Doppler-derived preoperative mitral regurgitation volume predicts postoperative left ventricular dysfunction after mitral valve repair

Doppler-derived preoperative mitral regurgitation volume predicts postoperative left ventricular dysfunction after mitral valve repair Tetsuhiro Yamano, MD,a A. Marc Gillinov, MD,b Nozomi Wada, MD,a Yoshiki Matsumura, MD,a Manatomo Toyono, MD,a James D. Thomas, MD,a and Takahiro Shiota, MD c Cleveland, OH; and Los Angeles, CA

Background Unexpected postoperative left ventricular (LV) dysfunction after valve repair for mitral regurgitation (MR) occurs in some patients with normal preoperative LV function. Identification of factors that predispose to such LV dysfunction would enhance our understanding of the indications and outcomes of surgery. Methods

We retrospectively analyzed pre- and postoperative (median fourth day) echocardiograms of 174 patients undergoing valve repair for pure and isolated MR. Preoperative MR volume was quantified by the quantitative Doppler and/or proximal isovelocity surface area method.

Results

There was an incremental predictive value of MR quantification over the current recommendations (global χ2 from 48.14 to 81.57, P b .001; Hosmer-Lemeshow test, P = .98), for postoperative LV dysfunction, defined as ejection fraction b50%. The independent predictors were MR volume and LV end-systolic dimension (P b .001 and P = .01, respectively). Sixty-nine patients underwent surgery before development of the current surgical criteria, namely, symptoms, atrial fibrillation, preoperative LV dysfunction, or pulmonary hypertension. Of these, MR volume was the only independent significant predictor (P b .001) of unexpected postoperative LV dysfunction that developed in 14 patients (20%). Unexpected LV dysfunction could be predicted with sensitivity of 86% (95% CI 67%-100%) and specificity of 89% (95% CI 81%-97%), using the optimal cutoff of 80 mL for MR volume.

Conclusions

Doppler-derived preoperative MR volume is a powerful predictor of unexpected postoperative LV dysfunction. Prompt mitral valve repair may be beneficial for patients with high likelihood of successful repair and MR volume ≥80 mL. (Am Heart J 2009;157:875-82.)

The optimal timing of surgical indication for chronic mitral regurgitation (MR) remains a matter of controversy.1-3 Postoperative left ventricular (LV) dysfunction frequently occurs after surgical correction of MR4,5 and causes congestive heart failure, which results in a poor prognosis.4,6,7 Thus, a key concern is to avoid LV functional deterioration after mitral valve (MV) surgery to improve longevity.4,5,7,8 The current clinical guidelines have defined surgical indications, taking into account prevention of postoperative LV dysfunction.9 However, we occasionally encounter “unexpected” LV

From the aDepartment of Cardiovascular Medicine, Cleveland Clinic Foundation, Cleveland, OH, bDepartment of Thoracic and Cardiovascular Surgery, Cleveland Clinic Foundation, Cleveland, OH, and cDepartment of Cardiology and David Geffen School of Medicine, University of California, Los Angeles, CA. Submitted October 1, 2008; accepted March 4, 2009. Reprint requests: Takahiro Shiota, MD, Cedars-Sinai Medical Center, Department of Cardiology and David Geffen School of Medicine, University of California, 8700 Beverly Blvd, Rm 5618, Los Angeles, CA 90048. E-mail: [email protected] 0002-8703/$ - see front matter © 2009, Mosby, Inc. All rights reserved. doi:10.1016/j.ahj.2009.03.001

dysfunction after MV repair, even when the surgery is performed for patients before they develop class І or IIa (atrial fibrillation or pulmonary hypertension) indications for surgery. There has been a long-standing concept that regurgitant volume itself is an important contributor to LV response after MV surgery.10,11 However, there are no studies validating this concept in a large number of patients with a truly quantitative approach. We hypothesized that Doppler-derived preoperative MR volume could be an important determinant of postoperative LV function and sought to clarify the incremental value of MR quantification to the current recommendations, with regard to the prediction of LV dysfunction after MV repair.

