Enhanced Regional Deformation at the Anterior Papillary Muscle Insertion Site After Chordal Transsection
Background Clinical and experimental studies of mitral valve replacement have shown a depression of ventricular function after chordal transsection; most recent studies have proposed that this is secondary to a depression of local function near the papillary muscle insertion site. However, there is no direct experimental evidence for changes in local fiber shortening in the wall of the left ventricle overlying the papillary muscle. Accordingly, we investigated the effect of chordal transsection on left ventricular shape and on three-dimensional regional deformation of the myocardium near the insertion of the anterior papillary muscle.
Methods and Results In six open-chest dogs, two sets of three transmural columns of radiopaque markers were implanted in the anterior wall, one set at the tip of the papillary muscle (basal) and one at the site of papillary muscle fiber insertion (apical). A Björk-Shiley mitral valve was placed in the left atrium adjacent to the native valve. Markers were then tracked with biplane cineradiography, and deformation was quantified with the use of finite strain analysis. Chordal transsection resulted in reduced left ventricular end-systolic pressure and slowed relaxation. After chordal transsection, outward displacement of the ventricular wall and transverse shearing deformation were observed in the area of the papillary muscle during isovolumic contraction. Circumferential and radial strains during ejection were maintained at our basal site and enhanced on our apical site.
Conclusions Chordal transsection led to enhanced local shortening and wall thickening and regional strain nonuniformity. These results indicate that chordal transsection induces an unloading of myocardium at the papillary muscle insertion site and that the resulting heterogeneity of regional function is the mechanism for the reduced global function and slowed ventricular relaxation.
The importance of the mitral apparatus to left ventricular function has been suggested in many clinical studies of mitral valve replacement.1 2 3 4 5 6 7 8 9 10 These studies documented that surgical mitral valve replacement with chordal preservation maintained postoperative ejection performance, while left ventricular ejection performance was reduced in patients undergoing mitral valve replacement with chordal transsection.1 2 7 9 Moreover, experimental studies11 12 13 14 15 16 have described a decrease in global left ventricular function after chordal transsection that supports the clinical findings. Several experimental13 14 17 and clinical studies5 7 8 18 that used echocardiography or piezoelectric dimension gauges to assess global left ventricular dimensions have reported a decline in segmental function in the area of the papillary muscle as well as a depression of global function. Indeed, Pitarys et al8 and Corin et al18 have shown that segmental function is reduced only in the area of the papillary muscle after chordal transsection. A variety of studies have also reported changes in global ventricular shape after chordal transsection.9 15 16 Others have shown that there is a local shape change in the left ventricular wall overlying the papillary muscle.8 18 Only one study to date has directly examined local dimension changes in the wall of the left ventricle overlying the papillary muscle: In this study Yun et al17 showed that preload recruitable stroke work calculated from left ventricular pressure and a local wall thickening gauge is depressed after chordal transsection.
Although the exact mechanism of the global depression of function observed after chordal transsection remains unclear, most evidence points to a local depression of function near the papillary muscle. Both loss of isovolumic recruitment of fiber length9 11 15 19 originally proposed by Rushmer20 and increased systolic stress secondary to local or global systolic shape changes have been proposed to explain this depression. Unfortunately, no study to date has directly examined local function at the site of the papillary muscle insertion before and after chordal transsection. This type of study seems essential to directly test the proposal that there is a local depression of function. Indeed, in the absence of direct myocardial injury this depression seems unlikely. Thus, in the present study, we examined the effects of transsection of the chordae of the anterior papillary muscle (APM) on the three-dimensional regional deformation of the myocardium near the insertion of the papillary muscle. This approach allowed us to directly measure strains across the left ventricular wall.21
In contrast to previous studies after chordal transsection, we found enhanced regional circumferential systolic shortening and increased wall thickening in the area of the papillary muscle insertion. These results support the hypothesis that after chordal transsection, myocardium near the papillary muscle is unloaded and local shortening is enhanced. We propose that the depression of global function observed after chordal transsection is secondary to local heterogeneity of function. This proposal is supported by the slowed ventricular relaxation observed in this and other studies.
