Functional mitral regurgitation (MR) at different phases of the regurgitant period may respond differently to cardiac resynchronization therapy (CRT). The aims of this study were to examine the impact of CRT on the phasic changes of MR (early vs late systole) and to explore the mechanisms of such changes.
Instantaneous MR flow rate and total MR volume were evaluated in 60 patients who had more than mild functional MR before and 3 months after CRT. In addition, indices of global left ventricular (LV) remodeling, mitral deformation, and LV systolic dyssynchrony were assessed.
CRT diminished MR volume (38 ± 18 vs 32 ± 20 mL) by reducing both the early (72 ± 47 vs 58 ± 48 mL/sec) and late (48 ± 42 vs 40 ± 42 mL/sec) systolic components (all p values < .01). In patients with ≥10% reductions in total MR volume but not in patients without this improvement, there were significant reductions in LV end-systolic volume, increases in LV +dP/dt, decreases in mitral valvular tenting, and improvements of systolic dyssynchrony at 3 months (all P values < .05). By multivariate regression, the reductions in LV end-systolic volume and tenting area were independent determinants of a reduction in total MR volume: the reductions in LV end-systolic volume and global dyssynchrony determined the reduction in early systolic MR, and the reductions in tenting area and global dyssynchrony determined reduction in late systolic MR.
CRT decreases MR volume by reducing both early and late systolic MR. The determinants of the phasic improvement in functional MR are different.
Functional mitral regurgitation (MR) is a common finding in patients with chronic left ventricular (LV) systolic dysfunction and is associated with poor prognosis. The mechanisms of functional MR are multifactorial. Although LV remodeling is considered a prerequisite for functional MR, mitral deformation caused by subvalvular traction secondary to LV remodeling seems to be the predominant reason. Also, LV systolic dyssynchrony and reduction of LV contractility might contribute to the pathogenesis of functional MR.
Cardiac resynchronization therapy (CRT) is an established therapy for patients with advanced chronic heart failure and prolonged QRS duration. Where myocardial structure and function are concerned, apart from enhancement of LV systolic function and LV reverse remodeling, improvements of MR have also been observed in many studies. Regarding the possible mechanisms of MR improvement after CRT, previous studies have suggested that improvement of functional MR after CRT is related to the increase in mitral valve coaptation force and reduction of papillary muscle dyssynchrony. However, the pattern of MR is more complex and not homogenous throughout the cardiac cycle. In fact, MR has a biphasic profile, with early and late systolic peaks and a midsystolic period of minimal regurgitant flow. Furthermore, early and late systolic MR may respond differently to CRT, because their pathogenesis is contributed to by different factors. Therefore, the assessment of changes in MR after CRT and the determination of related mechanisms must take this into consideration. The aim of the present study was to examine the improvements of early and late systolic MR after CRT, in particular to explore the potential determinants for the improvement in both components of functional MR.
From January 2004 to December 2008, 106 consecutive patients who had current indications for CRT received device implantation and finished follow-up of 3 months (i.e., evidence of systolic dysfunction with LV ejection fraction ≤ 35% despite optimal medical therapy, New York Heart Association class III or IV, and QRS duration > 120 ms). Those with morphologic abnormalities of the mitral valve apparatus (including papillary muscle elongation or retraction after myocardial infarction), clinical or echocardiographic evidence of other structural cardiac diseases, atrial flutter or fibrillation, greater than mild aortic regurgitation, or less than mild MR were excluded. Consequently, 60 patients were included in the present study. Among them, 48 patients were shared in the analysis of a recently published report in which the impact of prepacing MR and MR changes on LV reverse remodeling after CRT was investigated. The atrioventricular interval was optimized using Ritter’s method on day 1 after implantation. Clinical and echocardiographic assessments were performed before and 3 months after CRT implantation. The study protocol was approved by the ethics committee of the institution, and written informed consent was obtained from all patients.
Echocardiographic evaluation (Vivid 7 and EchoPAC PC version 7.0.0; GE Vingmed Ultrasound AS, Horten, Norway) included quantification of MR, LV function, and LV global remodeling; assessment of mitral valvular deformation; and examination of LV global and regional dyssynchrony. All analyses were performed in a blinded fashion by a single observer. At least three consecutive cardiac cycles were measured, and the average value was obtained for each parameter.