Methods Study patients Our study subjects consisted of 249 consecutive patients who underwent MV repair for chronic MR from June through December 2005. Patients were excluded if they had mitral stenosis of any degree or aortic stenosis or regurgitation of more than mild degree. We excluded 2 patients because of unavailability of preoperative (within 180 days before surgery)

American Heart Journal May 2009

876 Yamano et al

Table I. Pre- and postoperative variables Postoperative LV dysfunction⁎ Variables

Total (N = 174)

Present (n = 52)

Absent (n = 122)

56 (46-63) 103 (59) 29 (17)

58 (47-66) 28 (54) 10 (19)

55 (46-61) 75 (61) 19 (16)

Age at surgery† (y) Preoperative NYHA class I, n (%) Atrial fibrillation, n (%) Pre LV end-diastolic dimension† (mm) LVDs† (mm) LV end-diastolic volume† (mL) LV end-systolic volume† (mL) LVEF† (%) Left atrial dimension† (mm) PASP† (mm Hg) MR volume† (mL) MR fraction†, % Mitral-valve flail, n (%) Anterior leaflet involvement, n (%) Postoperative residual MR, n (%)

57 35 166 52 67 46 36 69 57 81 43

(53-62) (31-39) (137-196) (41-70) (63-72) (42-51) (32-45) (54-92) (49-64) (47) (25) –

Post

Pre

51 (47-54)‡ 37 (32-41)‡ 119 (100-148)‡ 54 (41-74)§ 55 (47-59)‡ 42 (38-47)‡ 34 (31-37)‡ – – – – 24 (14)

63 (58-67) 39 (37-45) 199 (167-241) 72 (57-84) 64 (61-70) 51 (46-54) 40 (33-51) 97 (79-106) 64 (58-72) 31 (60) 14 (27) –

Post (50-58)‡ (39-46)‡ (115-167)‡ (65-97)‡ (38-46)‡ (40-49)‡ (31-41)‡ – – – – 9 (17)

54 43 137 81 43 45 35

Pre

Post

56 (51-60)O 33 (30-37)O 157 (126-178)O 47 (37-59)O 68 (65-73)O 45 (40-49)O 35 (31-42)# 61 (51-71)O 52 (46-60)O 50 (41)# 29 (24) –

(46-54)‡,O (31-38)§,O (93-135)‡,O (37-58)O (54-63)‡,O (37-45)‡,O (31-37)§ – – – – 15 (12)

49 34 111 48 57 41 34

Pre-and postoperative echocardiographic variables were obtained a median of 4 days preoperatively and 4 days postoperatively, respectively. NYHA, New York Heart Association. ⁎ Defined as postoperative LVEF b50%. † Values are median (interquartile range). ‡ P b .005 for difference between pre- and postoperative variable. § P b .05 for difference between pre- and postoperative variable. O P b .005 for difference between postoperative LV dysfunction present and absent. # P b .05 for difference between postoperative LV dysfunction present and absent.

Table II. Perioperative variables and complications Postoperative LV dysfunction⁎

Variables

Total (N = 174)

Present (n = 52)

Absent (n = 122)

Combined surgical 24 (14) 11 (21) 13 (11) procedure, n (%) Operative times 63 (49-73) 65 (52-78) 60 (49-71)‡ Ischemic† (min) Cardiopulmonary 77 (64-93) 80 (69-99) 75 (61-89)‡ bypass† (min) Perioperative complication Myocardial infarction, n (%) 1 (1) 0 (0) 1 (1) Heart failure, n (%) 2 (1) 1 (2) 1 (1) Reoperation for bleeding, n (%) 3 (2) 0 (0) 3 (2) Significant pericardial 5 (3) 1 (2) 4 (3) effusion, n (%) ⁎ Defined as postoperative LVEF b50%. † Values are median (interquartile range). ‡ P b .05 for difference between postoperative LV dysfunction present and absent.

or postoperative (within 30 days after surgery) echocardiograms; 48, with technical difficulties in exactly delineating LV endocardial border; and 25, with significant angiographic coronary artery stenosis N50%. Finally, 174 patients (54 women and 120 men) were included in the analyses. By surgical inspection, all subjects had primary organic valve diseases

Table III. Accuracy of preoperative echocardiographic variables for prediction of postoperative LV dysfunction⁎ Variables

AUC

95% CI

P

LV end-diastolic dimension LVDs LV end-diastolic volume LV end-systolic volume LVEF Left atrial dimension PASP MR volume MR fraction