The University of California, San Diego, is accredited by the American Association for Accreditation of Laboratory Animal Care (AAALAC). All experiments were conducted according to AAALAC guidelines for the use of animals in research and were approved by the local faculty animal use committee.
Six adult mongrel dogs weighing 27 to 35 kg were anesthetized with sodium pentobarbital, intubated, and ventilated with oxygen. Anesthesia was maintained with additional barbiturate injections (50 to 100 mg/h). In each study, the chest was opened via median sternotomy and left fourth intercostal space thoracotomy. The pericardium was opened and the heart suspended in a pericardial cradle. We elected to study the effects of sectioning of the anterior left ventricular chordae tendineae for several reasons. Hansen et al19 reported that the chordae of the anterior and posterior mitral leaflets had an additive influence on global left ventricular systolic performance, but the contribution of the anterior chordae was more important. Furthermore, Yun et al17 showed that the influence of the anterior chordae on local left ventricular systolic function was greater than that of the posterior chordae at both papillary muscle insertion sites. In addition, it is technically difficult to implant bead sets into the posterior wall of the left ventricle. To aid in locating the anterior papillary muscle and positioning marker beads, we inserted a finger into the left ventricle via the left atrial appendage. A purse-string suture (R-833 polyester fiber) was placed around the base of the left atrial appendage for hemostasis during insertion of the finger. The positions of the tip and insertion of the papillary muscle were outlined on the epicardial surface with a marking pen. Columns of gold beads (1-mm outer diameter, approximately 2-mm spacing) were inserted across the ventricular wall from the epicardial surface using a trocar. Three columns were placed at the tip of papillary muscle (basal site) and three at the papillary muscle insertion (apical site). After marker implantation, larger (1.6 mm) lead beads were sewn to the epicardial surface above each column, at the apex of the left ventricle (apex bead), and at the bifurcation of the left main coronary artery (base bead). A Konigsberg micromanometer was inserted into the left ventricular apex and matched to a fluid-filled, 120-cm 7F pigtail catheter passed from the femoral artery and positioned inside the left ventricle.
A second purse-string suture was placed to secure a Björk-Shiley mitral valve using the method of Ohtani et al.22 In brief, a No. 5 silk suture threaded on a semicircular (80-mm diameter), 16-gauge needle was passed below the coronary sinus through the right atrium, pushed out from the superior medial aspect of the right atrium, and anchored at the base of the left atrium. A 25-mm Björk-Shiley mitral valve secured within a grooved Teflon ring that was attached to a 2-mm ODX 40-cm long stainless steel rod was inserted into the left atrium through the proximal purse-string suture (polyester) and secured in place with the second purse-string suture (silk).
Three-dimensional bead positions were detected with biplane cineradiography. Our cameras recorded cinefilm asynchronously at 120 frames per second. Data were recorded with the respirator off at end expiration for approximately 5 to 8 seconds. During each run, the following were recorded on an 8-channel chart recorder (Gould Instruments): ECG, aortic pressure, left ventricular pressure, left ventricular dP/dt, and cinemarks for correlation of film frame and hemodynamic events. Satisfactory dP/dt data were obtained for five of the six dogs.
Before chordal transsection, we recorded two cine runs in a steady hemodynamic state at low (<8 mm Hg) and at high (≥8 mm Hg) left ventricular end-diastolic pressure (EDP) using warm homologous blood transfusion as needed to elevate EDP. Data obtained prior to chordal transsection at low EDP in four dogs in this study were reported previously23 ; data at high EDP matching the EDP after transsection are reported here. A control left ventricular injection of contrast (Hypaque, Wintrop Pharmaceuticals) via the pigtail catheter revealed that there was no regurgitation beyond the normal mitral valve. After the control data were recorded, the anterolateral papillary muscle chordae tendineae were cut using Love-Kerrison Rongeur Forceps (3-mm bite) inserted directly through the left ventricular wall (Fig 1A⇓). Warm homologous blood was infused to keep the left ventricular peak systolic pressure more than 90 mm Hg and to match the left ventricular end-diastolic pressure (within 3 mm Hg) with pretranssection values. After transsection, the forceps were removed and the left ventricular wall repaired with 2-0 silk mattress sutures. A repeat left ventriculogram revealed no regurgitation beyond the Björk-Shiley valve. We again recorded two cine runs in steady state at low and high EDP. After the study, the heart was fixed by high-pressure retrograde aortic root perfusion with buffered glutaraldehyde (2.5%), which closed the aortic valve and perfused the coronary vessels.24 All hearts were removed and stored in 10% buffered formalin (Fisher Scientific).