Quantification of Functional MR
The total MR volume was calculated by the quantitative Doppler method as previously described. Briefly, the total MR volume was obtained as the mitral stroke volume minus the LV outflow stroke volume. The mitral stroke volume was determined as the product of mitral annular area and the mitral time-velocity integral during diastole, while the LV outflow stroke volume was calculated by multiplying the LV outflow tract area and LV outflow velocity-time integral at systole.
Early and late systolic MR was estimated by the instantaneous MR flow rate at early and late systole, which was calculated using the proximal isovelocity surface area method. The dynamic change in the severity of MR was calculated by frame-by-frame analysis throughout the regurgitant period from mitral valve opening to mitral valve closure, where the first and last third were defined as early and late systole, respectively. The images at the timing of the maximum proximal flow convergence area at early and late systole were selected for assessing the instantaneous MR flow rate. The proximal flow convergence area of MR was obtained from the apical views in color Doppler flow images, with the Doppler color gain carefully adjusted to maximize signal without random noise velocities. The maximal radius ( r ) of the proximal flow convergence area was measured after a baseline shift of color flow to decrease the aliasing velocity ( v ). The MR flow rate was then calculated as 2π vr 2 .
At 3 months after CRT, a reduction of ≥10% in MR volume was considered a significant improvement in MR. Similarly, ≥10% reductions in early and late systolic MR flow rates were regarded as improvements in early and late systolic MR, respectively.
LV Global Remodeling and Systolic Function
LV end-diastolic volume, LV end-systolic volume (LVESV), and LV ejection fraction were calculated from the apical four-chamber and two-chamber views, using the biplane Simpson’s method. LV cavity length and width were measured in apical four-chamber view at end-systole, while LV sphericity index was estimated as the ratio of LV length to LV width. Doppler-derived LV +dP/dt was estimated by measuring the time interval between 1 and 3 m/sec on the MR continuous-wave Doppler spectrum using a sweep speed of 100 mm/sec.
Mitral Valvular Deformation
Mitral valvular deformation indices included valvular tenting area and tenting height. The mitral valve tenting area was measured by tracing between the atrial surface of the leaflets and the mitral annular plane from the parasternal long-axis view at midsystole. In addition, the tenting height was scaled vertically from the coaptation point to the mitral annular plane.
Global and Regional Dyssynchrony
LV mechanical dyssynchrony was assessed by two-dimensional color Doppler tissue imaging in three apical views (apical four-chamber, two-chamber, and three-chamber views). Myocardial velocity curves were reconstituted offline using the six-basal, six-mid segmental model of the left ventricle, as previously described, and the time to the peak systolic velocity during ejection phase (Ts) was measured in each segment. The maximal difference of Ts among the 12 LV segments (Ts-Dif) was calculated as the parameter for global systolic dyssynchrony, while the absolute difference in Ts between the midlateral and midinferior segments was taken to reflect regional dyssynchrony between the anterolateral papillary muscle (APM) and posteromedial papillary muscle (PPM) delay at their attaching sites.
Continuous variables are expressed as mean ± SD. Categorical data are summarized as frequencies and percentages. Paired t tests or Wilcoxon’s signed-rank tests were used as appropriate in comparisons between baseline and after CRT. Unpaired t tests or χ 2 tests were used as appropriate in comparisons between patients with and without significant improvements in total MR after CRT. Determinants of improvement in total, early systolic, and late systolic MR were identified by univariate logistic regression, followed by backward stepwise multivariate logistic regression. A P value < .05 was considered statistically significant.