0.79 0.82 0.79 0.82 0.68 0.72 0.63 0.88 0.80

0.71-0.86 0.75-0.89 0.72-0.87 0.75-0.89 0.59-0.77 0.64-0.81 0.52-0.73 0.82-0.94 0.73-0.87

b.001 b.001 b.001 b.001 b.001 b.001 .02 b.001 b.001

P value is the significance of the test comparing the AUC vs 0.5 (H0: AUC = 0.5, H1: AUC N0.5). ⁎ Defined as postoperative LVEF b50%.

causing MR. Other than 2 patients with healed infective endocarditis on a previously normal valve and rheumatic valve disease, the etiology of regurgitation was degenerative disease with leaflet prolapse and/or flail in 172 patients (99%).12 The study was approved by the ethics committee at the Cleveland Clinic.

Echocardiographic examinations Left ventricular volume and ejection fraction (EF) were measured by the biplane Simpson disk method13; and the postoperative LV dysfunction was defined as postoperative EF b50%.4,5,7 We diagnosed MV flail based on 2-dimensional (2D)

American Heart Journal Volume 157, Number 5

Figure 1

Yamano et al 877

Table IV. Logistic regression analysis of the current surgical criteria and MR quantification for prediction of postoperative LV dysfunction⁎: all patients Variables Preoperative NYHA non–class I Atrial fibrillation LVDs (per 1 mm) LVEF (per 1%) PASP (per 1 mm Hg) MR volume (per 1 mL)

Odds ratio

95% CI

P

1.14 0.82 1.17 0.91 1.02 1.08

0.35-3.74 0.20-3.41 1.04-1.32 0.82-1.01 0.97-1.07 1.05-1.12

.83 .79 .01 .08 .41 b.001

⁎ Defined as postoperative LVEF b50%.

to be 10 mm Hg.16 In patients with atrial fibrillation, 5 cardiac cycles were averaged for all measurements.17

Surgical procedures Receiver operating characteristic curves of MR volume, LVDs, and EF for postoperative LV dysfunction in all patients. The arrow indicates the optimal cutoff point of MR volume.

Figure 2

Valvular prolapse and/or flail involved the posterior leaflet in 129 patients (74%), anterior leaflet in 12 (7%), and both leaflets in 31 (18%). The most common repair technique for posterior leaflet prolapse was quadrangular resection, and anterior leaflet prolapse was usually treated by transfer of chords from the posterior leaflet, as described previously.18 In all cases, we also used a Cosgrove-Edwards annuloplasty ring which is a flexible posterior annuloplasty band (Edwards Lifesciences, Irvine, CA).19 All patients received both antegrade and retrograde blood cardioplegia during the surgical procedure.

Statistical analyses

Multivariate logistic regression models for the prediction of postoperative LV dysfunction.

echocardiograms.14 Quantification of MR volume was performed by the quantitative Doppler and/or proximal isovelocity surface area (PISA) method.15 The proximal isovelocity radius was measured at mid-systole with optimal flow convergence by which satisfactory hemispherical PISA was obtained. We used the mean value for analyses when we could use both methods. In patients for whom we could apply the quantitative Doppler method, we also calculated MR fraction as MR stroke volume/ mitral filling stroke volume. In patients with at least mild tricuspid regurgitation, the maximum regurgitation velocity was measured using continuous-wave Doppler, which was used to determine the pulmonary artery systolic pressure (PASP) by the simplified Bernoulli equation, with right atrial pressure assumed

All continuous variables are expressed as the medians (interquartile range) and compared with Mann-Whitney U or Wilcoxon signed rank test, as appropriate. The Pearson χ2 or Fisher exact test was used for categorical data, as appropriate. Receiver operating characteristic analyses were performed, and the area under the curve (AUC) was compared with the methods described by Hanley and McNeil.20 The optimal cutoff was determined as the point with the most favorable sensitivity and specificity. In the American College of Cardiology/American Heart Association guidelines,9 MV surgery is recommended for patients who develop either symptoms, new onset of atrial fibrillation, early signs of LV dysfunction (EF ≤60% and/or LV end-systolic dimension [LVDs] ≥40 mm), or pulmonary hypertension (PASP at rest ≥50 mm Hg). We classified these criteria into a 3-step model: clinical (symptom and atrial fibrillation), LV parameters (EF and LVDs), and PASP; then, we assessed the incremental predictive value of MR quantification to the current recommendations, using likelihood ratio tests. The HosmerLemeshow goodness-of-fit statistic was obtained to assess the adequacy of the final model. The serial echocardiographic changes were tested using 1-way analyses of variance for repeated measures, followed by Sheffe multiple comparison tests. Mitral regurgitation volumes obtained by the 2 methods were compared in 65 subjects in whom both methods could be applied, and interobserver variability was assessed in 20 randomly selected subjects. All statistical analyses were performed with the statistical package SPSS for Windows (version 11.0J, SPSS, Chicago, IL). A P value b.05 was considered significant.

