscript S refers to the sum of the chordal forcecomponents
- LAL, LPL
- horizontally projected anterior and posterior leaflet length
- LAPM, LPPM, LPM
- papillary muscle lengths (distance from APM/PPM tip to the annular plane through the midpoint of the corresponding half mitral systolic coaptation point).LPMis the mean of LAPMand LPPM
- LVP − LAP
- transmitral pressure difference
- mitral valve regurgitation
- occlusional leaflet area of the mitral valve. Subscript ACOM/PCOM refers to the anterolateral/posteromedial commissural half portion
- posteromedial papillary muscle
- mitral regurgitant fraction
- αAL, αPL
- coaptation angle of the anterior/posterior leaflet
- ϕAL, ϕPL
- anterior/posterior leaflet excursion angle
In vitro model
The experiments were conducted in a left heart simulator, comprising a computer-controlled, pressure-driven compressible bladder type pulsatile pump system (13). Hemodynamic conditions could be simulated at heart rates from 40 to 160 min−1, and cardiac output up to 10liter/min with physiologic pressure and flow wave forms. A 0.9% saline solution was used as blood analog fluid.
The measurements were conducted in a rigid left ventricular model with dimensions of the human left ventricle at the onset of systole (Fig.1). The model allowed independent adjustment of the papillary muscle positions over a clinically relevant range (Fig. 2).
A transparent millimetric foil placed on the model allowed readings of three-dimensional coordinates, where x referred to the horizontal plane, y to the sagittal plane andz to the axial plane. The centerpoint of the inner rigid annular ring was defined as origo (0,0,0). In this system the mitral valve structures could be characterizedthree-dimensionally.
Porcine mitral valves with intact chordae tendineae and papillary muscles with an intertrigonal distance of 25 mm were tested in the model. The valves were slightlyfixated in 1% glutaraldehyde solution to prolong the tissue integrity of the valve. Each valve had a total leaflet surface area that was nearly twice that of the annularring area, which is similar to the ratio of normal human mitral valves (14). The number and distribution of the chordae tendineaein the porcine mitral valve are similar to human mitral valves (15). The native mitral annulus was mounted on an inner ellipsoidring that was attached to a rigid round outer ring by a waterproof Dacron cloth stretched tightly between the two rings. The anatomic shape and size of the inner annuluswere determined by mitral valve sizers used by surgeons for sizing of mitral annuloplasty rings (Medtronic Inc.). The outer ring was mounted on the atrial section of themodel, so the anterior leaflet was positioned anteriorly between the mitral orifice and the left ventricular outflow tract.
The papillary muscles were wrapped in a Dacron cuff for support and then sutured to sewing disks on the ends of sigmoidal mounting rods, shown in Figure 1. Although the posteromedial and the anterolateral papillary muscles have been designed anatomically, they both lie in the posteriorhalf of the left ventricle with a line joining the tips running parallel to the coaptation line of the mitral leaflets (3). In ourmodel the medial scallop of the posterior leaflet was positioned posteriorly opposite to the anterior leaflet in the sagittal plane. Therefore, to simulate the papillarymuscle displacement that can occur in ischemic heart disease or dilated cardiomyopathy (8), the papillary muscles were retractedin apical (A), posterolateral direction (PL) and apical posterolateral (APL) directions (Fig. 2).
Chordae tendineae forces
Specially self-designed chordal force transducers were directly attached to the chordae tendineae of the mitral valve. Each transducer was constructed of a C-shaped copperring, 6 mm in diameter and 0.8 mm thick (Fig. 2). Two miniature strain gauges (model EA-06-031DE-350, Measurements Group, Raleigh,North Carolina) were glued to the inner and outer surface of the ring, coated and coupled in a Wheatstone half-bridge to a strain gauge indicator (model 1526, Brüel& Kjaer, Copenhagen, Denmark). Before use in the model, the transducers were individually calibrated and tested for sensitivity, hysteresis and linearity to forceunits in a range of 0 to 5 N and for electrical stability in a 0.9% saline solution. In each end of the C-template a thin slit and two small holes were incorporated to permitfixation of a chorda tendinea with a Prolene 6-0 suture. After fixation, the chordal fiber was cut between the two ends of the transducer.
