Quantitative Three-Dimensional Echocardiographic Correlates of Optimal Mitral Regurgitation Reduction during Transcatheter Mitral Valve Repair

Quantitative Three-Dimensional Echocardiographic Correlates of Optimal Mitral Regurgitation Reduction during Transcatheter Mitral Valve Repair Didem Oguz, MD, Mackram F. Eleid, MD, Sumandeep Dhesi, MD, Sorin V. Pislaru, MD, PhD, Sunil V. Mankad, MD, Joseph F. Malouf, MD, Vuyisile T. Nkomo, MD, MPH, Jae K. Oh, MD, David R. Holmes, MD, Guy S. Reeder, MD, Charanjit S. Rihal, MD, and Jeremy J. Thaden, MD, Rochester, Minnesota

Background: Patient selection for transcatheter edge-to-edge mitral valve repair (TMVR) remains challenging because of heterogenous mitral valve pathology and highly variable anatomy. The aim of this study was to investigate whether quantitative three-dimensional (3D) transesophageal echocardiographic modeling parameters are associated with optimal mitral regurgitation (MR) reduction in patients undergoing TMVR. Methods: Fifty-nine patients underwent 3D transesophageal echocardiography during TMVR. Volumetric data sets were retrospectively analyzed using mitral valve quantitative 3D modeling software (Mitral Valve Navigator). Optimal MR reduction was defined as less than moderate residual MR. Logistic regression was used to correlate 3D transesophageal echocardiographic quantitative data to procedural success. Results: Thirty-five patients had primary MR, 24 had mixed or secondary MR, and all patients had grade $ 3/ 4 MR before the procedure. Optimal MR reduction was achieved in 40 of 59 patients (68%). Univariate correlates of optimal MR reduction in patients with primary MR were lower mitral leaflet tenting volume (P = .049) and lower tenting height (P = .025); tenting height < 3 mm and tenting volume < 0.7 mL were associated with increased likelihood of optimal MR reduction (92% vs 48% [P = .01] and 81% vs 47% [P = .03], respectively). In mixed or secondary MR, annular height $ 5.5 mm was associated with increased likelihood of optimal MR reduction (94% vs 38%; P = .03). During follow-up, redo TMVR or surgical mitral valve replacement occurred exclusively in patients with suboptimal anatomy defined by 3D transesophageal echocardiography (10% vs 0%, P = .045). Conclusions: Quantitative 3D echocardiographic data are associated with favorable response to TMVR and could help optimize patient selection. (J Am Soc Echocardiogr 2019;-:---.) Keywords: Mitral regurgitation, MitraClip, 3D transesophageal echocardiography

Mitral regurgitation (MR) is one of the most common valve lesions encountered in clinical practice. It is increasingly prevalent with age, with moderate or greater MR estimated to be present in 9.3% of individuals >75 years of age,1 and it constitutes approximately 24.8% of all valve disease encountered in the health care setting.2

From the Department of Cardiovascular Medicine, Mayo Clinic, Rochester, Minnesota. Dr. Oguz received a research grant supporting this work from the Turkish Society of Cardiology. Conflicts of Interest: None. Reprint requests: Jeremy J. Thaden, MD, Department of Cardiovascular Medicine, Mayo Clinic, 200 First Street SW, Rochester, MN 55905 (E-mail: thaden.jeremy@ mayo.edu). 0894-7317/$36.00 Copyright 2019 by the American Society of Echocardiography. https://doi.org/10.1016/j.echo.2019.06.014

Available data indicate that surgery is underused for patients with MR, in part because of advanced age and increased comorbidities, which impart increased surgical risk.3,4 Transcatheter edge-to-edge mitral valve repair (TMVR) offers an alternative to surgery in patients at high or prohibitive surgical risk. TMVR is currently clinically approved for patients with primary MR,5 and recent data are conflicting on its efficacy in secondary MR.6,7 Patient selection for TMVR remains challenging because of diverse mitral valve pathology and highly variable anatomy.8,9 Available studies are discordant with respect to anatomic predictors of procedural safety and efficacy.10-13 Recent data from our experience indicate that patients with flail leaflets and a single jet of regurgitation are more likely to achieve optimal MR reduction,11 but it is difficult to quantify complex mitral valve anatomy using two-dimensional (2D) transesophageal echocardiographic (TEE) imaging alone. Three-dimensional (3D) TEE imaging provides a panoramic view of the mitral valve in a single image, allowing a clearer understanding 1