878 Yamano et al

American Heart Journal May 2009

Figure 3

Representative case showing depressed EF postoperatively despite mitral valve repair being indicated earlier than development of any of surgical criteria. Upper panels reveal preoperative (Pre) and lower, postoperative (Post) echocardiograms. Left panels reveal color Doppler; middle, 2D images at end-diastole (ED); and right, at end-systole (ES). In pre–color Doppler image, upper right panel shows 2D image clearly delineating posterior mitral valve flail. In this case, regurgitant volume was calculated at 136 mL. The EF by the Simpson method was decreased from 71% to 42% (Pre: LV end-diastolic dimension 67 mm and end-systolic dimension 35 mm, end-diastolic volume 261 mL and end-systolic volume 75 mL; Post: 52 and 42 mm, and 165 and 95 mL, respectively).

No extramural funding was used to support this work. The authors are solely responsible for the design and conduct of this study, all study analyses, and the drafting and editing of the paper.

Results Echocardiographic variables Postoperative LV dysfunction, which was measured 4 days (3-5 days) after the surgery, was noted in 52 patients (30%). Pre- and postoperative variables were compared between patients with postoperative LV dysfunction (EF 43% [38%-46%]) and those without (EF 57% [54%-63%]) (Table I). In patients with postoperative LV dysfunction, preoperative LV end-diastolic and end-systolic volumes were significantly larger (both P b .001); and preoperative EF was less than in those without (64% [61%-70%] vs 68% [65%-73%], P b .001). Pulmonary artery systolic pressure could be determined in 148 patients (85%) preoperatively and in 116 patients (67%) postoperatively. Preoperative PASP was significantly higher in patients with postoperative LV dysfunction (P = .02). Mitral

regurgitation volume could be measured in 65 subjects (37%) by both methods, in 94 (54%) by the Doppler approach alone, and 15 (9%) by just the PISA method. Patients with postoperative LV dysfunction had significantly greater MR volume and more frequently had MV flail than those without (P b .001 and P = .02, respectively). Postoperatively, MR disappeared in all patients except for 24 (14%) in whom just trace regurgitation was found. In 10 patients (6%), pre- and postoperative echocardiograms were recorded under different cardiac rhythm, that is, sinus rhythm or atrial fibrillation, 5 patients each moving from one rhythm to the other.

Perioperative variables and complications Concomitant surgical procedures were performed in 24 patients (14%), including 5 tricuspid valve repairs and 23 maze or pulmonary vein isolation procedures. The number of patients for whom combined procedures were performed was greater with borderline significance (P = .07); consequently, both ischemic and cardiopulmonary

American Heart Journal Volume 157, Number 5

Figure 4

Yamano et al 879

Table V. Logistic regression analysis of the current surgical criteria and MR quantification for prediction of postoperative LV dysfunction⁎: patients without criteria for surgery Variables LVDs (per 1 mm) LVEF (per 1%) PASP (per 1 mm Hg) MR volume (per 1 mL)

Odds ratio

95% CI

P

1.17 0.91 1.07 1.14

0.90-1.53 0.74-1.12 0.90-1.26 1.06-1.23

.24 .38 .46 b.001

⁎ Defined as postoperative LVEF b50%.

bypass times were longer in patients with postoperative LV dysfunction than in those without (P = .04 and .03, respectively) (Table II). There was no operative death within 30 days after the surgery. Perioperatively, 1 patient had a myocardial infarction diagnosed by electrocardiogram criteria and elevated cardiac enzymes, and 2 patients had heart failure requiring diuretic therapy or reintubation for pulmonary congestion. Prevalence of perioperative complications was not statistically different between patients with postoperative LV dysfunction and those without.