The force transducers measured the resulting force on each chorda tendinea representing the sum of the tethering force component (FT) directed toward thepapillary muscles, and the coapting force component (FC) directed toward the leaflets (Fig. 3).The chordal tethering and coapting force components were both projected along the chorda; each of them increased the chordal tension. Since the papillarymuscles were fixated to static mounting rods, the chordal tethering force reflected the force measurement during left ventricular diastole, when there was no pressuredifference between the left ventricle and the left atrium. The chordal tethering force was assumed to be zero when the papillary muscles were in the normal position, and thusFTfor a given papillary muscle setting was calculated by subtracting the observed FTwith the papillary muscles in normal position fromthe FTwhen the papillary muscles were displaced. The chordal coapting force was the increase in the chordal force measurement when the left ventricularpressure was applied to close the valve. The difference between the coapting and tethering force component (FC− FT) of each chordadefined the resultant force that determined leaflet position at the point of insertion.
The spatial orientation of the chordae connecting the papillary muscle tip and the midpoint of the corresponding half mitral systolic coaptation line were determined bytwo-dimensional (2D) echocardiographic (see below). Hence, the measured chordal forces could be resolved into x (horizontal), y (sagittal) and z (axial) components:FT(x), FT(y), FT(z), FC(x), FC(y), FC(z).
Theoretically, the total chordal force balance is given by the vector summation of the tethering and coapting force vectors of all the chordae supporting the anterior andposterior leaflet (12). Using only four chordal force transducers, we could not actually measure the total chordal force balance.However, we anticipate that relative changes of the chordal force distribution were reflected by our measurements and could be used to interpret the relation between leafletconfiguration and the overall chordal force distribution.
The transmitral pressure difference (LVP − LAP) was measured by a differential pressure transducer (model DP15 TL, Validyne Inc., Northridge, CA) coupled to anamplifier/signal conditioner (model CD12 A-1-A, Validyne Inc., Northridge, California).
Transmitral flow rates were measured in the left atrium 4 cm upstream of the valve by an electromagnetic cannular flow probe with a 25.4-mm internal bore (model EP680,Carolina Medical Instruments Inc., King, North Carolina), coupled to an analog flowmeter (model FM 501, Carolina Medical Instruments Inc., King, North Carolina).
Mitral valve imaging
Two-dimensional echocardiographic recordings were made with a phased array ultrasound sector scanner (Sonolayer SSA-270A, Toshiba Inc.) using a 3.75-MHz probe (PSF-37DT).Machine settings were adjusted to provide optimal imaging with the highest possible frame rate at a penetration depth of 12 cm.
The ultrasound transducer was positioned at the apex of the left ventricular model and rotated counterclockwise from the horizontal plane (0°) in 45°steps. Thus, 2D echocardiographic recordings from four different scanning planes were obtained: 1) 0°, 2) 45°, 3) 90° (apical long-axis view)and 4) 135° (apical four-chamber view). In addition, 2D echocardiographic recordings were obtained from parasternal long-axis and short-axis views. Allechocardiographic data were recorded on videotape, and representative heart cycles were later transferred synchronized with the transmitral flow curve to a computer foroff-line analysis in a commercially available software program (EchoPAC, VingMed, Norway).
Transmitral pressure difference, mitral flow, chordal force and pump trigger signals were stored on a tape recorder (TEAC RD 180T PCM, Teac Corp., Japan) for off-line analysison a computer after analog-to-digital conversion with a sample frequency of 300 Hz. By a dedicated software program (designed in LabVIEW 4.01), one mean pump cycle from eachchannel was obtained by ensemble averaging 10 pump cycles.