In postmortem examination, we confirmed that all the chordae tendineae attached to the APM were cut in each dog. Hearts were positioned on a cutting jig,21 and tissue blocks containing the markers and having edges parallel to the in situ circumferential and longitudinal directions were removed. In each block, the junction of the papillary muscle with the left ventricular wall was grossly identified and the blocks were carefully examined under fluoroscopy to determine the placement of markers relative to the papillary muscle border.
The three-dimensional coordinates of the implanted beads were reconstructed from the biplane images at end diastole, end isovolumic contraction, and end systole. End diastole was taken as the time of the peak QRS wave of the ECG. The micromanometer and aortic pressure profiles were carefully matched with the catheter in the left ventricle before it was withdrawn into the aortic root. Then, pressure at the nadir of the dicrotic notch from the fluid-filled aortic root catheter was used to estimate end-systolic pressure on the micromanometer tracing. In each selected beat, bead positions at aortic valve opening (deformed configuration) were compared with bead positions at end diastole (reference configuration) to calculate isovolumic contraction strain, and bead positions at end systole (deformed configuration) were compared with end isovolumic contraction (reference configuration) to calculate ejection strain. Moreover, bead positions at end diastole after chordal transsection (deformed configuration) were compared with those at end diastole before chordal transsection (reference configuration) to assess end-diastolic remodeling strain. Finite (large) deformation theory was used to calculate strains after the method described by Waldman et al.25 Groups of four beads formed tetrahedra with bases nearly parallel to the epicardial surface (three beads differing in depth by no more than 1.5 mm) and heights of between 1.9 and 4.1 mm. Holmes et al23 reported previously that a boundary between the myocardial wall and the papillary muscle could be distinguished grossly and was also revealed in muscle fiber angle measurements. Thus, we believed that the homogeneous strain theory was inappropriate at the papillary muscle border and excluded tetrahedra spanning the border between inner wall and papillary muscle from analysis. Tetrahedra within the myocardial wall were grouped into outer and inner halves of the wall at the basal and apical sites. Marker tetrahedra were used to calculate six independent finite strains. These strains were expressed in a local coordinate system derived from the apex and base markers and the three epicardial beads. In this system, the first axis is circumferential, the second axis is longitudinal, and the third axis is radial. The three normal strains describe shortening or lengthening along the circumferential (E11), longitudinal (E22), or radial (E33) axes. The three shear strains (E12, E13, E23) describe changes in angle between pairs of axes that were mutually perpendicular in the reference configuration. We used three distances as the indices of local shape change as shown in Fig 1B⇑. L indicated the distance from base to apex; L1 indicated the distance from base to a perpendicular from the deepest bead at the basal site; and S indicated the perpendicular distance from the base-apex axis to the deepest bead at the basal site.
Data from tetrahedra in the outer and inner halves of the wall in the area of APM insertion were averaged for each dog. All values are reported as mean±SD unless otherwise noted. Changes in regional strains and shape indices were analyzed using ANOVA with a repeated measures design (SuperANOVA version 1.1, Abacus Concepts). When significant differences were detected by ANOVA, contrasts were performed to determine which individual differences were statistically significant; in these cases, Bonferroni’s correction for multiple comparisons was used. End-diastolic remodeling strain data were compared with 0 using a one-sample t test (StatView, version 4.0, Abacus Concepts). An effect with P<.05 was considered statistically significant.