(P b .001). Substitution of MV flail for MR volume did not contribute to a significant improvement of this model (global χ2 from 48.14 to 49.38, P = .27). Furthermore, addition of operative times could not make better this model (global χ2 from 81.57 to 82.65, P = .58). In the final model, the independent predictors were MR volume and LVDs (Table IV); and the Hosmer-Lemeshow goodness-of-fit statistic was 0.98 for 8 degrees of freedom. Among patients in whom preoperative PASPs were determined, 69 patients (47%) were asymptomatic and did not have any definite criteria for surgery. Of these, 14 patients (20%) developed unexpected LV dysfunction after the surgery (Figure 3). In this patient group, receiver operating characteristic analysis for MR volume (AUC 0.92, 95% CI 0.85-0.99) indicated the same optimal cutoff as in the entire cohort (Figure 4), and MR volume was the only independent significant predictor of LV dysfunction (Hosmer-Lemeshow test, P = .91) (Table V). Unexpected LV dysfunction could be predicted with sensitivity of 86% (95% CI 67%-100%) and specificity of 89% (95% CI 81%-97%), using the optimal cutoff value of 80 mL for MR volume.

Predictive accuracy for postoperative LV dysfunction

Serial echocardiographic changes

Receiver operating characteristic curve of MR volume for postoperative LV dysfunction in patients without criteria for surgery. The arrow indicates the optimal cutoff point.

The AUC of MR volume tended to be greater than those of LVDs and end-systolic volume (both P = .08) and was significantly larger than those of LV end-diastolic dimension and volume (both P = .02), as well as EF, left atrial dimension, and PASP (all P b .001) (Table III). Also, AUC of MR volume was significantly greater than that of MR fraction (P = .02). The optimal cutoff value of MR volume was defined as 80 mL, with sensitivity of 75% (95% CI 63%-87%) and specificity of 87% (95% CI 81%-93%) for the prediction of postoperative LV dysfunction (Figure 1).

Incremental predictive value of MR volume Step-by-step addition of LV parameters and PASP resulted in a significant increase in the global χ2 in the prediction of postoperative LV dysfunction (P b .001 and P = .002, respectively) (Figure 2). However, subsequent addition of MR quantification to the combined model of clinical, LV parameters, and PASP yielded a further significant increase in the global χ2 from 48.14 to 81.57

Midterm follow-up (N30 days) echocardiograms were available in 33 patients (Table VI). Their preoperative echocardiographic variables were not statistically different from those of patients without follow-up echocardiograms (EF 68% [62%-73%] vs 67% [63%-72%], P = .70; MR volume 69 mL [54-88 mL] vs 69 mL [54-93 mL], P = .99). Follow-up echocardiography was performed at a median of 365 days (121-519 days, range 32-1226 days) postoperatively. At follow-up echocardiography, cardiac rhythm was different from pre- or postoperative echocardiography in 4 patients (12%). Ejection fraction showed a significant recovery from 55% (46%-58%) at the early postoperative period to 59% (52%-65%) at the follow-up (P = .001). However, among 10 patients who developed postoperative LV dysfunction early after the surgery, 6 cases still had LV dysfunction at the follow-up echocardiography (368 days [106-576 days]). Conversely, although the follow-up period was not different (365 days [115-515 days], P = .80), there was no one

American Heart Journal May 2009

880 Yamano et al

Table VI. Serial echocardiographic changes Postoperative LV dysfunction⁎ Total (N = 33) Variables LV end-diastolic volume (mL) LV end-systolic volume (mL) LVEF (%)

Present (n = 10)

Absent (n = 23)

Pre

Post

F/U

Pre

Post

F/U

Pre

Post

F/U

158 (139-175) 49 (41-57) 68 (62-73)

118 (102-130)† 53 (42-65)‡ 55 (46-58)†

110 (92-133)† 42 (34-62)§ 59 (52-65)†,§

175 (152-237) 70 (45-92) 62 (57-69)

130 (103-190)† 84 (56-113) 41 (30-46)†

133 (112-177)† 72 (54-98) 46 (42-55)†,O

154 (138-170) 44 (38-51)⁎⁎ 70 (67-74)⁎⁎

117 (102-127)† 50 (36-58)# 55 (53-60)†,#

106 (89-114)†,# 39 (30-45)§,# 64 (57-66)†,O,#

Values are median (interquartile range). Pre- and postoperative, and follow-up (F/U) variables were obtained a median of 24 days preoperatively, 4 days postoperatively, and 365 days postoperatively, respectively. ⁎ Defined as postoperative LVEF b50%. † P b .005 for difference from preoperative variable. ‡ P b .05 for difference from preoperative variable. § P b .005 for difference from postoperative variable. OP b .05 for difference from postoperative variable. # P b .005 for difference between postoperative LV dysfunction present and absent. ⁎⁎ P b .05 for difference between postoperative LV dysfunction present and absent.

who had LV dysfunction at the follow-up echocardiography among patients without LV dysfunction early after the surgery.