Chordae tendineae forces
Peak chordal FCand FTwere assessed from the chordal force signals. Coapting force component and FTwere resolved intox,y,z vectors knowing the orientation of the chordae (see below). The magnitude of the total chordal net force, ∑[FC−FT]S, the chordal net force of anterior leaflet, ∑[FC− FT]AL,and posterior leaflet, ∑[FC− FT]PL, and from anterolateral papillary muscle,∑[FC− FT]APM, and posteromedial papillary muscle, ∑[FC−FT]PPM, were determined as the sum of the square root square of ∑[FC− FT](x),∑[FC− FT](y) and ∑[FC− FT](z) of the respectivechordae.
Early systolic peak + (LV − LA) dP/dt were computed from the LVP − LAP pressure recordings.
From the transmitral flow data the following parameters were calculated: forward flow volume, regurgitant flow volume, stroke volume and mitral regurgitant fraction (RF).The closing volume represented the fluid displaced by the valve during closure. The precoaptation regurgitant volume was defined as the difference between the measured closingvolume in the test condition and the closing volume of the same valve for the normal papillary muscle position and same physiologic condition.
Mitral valve imaging
Echocardiographic data were used to assess mitral leaflet mobility and systolic leaflet configuration.
In each apical view, the following parameters were measured (Fig. 4): the perpendicularprojected distance from the annular plane to the anterior and posterior leaflet tip (DALand DPL); the projected length of the leaflets(LALand LPL); the leaflet coaptation angle (αALand αPL), and the leafletexcursion angle, which was defined as the angle through which each leaflet travelled (ϕALand ϕPL; not shown).Except for leaflet excursion angle, only one representative early midsystolic frame per view was analyzed.
Occlusional leaflet area
The occlusional leaflet area (OLA) was defined as the leaflet surface area required to cover the mitral orifice for a given leaflet geometry orientation and was usedas an integrated measure for mitral leaflet tenting. The relative increase of OLA for a certain papillary muscle position in comparison with the normal condition thusreflected the severity of incomplete mitral leaflet coaptation, and may also be a reliable measure to reflect the presence of a regurgitant orifice area (see below).Occlusional leaflet area was calculated as follows. The leaflet contour of the tented leaflet configuration in a given apical scanning plane may be approximately describedby four heights (h1..h4) and four radii (r1..r4) (Fig. 4).Assuming that this leaflet contour is swept through 45° of rotation about the center point of the annulus, OLA can be calculated as the sum of fractions of acone produced from the four scanning planes, according to (modified from Boltwood et al. ):(1)in which r4and h4can be solved trigonometrically. The measure of OLA does not account for leaflet infolding or compression of leaflettissue at the line of coaptation. Based on the equation for OLA, the occlusional leaflet area of the anterolateral and posteromedial commissural half portion of the mitralorifice (OLAACOMand OLAPCOM) could easily be calculated. Line 3–4 in the equation refers to the leaflet edge separation, whichby summation gives the effective regurgitant orifice area, and was canceled out in complete leaflet coaptation. However, the regurgitant orifice area itself may bedifficult to assess precisely by 2D echo due to limitations of the spatial resolution.
Chordae and papillary muscle spatial orientation
Assuming that the chordae tendineae with origin at the papillary muscle tip were inserting at the midpoint of the corresponding half mitral systolic coaptation line inearly systole, the spatial orientation of the chordae connecting these coordinates could be calculated. The x,y,z coordinates of the papillary muscle tips were directlyread in the spatial coordinate system of the left ventricular model. The mitral leaflet geometry assessed by 2D echo could be superimposed on this spatial coordinatesystem, having a well defined annular hinge and scanning planes.
From the parasternal short-axis view, the x,y coordinates of the medial and lateral midpoint of the half mitral systolic coaptation line could be identified (asterisksin Fig. 5). The transducer was then moved along the x axis to center the medial and lateralmidpoint of the half mitral systolic coaptation line in the scanning planes. In these positions the transducer was rotated 90° into parasternal long-axis view,in which the z component of the midpoint of the half mitral systolic coaptation line could be measured as the distance from the leaflet tips to the annular plane.