2 Oguz et al

of the underlying mitral valve pathology and location of the MR 2D = Two-dimensional jet. Moreover, 3D TEE mitral valve data sets can be used to 3D = Three-dimensional create 3D models of the mitral LVEDD = Left ventricular valve apparatus, which provide end-diastolic diameter quantitative measurements of the mitral valve annulus, mitral MR = Mitral regurgitation valve leaflets, and their spatial TEE = Transesophageal relationships. Quantitative 3D echocardiographic models of the mitral valve afford TMVR = Transcatheter edgea better understanding of the unto-edge mitral valve repair derlying anatomy and pathology and have the potential to better predict which patients will have optimal MR reduction with TMVR. We aimed to assess the whether parameters of quantitative 3D TEE mitral valve modeling correlate with optimal MR reduction following TMVR. Abbreviations

METHODS Patients This study was approved by the Mayo Clinic institutional review board. Consecutive patients with primary or mixed or secondary mechanism MR who underwent TMVR (MitraClip; Abbott Vascular, Santa Clara, CA) between May 1, 2015, and June 30, 2017, were included in the analysis. All patients were evaluated by a heart team including an interventional cardiologist and a cardiac surgeon with expertise in mitral valve surgery and determined to be at high risk for surgery. Four patients with suboptimal 3D TEE image quality and one patient who had prior mitral valve repair were excluded from the analysis. All patients received routine general anesthesia during the procedure. Pre- and postprocedural imaging assessment was performed while under general anesthesia. Care was taken to match pre- and postprocedural hemodynamics to facilitate comparison of MR reduction before and after the procedure. Echocardiography All patients underwent preprocedural transthoracic echocardiography. Left ventricular diameter, volume, and ejection fraction were measured in accordance with current guidelines.14 MR severity was assessed using color and continuous-wave Doppler echocardiography in conjunction with pulsed-wave Doppler assessment of the pulmonary veins and proximal isovelocity surface area quantitation when feasible. The severity of MR was graded using a multiparametric approach according to current American Society of Echocardiography guidelines.15 Primary MR was defined as regurgitation due to degenerative mitral valve leaflets, including prolapse or flail and/or abnormalities of the supporting apparatus. Secondary MR was defined as regurgitation due to tethering of structurally normal mitral leaflets (e.g., ventricular dilation or focal regional wall motion abnormalities) or mitral annular dilatation. Mixedmechanism MR was graded when there was visible leaflet tethering on the 2D images with coexistent abnormalities of the mitral leaflets, such that both were felt to contribute to the MR. Optimal MR reduction was defined as less than moderate residual MR (none,