Comparison and reproducibility of MR quantification There was a significant correlation between MR volume obtained by the quantitative Doppler and PISA methods (r = 0.77, P b .001, SEE = 15 mL). Bland-Altman analysis21 showed there was a trend toward overestimation of MR volume by the quantitative Doppler method (absolute difference, 6 ± 17 mL [mean ± SD]). In regard to the interobserver variability, linear regression showed acceptable agreement, as follows: r = 0.90, P b .001, SEE = 9 mL for the quantitative Doppler; and r = 0.90, P b .001, SEE = 10 mL for the PISA method.

Discussion Mitral regurgitation volume and postoperative LV dysfunction Surgical correction of MR now can be reasonably (class IIa) indicated even for asymptomatic patients without any of surgical criteria, namely, atrial fibrillation, LV dysfunction by either EF ≤60% or LVDs ≥40 mm, or pulmonary hypertension with PASP ≥50 mm Hg, provided that there is at least 90% likelihood of successful valve repair and that the repair is performed in a center experienced in this procedure.9 The rationale of “early surgery” is to avoid the occult onset of LV dysfunction from long-standing volume overload2 because the previous studies4,7 have demonstrated that postoperative LV dysfunction of EF b50% has a prognostic significance, emphasizing the importance of prevention of postoperative LV dysfunction. However, 20% of them still developed unexpected postoperative

LV dysfunction in the present study, suggesting that the surgical criteria in the current guidelines9 still might be insufficient for the prevention of postoperative LV dysfunction in some patients. In the present study, we found the relation between large preoperative MR volume and postoperative LV dysfunction and demonstrated that unexpected LV dysfunction could be accurately predicted using a cutoff value of 80 mL in MR volume. We also showed that MV flail was more frequently involved in patients with postoperative LV dysfunction because the lesion results in high degree of regurgitation.22 Our study also showed that LVDs remained as a predictor of postoperative LV dysfunction, as well as MR volume. Although further study is needed, these data suggest that we consider MV repair as soon as possible for patients with MR volume ≥80 mL for the prevention of postoperative LV dysfunction, if the MV anatomy is suitable for repair.

Value of MR quantification Enriquez-Sarano et al1 have reported that asymptomatic patients with severe MR based on an effective regurgitant orifice (ERO) area ≥40 mm2 have a much worse clinical outcome than previously assumed. Thus, although there has been conflicting data verifying adherence to the current surgical criteria,3 they have proposed that prompt surgery should be indicated for asymptomatic MR patients with ERO of at least 40 mm2 to improve outcome.1 Their study, however, did not investigate LV function after MV surgery. In the present study, we quantified not only severity of MR but also LV function before and after MV repair and demonstrated that quantitative severity of MR also could be important predictor of postoperative LV dysfunction. In their study,1 ERO of 40 mm2 corresponds to about 60 mL for MR

American Heart Journal Volume 157, Number 5

volume, although MR volume is not linearly associated with ERO.23 Therefore, the present study could support their concept of prompt surgery especially for patients with very severe MR, namely, MR volume ≥80 mL because severe MR is defined as MV volume ≥60 mL by the current American Society of Echocardiography guidelines.15 Our study provides additional confirmation that MR quantification is highly recommended for the clinical management of MR patients, especially to determine the optimal timing of surgical indication for asymptomatic patients.