The central papillary muscle lines (double arrows in Fig. 5) were defined between the papillary muscle tips through thecorresponding midpoints of the half mitral systolic coaptation line to the annular plane. The length of the papillary muscle lines (LAPMandLPPM) and the mean of LAPMand LPPM(LPM) designated the papillary muscle malalignment with respectto the mitral annulus.
Five valves were included in the study. Alterations of the chordal force distribution were accomplished by combinations of 1) eight different papillary muscle settings(Fig. 2), 2) two peak LVP − LAP settings (90 and 150 mm Hg) and 3) two heart rate settings (70 and 120min−1). Cardiac output was kept constant at 5 liters/min or at maximum obtainable flow rate during moderate to severe MR. Two different levels ofpeak + (LV − LA) dP/dt were accomplished by the two heart rate settings.
The data were analyzed using mixed models (17), which are unbalanced analysis of variance models, allowing more than one source ofvariation (random effects) and both nominal (valve, papillary muscle settings, high/low LVP − LAP and high/low heart rate) and continuous (e.g., LPMand∑[FC− FT]S) independent variables (fixed effects). The Proc Mixed procedure in SAS was used for theanalysis. Tests to determine whether the fixed effects were statistically significant were performed for each valve and for all valves at once. Significance level was p =0.05.
Three different models were applied:
The dependent variable Y (e.g., RF) was analyzed in relation to variation in papillary muscle settings, high/low LVP − LAP and high/low heart rate:whereαPMwas the level for the eight papillary muscle settings, γLVP − LAPwas the difference between high/lowheart rate and IAwas the indicator function equal to 1 if A was true (high pressure/heart rate) and 0 if A was false (low pressure/heart rate).
The dependent variable Y (e.g., RF) was analyzed in relation to variation in a continuous variable X (e.g., ∑[FC−FT]S), high/low LVP − LAP and high/low heart rate: where β was the regression coefficient of X, α the intercept andγ and I as stated above.
The range of the continuous variable X (RX) was defined as the difference between the maximum and minimum value of X in allexperiments in the same valve. Hence, the magnitude of β·RX, γLVP−LAPandγHRindicated the contribution of X, LVP − LAP and heart rate settings to the dependent variable, Y. Due to the variation betweenvalves, the magnitudes of β and β·RXare stated for each valve in tables or as minimum to maximum valueof five valves in the text.
The dependent variable Y (e.g., RF) was analyzed in relation to variation in two continuous variables X1and X2(e.g.,LPMand ∑[FC− FT]S), high/low LVP − LAP and high/low heart rate, thatis, the regression term β·X in model 2 was changed to β1·X1+β2·X2.
The error terms in the models contained a component for each of the sources of variation: one common for the four measurements on the same valve and same papillary musclesetting (model 2 and 3), one common for two measurements (high/low LVP − LAP) in the same valve, same papillary muscle setting and same heart rate and finally aresidual term.
With the papillary muscles mounted in normal position, 2D echocardiograms showed complete mitral leaflet coaptation with no detectable regurgitation in any of the valves atthe four test conditions. The clinically observed tented leaflet geometry was successfully reproduced by papillary muscle displacement.
Determinants of leaflet configuration
Table 1shows selected leaflet variables in relation to papillary muscle settings. Occlusional leaflet area was significantly related to papillarymuscle settings (p < 0.001) and LVP − LAP (p < 0.001) (model 1MP). A positive correlation was found betweenLPMand OLA (OLA: beta = 22 to 35 [mm2/mm], β·RLPM= 261 to 552mm2; p < 0.001). Increasing LVP − LAP reduced OLA in all valves (γLVP − LAP= −45to −10 [mm2]; p < 0.001). Occlusional leaflet area was variably related to peak + (LV − LA) dP/dt (OLA:γHR= −37 to +48 [mm2]).
Table 2shows the chordal force distribution in relation to papillary muscle settings. ∑[FC−FT]Swas a central factor governing variations of OLA (Table 3). A strong correlation was found between ∑[FC− FT]Sand OLA (Fig. 6). Table 3shows that the impact of LVP − LAP on OLA (γLVP −LAP= −14 to +47 [mm2]; p = NS in four valves) was compensated by the variations of ∑[FC−FT]S. Using ∑[FC− FT]Sand LPMin a covariateanalysis for OLA, the significant impact of LPMon OLA was entirely neutralized by variations of ∑[FC−FT]S(LPM: p = 0.46 vs. ∑[FC− FT]S: p< 0.001 [model 3MP]).