Journal of the American Society of Echocardiography - 2019

trace, or mild) on intraoperative echocardiography. The pre- and postprocedural assessments of MR severity were performed on intraprocedural TEE imaging with the patient under general anesthesia. The preprocedural MR assessment typically occurred before femoral vein access, and the postprocedural assessment was performed immediately after clip deployment as part of intraprocedural TEE imaging. Three-dimensional TEE data sets were acquired before implantation of the MitraClip device as part of routine perioperative TEE imaging using a Philips EPIQ system (Philips Medical Systems, Andover, MA). Full-volume (31 of 59 [53%]) or 3D zoom (28 of 59 [47%]) data sets with single-beat (45 of 59 [76%]) or multibeat (14 of 59 [24%]) acquisitions were obtained at the discretion of the performing echocardiographer. The median frame rate was 11 Hz (interquartile range, 10–18 Hz). Three-Dimensional Quantitative Measurements All images were digitally stored for offline analysis using mitral valve quantitative 3D modeling software (Mitral Valve Navigator; Philips Healthcare, Best, the Netherlands) which provides semiautomated 3D modeling and quantification of the mitral annulus and mitral valve apparatus.16 Measurements were performed by two echocardiographers (D.O. and J.J.T.), and all measurements were performed blinded to the result of the TMVR procedure (Supplemental Videos 1 and 2, available at www.onlinejase.com). The end-systolic frame was identified as the last systolic frame before mitral valve opening. Three orthogonal planes of the mitral valve annulus were initially displayed at end-systole. The planes were then manually aligned to obtain bicommissural, left ventricular outflow tract (long-axis) and short-axis views of the mitral valve, respectively. The bicommissural view was used to identify the anterolateral and posteromedial points of the mitral annulus. The long-axis view was used to define the anterior and posterior points of the mitral annulus, the aortic annulus, and the mitral leaflet coaptation point. The semiautomated software then created a preliminary mitral annular model on the basis of these initial data points. Manual correction of the mitral annular points was performed using axial rotation around the mitral annulus and confirming the placement of each annular point (16 total) in an orthogonal view. After confirming the accuracy of the mitral annular points, the mitral leaflet commissures were identified in a mitral valve short-axis plane. The 3D-rendered surgical en face view was also used to confirm the location of medial and lateral mitral valve commissures. Following confirmation of the commissural points, the software then created a preliminary model of the mitral valve leaflets and zone of leaflet coaptation. Manual confirmation and adjustment of the mitral leaflets was then performed, if necessary, using multiple long-axis views of the mitral valve in a plane that lies orthogonal to the bicommissural view. Figure 1 shows relevant 3D TEE mitral valve and mitral annular measurements. Statistical Analysis Continuous variables are presented as mean 6 SD. Categorical variables are expressed as frequencies and percentages. The association between optimal MR reduction and 2D and 3D TEE parameters was examined using c2 analysis for categorical variables or logistic regression for continuous variables. Repeat 3D modeling was performed in a random subset of 20 patients by two echocardiographers (D.O. and J.J.T.) for interobserver and intraobserver variability analysis; the echocardiographers were blinded to the results of the initial analysis. The

Journal of the American Society of Echocardiography Volume - Number -

HIGHLIGHTS  Mitral valve 3D parameters are associated with optimal MR reduction during TMVR.  In primary MR, reduced leaflet tenting was associated with optimal MR reduction.  In functional MR, reduced annular height was associated with optimal MR reduction.  Patients with suboptimal anatomy more frequently had recurrent MR or reintervention.

analysis was timed in the same subgroup of patients, and the time needed to perform the analysis was recorded. All statistical tests were two-tailed, and P values < .05 were considered to indicate statistical significance. Statistical analysis was performed using JMP version 12.0 (SAS Institute, Cary, NC).

RESULTS Baseline Characteristics Baseline clinical characteristics are listed in Table 1. The mean age was 79 6 9 years, 16 patients (27%) were female, and 35 (59%) had primary MR while 24 (41%) had a mixed or secondary MR etiology. Patients with mixed or secondary MR were more likely to have coronary artery disease (83% vs 60%, P = .0499), prior cardiac surgery (79% vs 31%, P = .0002), and preprocedural atrial fibrillation (88% vs 63%, P = .03). The mean Society of Thoracic Surgeons mortality score was 6.7 6 4.9% and was not statistically different between groups. The majority of patients had New York Heart Association functional class III or IV symptoms before the procedure (88%). At the time of the TMVR procedure, 28 patients (47%) had more than one clip implanted. Echocardiographic Data Preprocedural 2D echocardiographic data are displayed in Table 2. All patients had moderate to severe (14%) or severe MR (86%) at baseline. The mean left ventricular ejection fraction was 52 6 13%, and the mean left ventricular end-diastolic diameter (LVEDD) was 58 6 7 mm. Among patients with primary MR, 25 of 35 (71%) had a flail scallop, and there was a trend toward higher likelihood of optimal MR reduction in this group (82% vs 54%, P = .08). An MR jet origin at A2-P2 was present in 36 of 59 patients (61%), and this was similar in patients with versus those without optimal MR reduction. Three-dimensional quantitative echocardiographic data are displayed in Table 3. In primary MR, tenting volume (0.7 6 0.6 vs 1.1 6 0.7 mL, P = .045) and tenting height (3.5 6 1.7 vs 4.9 6 1.7 mm, P = .03) were significantly lower in patients with optimal MR reduction. In contrast, in patients with mixed or secondary MR tenting volume and tenting height were similar in those with versus without optimal MR reduction (2.2 6 1.2 vs 2.7 6 1.4 mL [P = .42] and 6.8 6 2.4 vs 6.7 6 2.2 mm [P = .89]). However, in mixed or secondary MR, mitral annular height was increased in patients with optimal MR reduction compared with those without (6.5 6 1.4 vs 4.6 6 1.4 mm, P = .01).