Hemodynamics and Local Shape Change
Hemodynamic indices before and after chordal transsection are shown in Table 1⇓. Heart rate did not change after chordal transsection. When compared at matched end-diastolic pressure, left ventricular end-systolic pressure decreased significantly after chordal transsection from 138±39 to 118±30 mm Hg. Although maximal dP/dt did not change, minimal dP/dt fell by an average of 20% after chordal transsection (P<.05). Diastolic changes were assessed by comparing dimensions at the same end-diastolic pressure before and after chordal transsection (Table 2⇓). Chordal transsection did not affect apex-base length (L) and S at end diastole. However, L1 at end diastole increased significantly by an average of 2.8 mm after chordal transsection, indicating a shift of the tip of the papillary muscle toward the apex. There were significant effects of chordal transsection on cardiac dimensional changes during systole. With chordae intact, dimensional changes were small during systole. After chordal transsection, apex-base length (L), L1, and S increased significantly during isovolumic contraction. During the ejection phase, there were no further changes in dimensions after chordal transsection.
Tetrahedra were selected and grouped by transmural location as described in “Methods.” There were 11±3 tetrahedra (range, 6 to 15) in the outer wall and 13±8 tetrahedra (range, 7 to 28) in the inner wall at the basal site. At the apical site, there were 10±9 tetrahedra (range, 2 to 26) in the outer wall and 12±13 tetrahedra (range, 2 to 38) in the inner wall. It was difficult to obtain tetrahedra that met our selection criteria with all four beads in the papillary muscle at either apical or basal sites. We were able to achieve this in only three hearts and have not included these data here.
The average systolic strain data at each site and each depth at baseline and after chordal transsection are shown in Tables 3⇓ and 4⇓ and Fig 2⇓. In analyzing the data, we have separated the total systolic strain into two components: isovolumic and ejection strains. Isovolumic strain describes the deformation occurring between end diastole (reference configuration) and aortic valve opening (deformed configuration). Ejection strain describes the deformation occurring between aortic valve opening (reference configuration) and aortic valve closure (deformed configuration).
Isovolumic Contraction Strains
Although there were some significant longitudinal (difference between basal and apical site) and transmural (between outer and inner wall) variations in isovolumic strains detected with ANOVA, isovolumic contraction strains prior to chordal transsection were quantitatively small and represented only a small portion of total systolic strain.
After chordal transsection, there were large increases in longitudinal lengthening (+E22) at the inner wall at both sites and twofold to threefold increases in positive longitudinal-radial shear in the inner wall. These changes are shown in Table 3⇑ and summarized in Fig 2⇑.
During control, wall thickening strain (E33) and longitudinal-radial shear strain (E23) were both greater in the inner wall than the outer wall. There were no significant differences of ejection strains between the basal and apical sites during control.
After chordal transsection, inner wall circumferential shortening strain (E11) and radial strain (E33) at the apical site were greater than at the basal site. There was a transmural increase in E11 and E33 only at the apical site (Table 4⇑).
In summary, with rare (and quantitatively small) exceptions, systolic finite strains (both isovolumic and ejection) in the outer wall did not change with chordal transsection. This was not the case for the inner wall, where there were significant changes in systolic strains with chordal transsection as well as regional differences. During isovolumic contraction, all strains and all changes with chordal transsection were small except longitudinal-radial shear strain (E23). Large positive values of E23 in the inner wall during isovolumic contraction were observed after chordal transsection. During the ejection phase, there were no significant differences between basal and apical sites in the control condition. However, chordal transsection influenced ejection strains differently at the basal and apical sites. Both circumferential shortening strain (E11) and wall thickening strain (E33) were enhanced at the apical site and unchanged at the basal site (Fig 2⇑).
End-Diastolic Remodeling Strains
We found a small shift in the position of the papillary muscle (only L1 lengthening, from 31.4±4.8 to 34.2±6.1 mm) at end diastole after chordal transsection. However, chordal transsection caused no significant remodeling deformation within the tissue (remodeling strains E11, E22, and E33 were not significantly different from zero in a one-sample t test at matched end-diastolic pressure) (Table 5⇓).