Increasing chordal tethering forces by displacement of the papillary muscle restricted leaflet mobility tending to pull the corresponding leaflets apart. A positivecorrelation was found between ∑[FC− FT]ALand anterior leaflet excursion angleϕAL(beta = 2.0 to 8.3 [°/N]; p < 0.001 [model 2MP]) and between ∑[FC− FT]PLand posterior leaflet excursion angle ϕPL(beta = 4.7 to 12.0 [°/N]; p <0.001 [model 2MP]). Increasing the z component of the tethering forces by apical papillary muscle displacement significantly increased DALandDPL(DALvs. ∑[FC− FT]AL[z]: beta = −3.3 to−1.2 [mm/N], p = 0.01; DPLvs. ∑[FC− FT]PL[z]: beta = −3.3 to−1.3 [mm/N]; p < 0.001 [model 2MP]). Similarly, increasing the y component of the tethering forces by posterolateral papillary muscledisplacement significantly increased LALand decreased LPL(LALvs. ∑[FC−FT]AL[y]: p < 0.01; LPLvs. ∑[FC−FT]PL[y]: p < 0.01 [model 2MP]).
Hence, apical papillary muscle displacement resulted in an apical shift of the leaflet coaptation line, whereas posterolateral displacement of the papillary muscles causeda remodeling of the mitral leaflet surface area in which the anterior leaflet covered a larger part of the mitral orifice, resulting in a posterior shift of the leafletcoaptation line (Fig. 7). Therefore, by posterolateral displacement of the papillary muscles thechordal coapting force component of the anterior leaflet increased due to a larger leaflet surface area disposed, which counterbalanced increasing tethering forces and therebyprevented excessive apical migration of the anterior leaflet. Leaflet infolding along the coaptation line reduced the effective posterior leaflet surface area disposed, therebydecreasing the chordal coapting force component of the posterior leaflet. Therefore, posterior leaflet mobility was restricted partly due to increasing chordal tethering forcespartly due to decreasing chordal coapting forces of the posterior leaflet.
Separate changes of the posterior papillary muscle (normal:PL, normal:APL papillary muscle settings) increased [FT]PAand[FT]PPand also caused a reduction of the chordal coapting forces of the posterior leaflet on both sides of the commissures([FC]APand [FC]PP) (Table 2). Posteromedial papillarymuscle displacement thus affected the corresponding half portion of occlusional leaflet area, OLAPCOM, but tended also to increase OLAACOM,even though the anterolateral papillary muscle position was unchanged (Table 1).
Determinants of regurgitant fraction
Mitral regurgitant fraction significantly increased with papillary muscle displacement (p < 0.001, model 1MP). Apical papillary muscledisplacement was sufficient to create MR; however, posterolateral and apical posterolateral displacement was necessary to create more important and clinically significant MR(Table 1). Increasing LVP − LAP tended to increase RF (γLVP − LAP=−0.03 to +3.9 [%]; p < 0.001), even though that LVP − LAP caused a reduction of OLA. In four of five valves, RF was virtually unaffected bychanges of heart rate (γHR= 0.2 to 2.4 [%], p = NS). Only in one valve was RF significantly higher at heart rate of 120min−1than of 70 min−1(γHR= 5.1 [%], p < 0.001). However, RF may beaffected by the relative systolic and diastolic duration, which was not fully comparable in the two heart rate settings. In contrast, the early systolic precoaptationregurgitant volume was actually significantly reduced with increasing heart rate (γHR= −0.06 to −0.008 [liter]; p< 0.001, Model 1MP).
By application of the premise of the chordal force balance, incomplete mitral leaflet coaptation can be a result of: 1) abnormally increased tension caused by displacement ofthe annular (9)and papillary muscle attachment (3–8)and2) decreased global left ventricular systolic function, decreasing the coapting forces to close the mitral leaflets (5,7,18).