Oguz et al 3

Factors Associated with Optimal MR Reduction Postprocedural MR was graded none or trivial in two of 59 (3%), mild in 38 of 59 (64%), moderate in 14 of 59 (24%), and moderate to severe in five of 59 (9%) patients. Optimal MR reduction (less than moderate) was present in 40 of 59 patients (68%). The clinical factors of patient age, atrial fibrillation, diabetes mellitus, hypertension, prior myocardial infarction, and prior coronary artery bypass surgery did not correlate with optimal MR reduction (P = NS for all). Univariate echocardiographic correlates of optimal MR reduction for patients with primary and mixed or secondary MR are shown in Table 4. Two-dimensional flail gap, flail width, regurgitant orifice area, regurgitant volume, A2-P2 jet location, and the number of implanted clips were not correlated with optimal MR reduction. In patients with primary MR, smaller tenting height (odds ratio, 1.61 per 1-mm increase; 95% CI, 1.03–2.52; P = .025) and smaller tenting volume (OR, 3.08 per 1-mL increase; 95% CI, 0.95–9.98; P = .049) were associated with increased likelihood of optimal MR reduction. Tenting height < 3 mm and tenting volume < 0.7 mL were optimal cut points to maximize sensitivity and specificity; sensitivity and specificity, respectively, for these cut points were 50% and 92% for tenting height and 63% and 77% for tenting volume. Patients with primary MR and tenting height < 3 mm were significantly more likely to have optimal MR reduction (92% vs 48%; relative risk, 1.92; 95% CI, 1.21–3.04; P = .01), as were patients with tenting volume < 0.7 mL (81% vs 47%; relative risk, 1.72; 95% CI, 1.01– 2.91; P = .03). Figure 2A shows the percentage of patients with primary MR achieving optimal MR reduction according to tenting height < 3 or $3 mm. Patients with primary MR and tenting height $ 3 mm had larger LVEDD (58 6 6 vs 53 6 6 mm, P = .03), anterior leaflet angle (16 6 6 vs 10 6 6 , P = .007), and posterior leaflet angle (23 6 8 vs 14 6 10 , P = .01) and smaller nonplanar mitral leaflet angle (141 6 11 vs 157 6 15 , P = .006). In patients with primary MR, LVEDD was weakly correlated with tenting height (r = 0.37, P = .04) and tenting volume (r = 0.45, P = .01). Patients with primary MR and tenting height $ 3 mm had similar mitral annular area (1,243 6 255 vs 1,294 6 284 mm2, P = .61) and mitral leaflet area (1,438 6 316 vs 1,493 6 369 mm2, P = .66) compared with those with tenting height < 3 mm. When evaluating the subset of patients with flail leaflet versus the remaining cohort, patients with flail leaflet had lower mean tenting height (3.7 6 1.6 vs 6.2 6 2.4 mm, P < .001) and tenting volume (0.8 6 0.6 vs 2.0 6 1.2 mL, P < .001). However, tenting height $ 3 mm was present in 16 of 25 patients (64%) and tenting volume $ 0.7 mL was present in 12 of 25 patients (48%) with flail leaflet. Both tenting height (P = .038) and tenting volume (P = .018) remained significantly correlated with optimal MR reduction on logistic regression analysis in the subset of patients with flail leaflet. In patients with mixed or secondary MR, tenting height did not influence outcomes (P = NS). However, decreased annular height (odds ratio, 0.23 per 1-mm decrease; 95% CI, 0.06–0.94) was associated with reduced likelihood of optimal MR reduction. Mitral annular height > 5.5 mm had sensitivity of 83% and specificity of 83% to predict optimal MR reduction. Patients with mixed or secondary MR and annular height > 5.5 mm were significantly more likely to have optimal MR reduction (94% vs 38%; relative risk, 2.50; 95% CI, 1.01–6.17; P = .03). Figure 2B shows the percentage of patients with mixed or secondary MR achieving optimal MR reduction according to annular height > 5.5 versus #5.5 mm.