Limitations The correlation of MR volume quantified by the quantitative Doppler and PISA methods was clinically acceptable, and good reproducibility was observed. However, as MR volume in this study might be somewhat overestimated considering the total stroke volume by 2D echocardiography, certain measurement errors of these methods are inevitable.15,24 We did not use ERO as a measure of MR severity because technical difficulties in obtaining the full jet envelope of eccentric MR15,23 and the retrospective nature of this study prevented us from obtaining ERO in more than half of the subjects (54%). We defined postoperative LV dysfunction as EF b50% according to the past studies.4,5,7 However, EF is only one aspect of true LV systolic function in which many other components are involved.8 Although operative times were not independent significant predictors, the impact of surgical complexity as well as combined procedures could not be negated. We also did not assess effects of atrial fibrillation on LV parameters. The main limitation of our study was that we did not have long-time follow-up in most of our patients, and the postoperative LV function was assessed only a median of 4 days after the surgery. However, as shown in Table VI, EF obtained at the midterm follow-up period was still significantly lower in patients with postoperative LV dysfunction early after the surgery than in those without. In addition, although the number was small, more than half of the patients who showed LV dysfunction early after the surgery still had LV dysfunction at the follow-up echocardiography. Therefore, the clinical significance of early postoperative LV dysfunction holds true despite functional recovery during follow-up.

Conclusions In the present study, there was a significant incremental predictive value of MR quantification with regard to LV dysfunction after MV repair, and the Doppler-derived preoperative MR volume ≥80 mL accurately predicted unexpected postoperative LV dysfunction in asymptomatic patients undergoing MV repair before development of the current surgical

Yamano et al 881

criteria. For patients with high likelihood of successful repair and preoperative MR volume of at least 80 mL, prompt MV repair may be beneficial for the prevention of postoperative LV dysfunction.

Acknowledgements A. Marc Gillinov has served as a consultant and speaker for Edwards Lifesciences, Irvine, CA; Medtronic, Inc, Minneapolis, MN; and St Jude Medical, St Paul, MN. He has an equity interest in Viacor, Inc, Wilmington, MA.

References 1. Enriquez-Sarano M, Avierinos JF, Messika-Zeitoun D, et al. Quantitative determinants of the outcome of asymptomatic mitral regurgitation. N Engl J Med 2005;352:875-83. 2. Ling LH, Enriquez-Sarano M, Seward JB, et al. Early surgery in patients with mitral regurgitation due to flail leaflets: a long-term outcome study. Circulation 1997;96:1819-25. 3. Rosenhek R, Rader F, Klaar U, et al. Outcome of watchful waiting in asymptomatic severe mitral regurgitation. Circulation 2006;113: 2238-44. 4. Enriquez-Sarano M, Tajik AJ, Schaff HV, et al. Echocardiographic prediction of left ventricular function after correction of mitral regurgitation: results and clinical implications. J Am Coll Cardiol 1994;24:1536-43. 5. Matsumura T, Ohtaki E, Tanaka K, et al. Echocardiographic prediction of left ventricular dysfunction after mitral valve repair for mitral regurgitation as an indicator to decide the optimal timing of repair. J Am Coll Cardiol 2003;42:458-63. 6. Enriquez-Sarano M, Schaff HV, Orszulak TA, et al. Congestive heart failure after surgical correction of mitral regurgitation. A long-term study. Circulation 1995;92:2496-503. 7. Crawford MH, Souchek J, Oprian CA, et al. Determinants of survival and left ventricular performance after mitral valve replacement. Department of Veterans Affairs Cooperative Study on Valvular Heart Disease. Circulation 1990;81:1173-81. 8. Starling MR, Kirsh MM, Montgomery DG, et al. Impaired left ventricular contractile function in patients with long-term mitral regurgitation and normal ejection fraction. J Am Coll Cardiol 1993; 22:239-50. 9. Bonow RO, Carabello BA, Chatterjee K, et al. ACC/AHA guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol 2006;48: e1-e148. 10. Boucher CA, Bingham JB, Osbakken MD, et al. Early changes in left ventricular size and function after correction of left ventricular volume overload. Am J Cardiol 1981;47:991-1004. 11. Clancy KF, Hakki AH, Iskandrian AS, et al. Forward ejection fraction: a new index of left ventricular function in mitral regurgitation. Am Heart J 1985;110:658-64. 12. Waller BF, Morrow AG, Maron BJ, et al. Etiology of clinically isolated, severe, chronic, pure mitral regurgitation: analysis of 97 patients over 30 years of age having mitral valve replacement. Am Heart J 1982; 104:276-88. 13. Lang RM, Bierig M, Devereux RB, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography's Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction

American Heart Journal May 2009

882 Yamano et al

14.

15.

16.

17.