Our study emphasizes that redistribution of the chordal force balance leads to incomplete mitral leaflet coaptation.
Papillary muscle malalignment increased the chordal tethering forces and induced functional MR based on the tented leaflet configuration as observed clinically (3).
Increasing chordal tethering forces restricted leaflet mobility, impeded mitral valve coaptation and increased regurgitation. In addition, redirection of the chordal forcecomponents by a posterolateral shift of the papillary muscles caused a remodeling of the mitral leaflet surface area. This process can be interpreted as follows. Increasedtethering forces in posterior direction restricted posterior leaflet excursion angle. Leaflet separation was partly compensated by the anterior leaflet covering a larger part ofthe mitral valve orifice. Therefore, the mitral leaflet coaptation line moved in an apical and posterior direction in an asymmetrical coaptation pattern, which has also beenreported clinically (10). As a consequence, the coapting forces of the anterior leaflet increased due to exposure of a larger leafletsurface area. Thereby, excessive apical migration of the anterior leaflet was prevented. Conversely, a reduction of the posterior leaflet surface area due to leaflet infoldingcaused less coapting force to counterbalance increasing tethering forces, resulting in further leaflet malalignment (compare Fig. 7). Theanatomical substrate for anterior leaflet expansion may be unfolding of the leaflet tissue at the coaptation zone (rough zone) and stretching due to horizontal stress.
Lateralization of the chordal tethering forces might have exacerbated the degree of incomplete mitral leaflet coaptation by diverting the central parts of the leaflets awayfrom closure (8).
Recent clinical studies of functional MR have shown early and late systolic peaks in mitral regurgitant flow and regurgitant orifice area, which decreases in midsystolecoinciding with maximum transmitral pressure difference (19). A possible explanation for this phenomenon includes increased leftventricular pressure, which closes the leaflets more effectively. This mechanism was partly supported by our data, which demonstrated that increasing chordal coapting forces bychanging the left ventricular pressure reduced OLA. However, any favorable reduction of the regurgitant orifice area for a given subvalvular geometry was counteracted by increasingregurgitant flow velocities causing the observed increase of regurgitant volume to driving pressure.
Increasing peak + (LV − LA) dP/dt in early systole was associated with a reduction of the precoaptation regurgitant volume, indicating accelerated mitral leafletcoaptation and a relative delay of fluid momentum. However, increasing peak + (LV − LA) dP/dt did not cause more effective closure of the mitral leaflets (OLAinsensitive to peak + [LV − LA] dP/dt). Therefore, we do not consider a slow rate of rise of pressure, as speculated by Kaul et al. (5), to be a primary determinant of incomplete mitral leaflet coaptation and functional MR.
Using three-dimensional echocardiography to explore geometric determinants of functional MR in a canine model of global left ventricular dysfunction, Otsuji et al.(8)showed that MR occurred in relation to increased papillary muscle-to-annulus tethering length after papillary muscle displacementin lateral and posterior direction rather than in apical direction along the left ventricular long axis. In agreement with our findings, it was suggested that a lateral andposterior shift of the papillary muscles redirects papillary muscle tension away from the axial direction, diverting the leaflets away from closure. Our in vitro study elucidatedthis issue further by demonstrating decreased coapting forces of the posterior leaflet due to leaflet remodeling that enhanced leaflet malalignment.
Recent studies in our own laboratory have shown that annular dilation was a necessary condition to generate more significant levels of MR (13). However, annular dilation was not sufficient to develop the tented leaflet geometry without papillary muscle displacement. Therefore, in the present studywe kept mitral annular dimension constant to demonstrate that altered chordal force distribution accomplished by papillary muscle displacement reproduces tented mitral leafletconfiguration and functional MR as observed clinically.
Functional MR appears in an orchestra of geometric and hemodynamic variables that accompany acute and chronic left ventricular dysfunction. The chordal force distribution isthe key factor, which mediates the relative contribution and interaction of these variables, and thereby determines systolic mitral leaflet configuration.