4 Oguz et al

Journal of the American Society of Echocardiography - 2019

Figure 1 Three-dimensional TEE mitral valve modeling and relevant measurements. (A) Anterolateral-posteromedial diameter (red double arrow). (B) Anteroposterior diameter (red double arrow). (C) Commissure-to-commissure distance (red double arrow). (D) Annular height (red double arrow). (E) Tenting volume (green shading). (F) Tenting height (red double arrow). (G) Nonplanar mitral valve angle (NPA), anterior leaflet angle (ALA), and posterior leaflet angle (PLA). (H) Three-dimensional mitral annular circumference (red line). A, Anterior; AL, anterolateral; Ao, aortic annulus; P, posterior; PM, posteromedial.

Oguz et al 5

Journal of the American Society of Echocardiography Volume - Number -

Table 1 Baseline clinical characteristics

All patients (N = 59)

Patients with primary MR (n = 35)

Patients with mixed/ secondary MR (n = 24)

Age (y)

79 6 9

80 6 9

77 6 9

.073

Sex, female

16 (27)

10 (29)

6 (25)

.76

Prior stroke or TIA

7 (12)

2 (6)

5 (21)

.079

Peripheral arterial disease

9 (15)

4 (11)

5 (21)

.33

Current smoker

4 (7)

2 (6)

2 (8)

.70

15 (25)

7 (20)

8 (33)

.25

Parameters

Diabetes

P

Hypertension

44 (75)

27 (77)

17 (71)

.59

Hyperlipidemia

38 (64)

23 (66)

15 (63)

.80

1.7 6 1.4

1.6 6 1.2

1.9 6 1.6

.19

COPD

20 (34)

13 (37)

7 (29)

.52

Prior myocardial infarction

15 (25)

8 (23)

7 (29)

.59

Coronary artery disease

41 (69)

21 (60)

20 (83)

.0499

Atrial fibrillation

43 (73)

22 (63)

21 (88)

.03

Creatinine

Preprocedural NYHA functional class III IV

.68 39 (66)

24 (69)

Mean tenting height increased slightly after the procedure in patients with primary MR (4 6 2 vs 5 6 3 mm, P = .001) and remained weakly correlated with postprocedural LVEDD (r = 0.53, P = .004). Mitral annular height did not change significantly before versus after the procedure (7 6 2 vs 7 6 3 mm, P = .29). Patients with prior coronary artery bypass surgery had smaller annular height (6 6 1 vs 7 6 3 mm, P = .009), and there was a trend for reduced annular height in patients with prior myocardial infarction (7 6 2 vs 6 6 2 mm, P = .06). Mitral annular height was not significantly associated with atrial fibrillation, diabetes mellitus, hypertension, or LVEDD (P = NS for all). Reproducibility and Timed Analysis Inter- and intraobserver variability was assessed in a subgroup of 20 patients; mean per patient time for the analysis was 5.2 6 1.3 min. Intraobserver linear correlations were good when calculating tenting volume (r = 0.98, P < .0001), tenting height (r = 0.93, P < .0001), and annular height (r = 0.88, P < .0001)(Supplemental Figure 1, available at www.onlinejase.com). Single-measures intraclass correlation coefficient for interobserver variability was 0.97 (95% CI, 0.93–0.99) for tenting volume, 0.93 (95% CI, 0.82–0.97) for tenting height, and 0.85 (95% CI, 0.66–0.94) for annular height. Single-measures intraclass correlation coefficient for intraobserver variability was 0.94 (95% CI, 0.86–0.98) for tenting volume, 0.89 (95% CI, 0.75–0.95) for tenting height, and 0.76 (95% CI, 0.50–0.90) for annular height.