The importance of the mitral apparatus to left ventricular function has been suggested in several clinical studies1 2 7 9 that showed that mitral valve replacement with chordal transsection depressed postoperative ejection performance compared with mitral valve replacement with chordal preservation. In the present study, sectioning of the chordae tendineae to the anterior papillary muscle in the dog resulted in a decease in left ventricular systolic pressure. This decrease in systolic pressure was associated with a local systolic outward displacement of the ventricular wall in the area of the papillary muscle and enhanced circumferential shortening and wall thickening in the myocardium overlying the papillary muscle insertion site. Although previous studies have shown bulging in the wall over the papillary muscle, this is the first study to document enhanced function and transverse shearing deformation in the inner wall at the site of the papillary muscle insertion and a shift in the diastolic position of the papillary muscle. Since systolic wall motion abnormalities produced, for example, by changing the local activation sequence have been shown to produce decreases in pressure generation,26 27 28 we propose that the alteration of performance that occurs with chordal section is due to heterogeneity of shortening caused by local “unloading” of the myocardium at the papillary muscle insertion site and not (as suggested by previous studies) due to a local depression of regional function.
Effects of Chordal Transsection During Diastole
To examine whether there were changes in the diastolic configuration of the papillary muscle, marker positions were compared before and after chordal transsection at the same end-diastolic pressure. At end diastole, we found that the dimension L1 increased significantly after chordal transsection. This motion of the papillary muscle toward the apex is consistent with severing the connection between the papillary muscle and the base of the heart and suggests that the chordae tendineae are normally under tension even during diastole (Fig 3⇓).
Earlier studies reported that base-apex distance at end diastole increased significantly after chordal transsection,14 15 16 29 but we did not find a significant change in L. This difference may be explained by the fact that only the anterior chordae were transsected in our study, while all chordae attached to the anterior and posterior papillary muscles were disrupted in earlier studies. We found no significant change in the perpendicular distance from the base-apex axis to the deepest bead (S) at end diastole. Salter et al29 and Sarris et al14 also found that the minor-axis end-diastolic dimension did not change after chordal transsection. Shintani and Glantz16 reported that septum–free wall and anteroposterior left ventricular dimensions at end diastole increased significantly after chordal disruption, but EDP was allowed to increase in their study. Yun et al15 found that the end-diastolic septal-lateral axis dimension increased significantly, but anterior-posterior axis dimension did not change significantly (P=.06) at matched EDP. However, even these latter two studies found a more ellipsoidal shape at end diastole despite the increased minor axis dimension.15 16 On the basis of our results and the ellipsoidal shape change in earlier studies, we conclude that the end-diastolic shape change after chordal transsection occurs primarily along the major axis. Our results indicate that the majority of this change occurs at the base of the heart between the papillary muscle insertion and the mitral annulus. This shape change after chordal transsection may be caused by the loss of the tethering effect of the chordae, which are normally parallel to the base-apex axis when the mitral valve is open (Fig 3⇑). We found no significant end-diastolic remodeling strains after chordal transsection. This would indicate that the increase in L1 reflected a transposition (rigid body motion) of the papillary muscle without deformation of the overlying inner wall tissue. This also supports our impression that the tip of the papillary muscle is not well integrated with the overlying myocardium. Moreover, the absence of significant diastolic remodeling strains argues against diastolic changes in local muscle fiber length as the cause of changes in regional function after chordal transsection.
Effects of Chordal Transsection During Isovolumic Contraction
We found that the shape of the left ventricular wall in the area of papillary muscle insertion changed substantially during isovolumic contraction. Increased base-apex distance and local bulging were observed at end isovolumic contraction after chordal transsection (Fig 4⇓), and these changes continued to end systole. If chordae tendineae tension increased rapidly during isovolumic contraction, the loss of this force after chordal transsection would explain the shape changes we observed. Both lateral bulging and increased base-apex distance are consistent with the orientation of the chordae when the mitral valve is closed.
We found significant isovolumic longitudinal stretching (+E22) and shear deformation (E23) in the inner wall at both basal and apical sites after chordal transsection. Longitudinal stretching (+E22) during isovolumic contraction was consistent with significant lengthening of the base-apex distance. During isovolumic systole before chordal transsection, we found a negative longitudinal-radial shear strain (E23) in the inner wall at the apical site and essentially no shear at the basal site. However, we found large positive values of E23 during isovolumic systole after chordal transsection, consistent with observed motion of the papillary muscle toward the apex (Fig 2⇑ inset).