Previous in vivo studies of these mechanistic factors of functional MR have provided data of limited value for two reasons. First, proposed causes of MR could not bedeliberately varied to show that they induced regurgitation in an initially competent valve (11). Second, competing hemodynamic andgeometric factors could not be readily separated in vivo (5). Our in vitro model allowed us to test these interactions; to show thatpostulated changes, in fact, can induce regurgitation in this model, and to demonstrate their overall similarity to clinical observations.
Normally, the mitral valve acts as an integrated system, with dynamic changes in the relation between the annulus and papillary muscle–chordae tendinae complexthat permit efficient closure throughout the systole (20). The model reproduced at different time steps of the cardiac cycle theinstantaneous papillary muscle and annular relationships that occur with akinetic or dyskinetic function of the left ventricular posterior wall (13). Although the dynamic changes of the left ventricle as a whole were not simulated, the pressure and flow changes were. The fixation of the papillarymuscles to mounting rods even allowed us to separate the chordal coapting forces and the tethering forces in the chordal force measurements, which would be impossible toaccomplish in vivo.
Even though any clinical implications of this model must be stated circumspectly, this in vitro study can potentially help us to gain insight into therapeuticinterventions that modify mechanistic factors, for example, the role of new surgical approaches that can modify the geometric relationships, such as chordal reconstruction(21), papillary muscle repositioning (20,22)and resection of posterior wallmyocardium between the papillary muscles to move them closer together (23). Our study indicates that remodeling of the mitral valvesubstructures (e.g., chordal lengthening) may be a relevant surgical approach to “normalize” the chordal force distribution in functional MR. However, onemay be aware that chordal lengthening itself, instead of “slackening the reins” of the mitral leaflets, may adversely deteriorate left ventricularfunction due to loss of support of the left ventricular wall (valvular–ventricular interaction). Therefore, surgical correction could involve lengthening of theprimary chordae, possibly in combination with transposition or shortening of the secondary (strut) chordae with the aim of improving mitral valve function and interaction withthe left ventricle. Further studies on this issue are in progress.
We thank the AV Department, Aarhus University Hospital for artistic illustrations.
☆ This project has been financially supported by grants from the Danish Heart Foundation (96-1-3-58-22370; 96-2-2-20-22417), The Danish Medical Research Council and theResearch Initiative of Aarhus University Hospital.
- Received July 27, 1998.
- Revision received September 18, 1998.
- Accepted November 18, 1998.
- American College of Cardiology
- Frantz E.,
- Weininger F.,
- Oswald H.,
- Fleck E.
- Godley R.W.,
- Wann L.S.,
- Rogers E.W.,
- Feigenbaum H.,
- Weyman A.E.
- Kaul S.,
- Spotnitz W.D.,
- Glasheen W.P.,
- Touchstone D.A.
- Kono T.,
- Sabbah H.N.,
- Rosman H.,
- Alam M.,
- Jafri S.,
- Goldstein S.
- Kono T.,
- Sabbah H.N.,
- Rosman H.,
- et al.
- Otsuji Y.,
- Handschumacher M.D.,
- Schwammenthal E.,
- et al.
- Boltwood C.M.,
- Tei C.,
- Wong M.,
- Shah P.M.
- Izumi S.,
- Miyatake K.,
- Beppu S.,
- et al.
- Salisbury P.F.,
- Cross C.E.,
- Rieben P.A.
- He S.,
- Fontaine A.A.,
- Schwammenthal E.,
- Yoganathan A.P.,
- Levine R.A.
- Brock R.C.
- Yacoub M.
- Littell R.C.,
- Milliken G.A.,
- Stroup W.W.,
- Wolfinger R.D.
- Schwammenthal E.,
- Chen C.,
- Benning F.,
- Block M.,
- Breithardt G.,
- Levine R.A.
- Frater R.W.M.,
- Cornelissen P.,
- Sisto D.
- Liel-Cohen N.,
- Guerrero J.L.,
- Otsuji Y.,
- et al.