15 (63)

13 (22)

7 (20)

6 (25)

Mean STS mortality score (%)

6.7 6 4.9

6.2 6 5.6

7.3 6 3.6

Prior cardiac surgery

30 (51)

11 (31)

19 (79)

DISCUSSION .06 .0002

COPD, Chronic obstructive pulmonary disease; NYHA, New York Heart Association; STS, Society for Thoracic Surgeons; TIA, transient ischemic attack. Continuous variables are expressed as mean 6 SD; dichotomous variables are expressed as number (percentage).

A group with suboptimal anatomy was identified by combining patients with mixed or secondary MR with annular height # 5.5 mm and patients with primary MR with tenting height $ 3 mm (n = 31 [53%]). Patients with suboptimal anatomy were more likely to have significant residual MR after the procedure (Figure 3). The mean follow-up time was 254 days. After the TMVR procedure, two of 59 patients (3%) had surgical mitral valve repair or replacement, one patient (2%) had redo TMVR, 17 patients (29%) had recurrent moderate to severe or greater MR, and mortality occurred in six patients (10%). Surgical mitral valve replacement or redo TMVR occurred exclusively in patients with suboptimal anatomy defined by 3D echocardiography (10% vs 0%, P = .045). Patients with suboptimal anatomy were more likely to develop recurrent moderate to severe or greater MR or undergo redo mitral valve intervention during intermediate-term follow-up (P = .01; Figure 4). Postprocedural Measurements There was no significant change in LVEDD before and after the procedure (58 6 7 vs 57 6 7 mm, P = .30). Patients with primary MR and increased preprocedural tenting height ($3 mm) had increased postprocedural tenting height (6 6 3 vs 4 6 2 mm, P = .007).

Transcatheter mitral valve repair and replacement is a novel and rapidly evolving field that offers a therapeutic option for patients with significant MR at high or prohibitive risk for conventional open surgery. However, transcatheter mitral therapies also offer a unique set of challenges with respect to patient selection to optimize procedural efficacy. Recent randomized data in patients with secondary MR have shown conflicting results with respect to procedural efficacy and patient outcomes.6,7 Our data suggest that quantitative 3D echocardiography and mitral valve modeling can play an important role in patient selection and improving procedural efficacy. Important findings include the following: (1) ‘‘specifically oriented’’ 3D quantitative modeling of the mitral valve was a more powerful predictor of optimal MR reduction compared with ‘‘conventional’’ 2D TEE parameters, (2) mitral leaflet tenting in patients with primary MR is associated with a lower likelihood of optimal MR reduction, (3) reduced annular height in patients with mixed or functional MR is associated with a lower likelihood of optimal MR reduction, and (4) patients with suboptimal anatomy defined by 3D TEE imaging are more likely to have recurrent moderate to severe or greater MR and undergo redo mitral valve intervention during intermediate-term follow-up. Primary MR is characterized by abnormalities of the mitral leaflets or apparatus. Primary MR constituted 59% of our cohort, and the majority of these patients (71%) had flail leaflets. There was a nonsignificant trend toward improved MR reduction in patients with flail leaflets compared with those without, although this did not reach statistical significance, perhaps because of the small sample size. Previous data from our center have shown that flail leaflet is typically associated with optimal MR reduction,11 and our present results are concordant. However, frequently cited markers of procedural success such as A2-P2 jet location, flail width, and flail gap were not predictive of

6 Oguz et al

Journal of the American Society of Echocardiography - 2019

Table 2 Preprocedural 2D echocardiographic data Primary MR (n = 35) Optimal MR reduction (n = 22)

All patients (N = 59)

Parameter

Mixed/secondary MR (n = 24)

Suboptimal MR reduction (n = 13)

P

Optimal MR reduction (n = 18)

Suboptimal MR reduction (n = 6)

P

LVEF (%)

52 6 13

59 6 9

61 6 9

.27

42 6 12

40 6 10

.62

LVESD (mm)

41 6 9

37 6 5

36 6 7

.39

49 6 9

47 6 11

.82

LVEDD (mm)