Effects of Chordal Transsection on Strains During Ejection
Circumferential shortening (E11) and wall thickening (E33) were enhanced after chordal transsection at the apical site, but no changes were observed at the basal site. We did not find any decline of ejection strains near the papillary muscle after chordal transsection despite systolic bulging in this area.
Because global left ventricular function and regional segmental wall motion were depressed after chordal transsection, earlier studies speculated that a loss of inward force generated by attachment of the papillary muscle to the annulus may decrease circumferential shortening.5 9 12 15 Rushmer20 speculated that in the normal heart, the early movement of the atrioventricular ring toward the apex produced by papillary muscle shortening during isovolumic contraction may increase preload in circumferential midwall fibers and enhance circumferential strain during the ejection phase via the Frank-Starling mechanism. A loss of this early systolic shape change after chordal transsection would induce a decline in circumferential strain during ejection. However, this hypothesis is not consistent with our data. We did not find any circumferential remodeling strain or change of circumferential strain during isovolumic contraction that might affect ejection strain via the Frank-Starling mechanism, nor did we observe a decline of ejection strains. Indeed, our data indicate that regional function in the area of papillary muscle insertion may be enhanced by the unloading effect caused by a loss of the connection of the tip of the papillary muscle to the base of the heart after chordal transsection. Our results appear inconsistent with earlier studies describing a reduction of regional left ventricular function in the area of papillary muscle insertion after chordal transsection.13 14 17 Hansen et al13 and Sarris et a14 evaluated regional systolic function at the base of the papillary muscle using the regional end-systolic pressure-dimension relation and concluded that there was a decline of regional end-systolic elasticity. Yun et al17 evaluated segmental left ventricular function using the slope of the segmental stroke work–end-diastolic wall thickness relation and found a decline in segmental function not only in the areas subtending papillary muscle insertions but also in remote left ventricular regions. However, all of these studies used uniaxial measurements to examine regional deformation. Since we observed large changes in transverse shear with chordal transsection, it is quite likely that the uniaxial measurements were affected by the shearing deformation.21 23 For example, the positive E23 shears observed at both sites would be reflected in the dimension measurements of Hansen et al and Sarris et al as a chord lengthening. We did not find a change in ejection strains at the base of the papillary muscle although strains at our apical site were enhanced. These data support our hypothesis that it is regional heterogeneity of function that is associated with the decrease in global left ventricular function. Because ejection pressure was depressed after chordal transsection, a part of the enhancement of circumferential shortening and wall thickening may derive not only from the direct unloading at the papillary muscle insertion site but also from reduction of afterload. However, the unloading effect was limited to the apical site, which argues against a global effect of pressure reduction.
To explain the different effect of chordal transsection at the apical papillary muscle site, we considered the muscular architecture of the left ventricle in the area of the papillary muscle insertion.30 At the apical site, papillary muscle fibers insert to all layers. In contrast, papillary muscle fiber insertions at the basal site are much shallower and are limited to inner wall, more longitudinal fibers. We propose that the deep insertion of papillary muscles at the apical site enhanced the unloading effect of chordal transsection.
Cobbs et al31 described in detail the pathological findings among seven patients who died after mitral valve replacement and suggested that spontaneous rupture of the left ventricle could occur in some patients after excision of the mural leaflet and the papillary muscles. In that study, they speculated that there was abnormal longitudinal traction on the left ventricular endocardium that sometimes separated along natural lines of cleavage or foci of minor trauma after mitral valve replacement. Based on our results, division of chordae tendineae enhanced the longitudinal stretching in the inner wall over the papillary muscle and radial-longitudinal shear (pulling down the tip of papillary muscle toward the apex) during isovolumic contraction. Both these factors would tend to increase the chance of endocardial injury in the basal areas. In clinical settings, left ventricular wall stress is considerably greater in the chronically dilated heart particularly in patients with marginal cardiac reserve.32 Moreover, the shape change during systole after mitral valve replacement in clinical patients may be greater than in our results, because all chordae tendineae attached to both anterior and posterior papillary muscles are disrupted. These factors would all work to enhance the possibility of endocardial injury.