Improvements in Total Functional MR and Its Early and Late Systolic Components After CRT
Baseline characteristics of the study patients are listed in Table 1 . At 3 months after CRT, significant reductions in total MR volume (38 ± 18 vs 32 ± 20 mL, P < .001), as well as the early systolic (72 ± 47 vs 58 ± 48 mL/sec, P < .001) and late systolic (48 ± 42 vs 40 ± 42 mL/sec, P = .007) components of MR were observed ( Table 2 ). Interestingly, both the baseline (72 ± 47 vs 48 ± 42 mL/sec, P < .001) and improvement (−14 ± 23 vs −7 ± 20 mL/sec, P < .001) in functional MR was greater in early than in late systolic MR. The mean reduction in total MR volume was 18.1 ± 32.0%, and significant improvements in total MR were found in 32 patients (53.3%) using the cutoff value of ≥10%. Similarly, significant improvements in early and late systolic MR were identified in 32 (53.3%) and 24 (40%) patients, respectively. As a result, improvements of both early and late systolic MR were observed in 18 patients (30%), improvements of only early systolic MR in 15 (25%), improvements of only late systolic MR in 6 (10%), and a lack of improvement in either phase in 21 (35%). With regard to the etiology of heart failure, though no significant difference was observed in baseline total MR volume and its early and late systolic components (all P values > .05), at 3 months after CRT, the total MR volume (42 ± 19 vs 36 ± 21 mL, P = .001) as well as the early systolic (79 ± 47 vs 64 ± 50 mL/sec, P = .002) and late systolic (55 ± 43 vs 47 ± 45 mL/sec, P = .034) MR all decreased significantly in nonischemic patients, while only the total MR volume (33 ± 17 vs 27 ± 19 mL, P = .010) and early systolic MR (63 ± 46 vs 51 ± 45 mL/sec, P = .002) decreased, with late systolic MR (39 ± 39 vs 33 ± 38 mL/sec, P = .105) unchanged in ischemic patients.
|Age (years)||66 ± 12|
|ACE inhibitors or ARBs||90%|
|Entire population||Patients with improvement in total MR||Patients without improvement in total MR ∗|
|( n = 60)||( n = 32)||( n = 28)|
|Parameter||Baseline||3 Months||P||Baseline||3 Months||P||Baseline||3 Months||P|
|MR volume (mL)||38 ± 18||32 ± 20||<.001||35 ± 16||22 ± 14||<.001||41 ± 21||43 ± 24||.096|
|Early systolic MR flow rate (mL/sec)||72 ± 47||58 ± 48||<.001||67 ± 36||40 ± 30||<.001||77 ± 57||78 ± 56||.819|
|Late systolic MR flow rate (mL/sec)||48 ± 42||40 ± 42||.007||41 ± 31||25 ± 25||.001||55 ± 51||58 ± 51||.021|
|LVEDV (mL)||181 ± 56||168 ± 62||<.001||169 ± 44||147 ± 47||<.001||195 ± 65||192 ± 68||.501|
|LVESV (mL)||135 ± 48||114 ± 52||<.001||126 ± 41||93 ± 42||<.001||145 ± 54||137 ± 53||.005|
|LV ejection fraction (%)||27 ± 6||34 ± 8||<.001||27 ± 6||38 ± 8||<.001||27 ± 6||30 ± 7||.004|
|LV +dP/dt (mm Hg/sec)||695 ± 258||808 ± 291||.001||675 ± 268||885 ± 333||<.001||716 ± 252||732 ± 224||.635|
|LV cavity length (cm)||8.2 ± 0.9||8.0 ± 0.9||.066||8.1 ± 0.7||7.8 ± 0.9||.002||8.2 ± 1.1||8.1 ± 0.9||.904|
|LV cavity width (cm)||5.0 ± 0.7||4.6 ± 0.8||<.001||4.9 ± 0.7||4.4 ± 0.8||<.001||5.2 ± 0.8||5.0 ± 0.7||.031|
|LV sphericity index||1.6 ± 0.2||1.7 ± 0.2||<.001||1.7 ± 0.2||1.8 ± 0.3||.001||1.6 ± 0.2||1.7 ± 0.2||.114|
|Tenting area (cm 2 )||2.3 ± 0.8||2.1 ± 0.8||<.001||2.3 ± 0.6||1.9 ± 0.6||<.001||2.2 ± 0.9||2.2 ± 0.8||.559|
|Tenting height (cm)||1.1 ± 0.3||1.0 ± 0.3||<.001||1.1 ± 0.3||1.0 ± 0.3||.001||1.1 ± 0.3||1.1 ± 0.3||.099|
|APM-PPM delay (ms)||62 ± 36||63 ± 37||.858||70 ± 32||49 ± 34||.003||51 ± 39||79 ± 34||.006|
|Ts-Dif (ms)||103 ± 39||91 ± 41||.038||107 ± 39||76 ± 35||<.001||109 ± 41||98 ± 38||.091|
∗ P <.05 compared with patients with improvement in total MR.