58 6 7

56 6 6

58 6 7

.46

60 6 7

58 6 10

.89

LAVI (mL/m2)

70 6 24

64 6 24

73 6 17

.12

66 6 22

93 6 36

.09

Preprocedural mitral mean diastolic gradient (mm Hg)*

2.6 6 1.4

2.6 6 1.2

3.2 6 2

.66

2.3 6 1

2.3 6 1

0.46 6 0.19

0.54 6 0.27

0.42 6 0.09

Preprocedural regurgitant volume (mL)

69 6 19

74 6 27

Flail scallop

25 (42)

18 (82)

Preprocedural ERO (cm2)

.72

.19

0.40 6 0.11

0.50 6 0.11

67 6 10

.5

66 6 16

63 6 9

.92

7 (54)

.08

0 (0)

0 (0)

—

.09

Flail gap > 10 mm

2 (3)

1 (5)

1 (8)

.49

0 (0)

0 (0)

—

Flail width > 15 mm

5 (8)

4 (18)

1 (8)

.65

0 (0)

0 (0)

—

A2-P2 flail

20 (34)

15 (25)

5 (8)

.09

0 (0)

0 (0)

—

A2-P2 jet origin

36 (61)

13 (59)

8 (62)

.89

11 (61)

4 (67)

.81

Multiple regurgitant jets

35 (59)

12 (20)

9 (69)

.39

11 (61)

3 (50)

.63

ERO, Estimated regurgitant orifice; LAVI, left atrial volume index; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter. Continuous variables are expressed as mean 6 SD; dichotomous variables are expressed as number (percentage). *Preprocedural mean diastolic gradient was available in 42 patients.

Table 3 Preprocedural 3D echocardiographic data Primary MR (n = 35)

Parameter

All patients (N = 59)

Optimal MR reduction (n = 22)

Suboptimal MR reduction (n = 13)

Mixed/secondary MR (n = 24) P

Optimal MR reduction (n = 18)

Suboptimal MR reduction (n = 6)

P

AL-PM diameter (mm)

40.5 6 5.3

41.2 6 6.2

41.2 6 4.7

.88

39.6 6 4.4

39.8 6 6.4

.97

AP diameter (mm)

36.2 6 3.8

36.4 6 3.8

36.9 6 2.4

.58

36.2 6 3.4

34.5 6 6.9

.18

6.5 6 2.2

7.3 6 2.5

6.4 6 2.4

.27

6.5 6 1.4

4.6 6 1.4

.01

3D circumference (mm)

130.5 6 14.1

132.4 6 16.5

132.2 6 10.8

.88

128.2 6 11.2

126.3 6 19.2

.57

3D minimum area (mm2)

1,271 6 271

1,302 6 308

1,299.8 6 218

.81

1,230 6 216

1,218 6 411

.55

1.5 6 1.2

0.7 6 0.6

1.1 6 0.7

2.2 6 1.2

2.7 6 1.4

.42

Annular height (mm)

Tenting volume (mL)

.045

Prolapse volume (mL)

0.7 6 1.1

1.2 6 1.6

0.8 6 0.9

.46

0.2 6 0.2

0.1 6 0.2

.21

Tenting height (mm)

5.1 6 2.4

3.5 6 1.7

4.9 6 1.7

.03

6.8 6 2.4

6.7 6 2.2

.89 .13

Prolapse height (mm)

3.3 6 2.6

4.6 6 2.7

3.7 6 3

.23

2.1 6 1.1

1.4 6 0.8

Nonplanar angle ( )

140 6 18

148 6 15

145 6 14

.54

130 6 15

128 6 22

.95

15 6 8

13 6 6

15 6 8

.53

17 6 7

21 6 16

.79

Anterior leaflet angle ( ) Posterior leaflet angle ( ) CC diameter (mm)

25 6 12

19 6 10

21 6 9

.40

34 6 11

31 6 13

.48

22.2 6 4.9

21.9 6 4.6

23.5 6 4.9

.50

21.5 6 4.2

22.3 6 7.8

.89

AL-PM, Anterolateral-posteromedial; AP, anteroposterior; CC, commissure-to-commissure. Data are expressed as mean 6 SD.