Effect of Chordal Transsection on Relaxation
Peak negative dP/dt was reduced after chordal transsection, a finding consistent with earlier studies.14 16 The time constant of left ventricular relaxation is increased with wall motion asynchrony or nonuniformity of regional ventricular function.33 34 Therefore, Shintani and Glantz16 speculated that wall motion asynchrony or nonuniformity of left ventricular function produced after chordal transsection slowed left ventricular relaxation. Local hyperkinesia, nonuniformity of regional ejection strains, and the reduction of minimal dP/dt after chordal transsection in our study support their speculation.
Shintani and Glantz16 also reported a decrease of maximal dP/dt; this result is not consistent with our observation that maximal dP/dt did not change significantly. The differential effect of chordal uncoupling on maximal and minimal dP/dt may reflect the fact that only the anterior chordae were transsected in our study, while all chordae were disrupted in Shintani and Glantz’s work. Sarris et al14 reported that the percent change in maximal dP/dt (−19%) after chordal transsection was smaller than the change in minimal dP/dt (−32%), in agreement with our results. Moreover, Yun et al17 showed that transsection of only the anterior chordae did not change maximal dP/dt despite a significant decrease of maximal dP/dt after transsection of both anterior and posterior chordae tendineae.
Several potential limitations of this study should be considered. First, we used open-chest dogs in which the pericardium was also widely opened. Both these factors might enhance the magnitude of the changes seen in the current study. The depression of function observed with anesthesia might enhance the gradients in function observed. All chordae attached to both anterior and posterior papillary muscles are disrupted in standard mitral valve replacement. In these clinical settings, regional heterogeneity of function may be enhanced or reduced compared with our study.
Another potential difficulty is the method of cutting the chordae tendineae. It is possible that the ring supporting the mitral prosthesis could have restrained normal motion of the mitral annulus. However, the ring in this case is anchored in atrial tissue (≈5 mm from the annulus) and seems unlikely to restrain annular motion. Moreover, Rayhill et al35 have shown that the presence of rigid ring fixation does not influence left ventricular systolic performance. Love-Kerrison Rongeur Forceps were inserted through the anterior wall into the left ventricle. Although this procedure caused some myocardial injury, we did not find a significant reduction of any ejection strains; indeed, strains were enhanced at the apical site and unchanged at the basal site. However, since the site of injury was basal to both sites, it is possible that the injury enhanced the changes in inner wall longitudinal shortening. We think that this is unlikely, since the injury was somewhat lateral to the site of measurement and was also transmural. No changes in outer wall strains were observed.
Transsection of the chordae to the anterior papillary muscle in a preparation where mitral regurgitation was prevented led to a systolic bulge in the area of the muscle and a diastolic apical displacement of the tip of the papillary muscle. These changes were accompanied by enhanced systolic deformation at the papillary muscle insertion site without changes at an adjacent site. The data support the concept that the chordae are under tension at all times including diastole. Moreover, the results indicate that the loss of ventricular function after chordal transsection is due to heterogeneity of regional wall motion at the site of insertion of papillary muscle fibers into the ventricular wall, and not to local depression of regional function as implied by earlier studies.
This study was supported by grants from the National Heart, Lung, and Blood Institute (HL-32583), the Health Research Council of New Zealand, and the Whitaker Foundation.
- Received March 28, 1995.
- Revision received August 10, 1995.
- Accepted September 18, 1995.
- Copyright © 1996 by American Heart Association
David TE, Uden DE, Strauss HD. The importance of the mitral apparatus in left ventricular function after correction of mitral regurgitation. Circulation. 1983;68(suppl II):II-76-II-82.
David TE, Ho WC. The effect of preservation of chordae tendineae on mitral valve replacement for postinfarction mitral regurgitation. J Thorac Cardiovasc Surg. 1986;74(suppl I):I-116-I-120.
Rozich JD, Carabello BA, Usher BW, Kratz JM, Bell AE, Zile MR. Mitral valve replacement with and without chordal preservation in patients with chronic mitral regurgitation. Circulation. 1992;86:1718-1726.