As shown in Table 2 , LV reverse remodeling was evident at 3 months, with reductions in LV volumes and gains in ejection fractions (all P values < .001). Improvement in LV geometry was achieved by a greater extent of decrease in LV width than in LV length, which resulted in higher values of sphericity index (i.e., a less globular shape; all P values < .001). Improvement in mitral valvular deformation was observed, as reflected by reductions in tenting area and tenting height (both p values < .001). In addition, LV global dyssynchrony decreased after CRT ( P = .038) ( Table 2 ).
Comparisons Between Patients With and Without Improvements in MR After CRT
According to the percentage reduction in total MR volume after CRT, the 60 patients were divided into two groups: those with improvements in total MR defined by a reduction of ≥10% (group 1, n = 32) and those without (group 2, n = 28). The etiology of heart failure was identical between the two groups (percentage of ischemic patients, 46.4 vs 46.9%, χ 2 = 0.001, P = 1.000). Table 2 shows the comparisons between the two groups before and after CRT. Despite no differences in baseline total MR volume as well as early and late systolic MR flow rate, group 1 showed reductions in all three MR parameters by 40% ( P < .001), 42% ( P < .001), and 39% ( P = .001), respectively. However, group 2 showed a significant increase in late systolic MR flow rate by 10% ( P = .021), with no change in total MR volume and early systolic MR flow rate ( Table 2 ). In group 1, improvements of both early and late systolic MR were observed in 18 patients (56.3%), improvements of only early systolic MR in 11 (34.4%), and improvements of only late systolic MR in three (9.4%), while none of the patients showed an absence of improvement in both components. On the contrary, these figures were zero (0%), four (14.3%), three (10.7%), and 21 (75%) patients, respectively, in group 2 (group 1 vs group 2, χ 2 = 42.19, P < .001).
The baseline echocardiographic parameters of LV volume, geometry, contractility, and mitral valvular deformation were comparable between the two groups ( Table 2 ). At 3-month follow-up, group 1 showed evidence of LV reverse remodeling, with significant decreases in both LVESV and LV end-diastolic volume (both P values < .001) and a dramatic increase in ejection fraction by 11% (absolute value; P < .001). In contrast, there was no reduction in LV end-diastolic volume, only a mild reduction in LVESV and an increase in ejection fraction by 3% ( P = .004) in group 2. Most of the other echocardiographic parameters assessed were improved only in group 1, which included a decrease in LV cavity length ( P = .002), an increase in sphericity index ( P = .001), an increase in +dP/dt ( P < .001), and a decrease in mitral valve tenting area ( P < .001) and tenting height ( P = .001) ( Table 2 ). For dyssynchrony assessment, parameters for regional dyssynchrony of APM-PPM delay at baseline had a trend of being greater in group 1 than group 2 (70 ± 32 vs 51 ± 39 ms), although this did not reach statistical significance. Interestingly, both Ts-Dif ( P < .001) and APM-PPM delay ( P = .003) were reduced significantly after CRT in group 1, indicating an improvement of systolic dyssynchrony. On the other hand, the APM-PPM delay was further prolonged after CRT ( P = .006) in group 2.
Determinants of Improvement in Total, Early, and Late Systolic MR After CRT
Determinants of improvement in total MR (i.e., ≥10% reduction) at 3 months after CRT were examined, in which changes in LV volume, geometry, LV +dP/dt, mitral valvular deformation, and global and regional measures of LV systolic dyssynchrony were included in the regression analysis ( Table 3 ). In the univariate model, the changes in LVESV, LV cavity width, LV +dP/dt, tenting area, Ts-Dif, and APM-PPM delay were correlated with the improvement in total MR. However, in multivariate analysis, only the reductions of LVESV (β = 0.837, P = .004) and tenting area (β = 0.880, P = .012) were confirmed to be independent determinants of improvement in total MR ( Table 3 ).
|Univariate model||Multivariate model|
|Parameter||β||95% CI||P||β||95% CI||P|
|Δ LVESV (%)||0.835||0.763–0.914||<.001||0.837||0.742–0.945||.004|
|Δ LV +dP/dt (mm Hg/sec)||1.004||1.001–1.007||.008||NS|
|Δ LV cavity width (%)||0.932||0.881–0.987||.016||NS|
|Δ Tenting area (%)||0.891||0.834–0.952||.001||0.880||0.796–0.972||.012|
|Δ Ts-Dif (ms)||0.968||0.950–0.986||.001||NS|
|Δ APM-PPM delay (ms)||0.971||0.955–0.988||.001||NS|