18.

with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiogr 2005;18: 1440-63. Mintz GS, Kotler MN, Segal BL, et al. Two-dimensional echocardiographic recognition of ruptured chordae tendineae. Circulation 1978;57:244-50. Zoghbi WA, Enriquez-Sarano M, Foster E, et al. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003;16:777-802. McQuillan BM, Picard MH, Leavitt M, et al. Clinical correlates and reference intervals for pulmonary artery systolic pressure among echocardiographically normal subjects. Circulation 2001;104: 2797-802. Gottdiener JS, Bednarz J, Devereux R, et al. American Society of Echocardiography recommendations for use of echocardiography in clinical trials. J Am Soc Echocardiogr 2004;17:1086-119. Gillinov AM, Cosgrove DM. Mitral valve repair for degenerative disease. J Heart Valve Dis 2002;11(Suppl 1):S15-20.

19. Gillinov AM, Cosgrove III DM, Shiota T, et al. Cosgrove-Edwards Annuloplasty System: midterm results. Ann Thorac Surg 2000;69: 717-21. 20. Hanley JA, McNeil BJ. A method of comparing the areas under receiver operating characteristic curves derived from the same cases. Radiology 1983;148:839-43. 21. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1: 307-10. 22. Pearson AC, St Vrain J, Mrosek D, et al. Color Doppler echocardiographic evaluation of patients with a flail mitral leaflet. J Am Coll Cardiol 1990;16:232-9. 23. Enriquez-Sarano M, Seward JB, Bailey KR, et al. Effective regurgitant orifice area: a noninvasive Doppler development of an old hemodynamic concept. J Am Coll Cardiol 1994;23: 443-51. 24. Enriquez-Sarano M, Bailey KR, Seward JB, et al. Quantitative Doppler assessment of valvular regurgitation. Circulation 1993;87: 841-8.

The following article is an AHJ Online Exclusive. Full text of this article is available at no charge at our website: www.ahjonline.com

Imaging and Diagnostic Testing Real-time 3-dimensional echocardiography early after acute myocardial infarction: Incremental value of echo-contrast for assessment of left ventricular function Gaetano Nucifora, MD,a,b Nina Ajmone Marsan, MD,a Eduard R. Holman, MD, PhD,a Hans-Marc J. Siebelink, MD, PhD,a Jacob M. van Werkhoven, MSc,a Arthur J. Scholte, MD,a Ernst E. van der Wall, MD, PhD,a,c Martin J. Schalij, MD, PhD,a and Jeroen J. Bax, MD, PhD,a Leiden and Utrecht, The Netherlands; Udine, Italy

Background

Accurate and reproducible assessment of left ventri-

graded for each visible segment as follows: 1 = normal, 2 = hypokinetic, 3 =

cular (LV) systolic function is important in patients with acute myocardial

akinetic, and 4 = dyskinetic.

infarction (AMI). Real-time 3-dimensional echocardiography (RT3DE) is an

Results

accurate technique, but it relies heavily on good image quality. The aim of the present study was to evaluate the incremental value of contrastenhanced RT3DE.

Methods

A total of 140 consecutive patients (58 ± 11 years, 78%

men) with ST-elevation AMI clinically underwent nonenhanced and contrastenhanced RT3DE within 24 hours from AMI to evaluate global and regional

During contrast-enhanced RT3DE, as compared with nonenhanced RT3DE, the number of segments with complete visualization of the endocardial border increased from 66% to 84% (P b .001); and the number of patients with a good-quality echocardiogram increased from 59% to 94% (P b .001). Intra- and interobserver agreement for assessment of global and regional LV systolic function improved during contrast-enhanced RT3DE, as compared with nonenhanced RT3DE.

LV systolic function. Endocardial border definition was graded for each of the

Conclusions

16 LV segments as follows: 0 = border invisible, 1 = border visualized only

RT3DE is frequently hampered by suboptimal echocardiographic quality.

partially, and 2 = complete visualization of the border. Three image-quality

Contrast-enhanced RT3DE is of incremental value, improving the endocardial

groups (good, fair, and uninterpretable) were identified. Left ventricular

border visualization and the reproducibility of LV function assessment.

volumes and ejection fraction were measured off-line. Wall motion was

(Am Heart J 2009;157:882.e1-882.e8.)

Assessment of LV systolic function in AMI patients with