Tischler MD, Cooper KA, Rowen M, LeWinter MM. Mitral valve replacement versus mitral valve repair. Circulation. 1994;89:132-137.
Hansen DE, Cahill PD, DeCampli WM, Harrison DC, Derby GC, Mitchell RS, Miller DC. Valvular-ventricular interaction: importance of the mitral apparatus in canine left ventricular systolic performance. Circulation. 1986;73:1310-1320.
Sarris GE, Fann JI, Niczyporuk MA, Derby GC, Handen CE, Miller DC. Global and regional left ventricular systolic performance in the in situ ejection canine heart. Circulation. 1989;80(suppl I):I-24-I-42.
Yun KL, Niczyporuk MA, Sarris GE, Fann JI, Miller DC. Importance of mitral subvalvular apparatus in terms of cardiac energetics and systolic mechanics in the ejecting canine heart. J Clin Invest. 1991;87:247-254.
Shintani H, Glantz SA. Effect of disrupting the mitral apparatus on left ventricular function in dogs. Circulation. 1993;87:2001-2015.
Yun KL, Fann JI, Rayhill SC, Nasserbakht F, Derby GC, Handen CE, Bolger AF, Miller DC. Importance of the mitral subvalvular apparatus for left ventricular segmental systolic mechanics. Circulation. 1990;82:(suppl IV):IV-89-IV-104.
Rushmer RF. Initial phase of ventricular systole: asynchronous contraction. Am J Physiol. 1956;184:188-194.
Waldman LK, Nosan D, Villarreal FJ, Covell JW. Relation between transmural deformation and local myofiber direction in canine left ventricle. Circ Res. 1988;63:550-562.
Ohtani M, Nikolic SD, Glantz SA. A new approach to volume clamping in dogs. Am J Physiol. 1991;261:H1335-H1343.
Holmes JW, Takayama Y, LeGrice I, Covell JW. Depressed regional deformation near anterior papillary muscle. Am J Physiol. 1995;269(Heart Circ Physiol. 38):H262-H270.
Yoran C, Covell JW, Ross JR. Rapid fixation of the left ventricle: continuous angiographic and dynamic recordings. J Appl Physiol. 1973;35:155-157.
Waldman LK, Fung Y, Covell JW. Transmural myocardial deformation in the canine left ventricle: normal in vivo three-dimensional finite strains. Circ Res. 1985;57:152-163.
Boerth RC, Covell JW. Mechanical performance and efficiency of the left ventricle during ventricular stimulation. Am J Physiol. 1971;221:H1686-H1691.
Badke FR, Boinay P, Covell JW. Effects of ventricular pacing on regional left ventricular performance in the dog. Am J Physiol. 1980;238:H858-H867.
Waldman LK, Covell JW. Effects of ventricular pacing on finite deformation in canine left ventricles. Am J Physiol. 1987;252:H1023-H1030.
Salter DR, Pellom GL, Murphy CE, Brunsting LA, Goldstein JP, Morris JM III, Wechsler AS. Papillary-annular continuity and left ventricular systolic function after mitral valve replacement. Circulation. 1986;74:(suppl I):I-121-I-129.
Yin FCP. Ventricular wall stress. Circ Res. 1981;49:829-842.
Aoyagi T, Iizuka M, Takahashi T, Ohya T, Serizawa T, Momomura S, Sato H, Mochizuki T, Matsui H, Ikenouchi H, Shin I, Ma Y, Sugimoto T. Wall motion asynchrony prolongs time constant of left ventricular relaxation. Am J Physiol. 1989;257:H883-H890.
Lew WYW, Rasmussen CM. Influence of nonuniformity on the rate of left ventricular pressure fall in the dog. Am J Physiol. 1989;256:H222-H232.
Rayhill SC, Castro LJ, Niczyporuk MA, Ingels NB Jr, Daughters GT, Derby GC, Tye TL, Bolger AF, Miller DC. Rigid ring fixation of the mitral annulus does not impair left ventricular function in the normal canine heart. Circulation. 1992;86(suppl II):II-26-II-38.