procedural success. It is possible that these would be correlated with MR reduction in a larger cohort, but our data indicate that 3D quantitative parameters are more strongly associated with acute procedural success. In patients with primary MR and increased leaflet mobility (Carpentier classification type II), mitral leaflet prolapse and flail are thought to be the dominant pathology. However, our data indicate

that a substantial proportion of these patients have some degree of regional mitral leaflet tenting that coexists with mitral leaflet prolapse (Figure 5). This finding is consistent with previous data, indicating that regional leaflet tenting may be an important mechanism of worsening MR in patients with posterior mitral leaflet prolapse.17 Our data take this a step further, indicating that two markers of leaflet tenting, maximal tenting height and tenting volume, are associated with

Oguz et al 7

Journal of the American Society of Echocardiography Volume - Number -

Table 4 Univariate correlates of optimal MR reduction Primary MR (n = 35) Parameter

P value

OR (95% CI)

Mixed/secondary MR (n = 24) AUC

P value

OR (95% CI)

AUC

0.0002 (1.08  10 ) 8

0.74 0.52

ERO

.09

63.6 (0.17–23,525.13)

0.64

.05

Regurgitant volume

.38

1.02 (0.98–1.06)

0.57

.72

1.01 (0.94–1.09)

Multiple MR jets

.39

0.53 (0.13–2.27)

0.57

.63

1.57 (0.25–10.09)

0.56

Multiple clips

.51

1.6 (0.40–6.46)

0.56

.34

0.4 (0.06–2.77)

0.61

A2-P2 jet location

.89

0.90 (0.22–3.68)

0.51

.81

0.79 (0.11–5.49)

0.53

Flail leaflet

.08

3.86 (0.83–17.94)

0.64

—

—

—

Flail gap

.49

0.35 (0.02–6.57)

0.54

—

—

—

Flail width

.79

1.04 (0.79–1.37)

0.55

—

—

Annular height, 3D*

.28

0.84 (0.61–1.16)

0.61

.004

0.23 (0.06–0.94)

0.84

Tenting volume, 3D†

.049

3.08 (0.95–10.00)

0.71

.45

1.34 (0.63–2.86)

0.61

Tenting height, 3D‡

.03

1.61 (1.03–2.52)

0.72

.90

0.97 (0.64–1.47)

0.52

—

ERO, Estimated regurgitant orifice; OR, odds ratio. *Per 1-mm decrease. † Per 1-mL decrease. ‡ Per 1-mm decrease.

optimal MR reduction in patients with primary MR undergoing TMVR. Even a small degree of tenting height or tenting volume ($3 mm or $0.7 ml, respectively) was associated with less than optimal rates of MR reduction. Additionally, in the subset of patients with flail leaflets, in whom mitral leaflet tenting is not frequently considered, tenting height and tenting volume remain important correlates of MR reduction. Patients with primary MR and increased tenting height had larger LVEDD, and tenting height correlated with LVEDD both before and after the procedure. Transcatheter edge-to-edge repair does not appear to significantly alter tenting height after the procedure; those with increased tenting before the procedure had persistently elevated tenting after the procedure. In some patients with primary MR, left ventricular eccentric remodeling may play a role in persistent leaflet tethering and ongoing MR after the procedure.18,19 These data indicate that it is important to look beyond the site of regurgitation when evaluating patients before the procedure because mitral leaflet tenting even remote from the site of regurgitation has potential implications in successful MR reduction. Further study is needed to evaluate whether patients with primary MR and increased leaflet tenting may respond more favorably to more aggressive medical therapy for heart failure, alternative mitral valve repair techniques, or mitral valve replacement. Three-dimensional TEE predictors of MR reduction differed in patients with primary versus mixed or secondary MR. In the subset of patients with mixed or secondary MR, parameters of mitral leaflet tenting were not associated with the degree of MR reduction. Instead, a parameter of mitral annular nonplanarity, annular height, was more closely associated. The normal mitral annulus has a saddle shape, with anterior and posterior points that are superior to the medial and lateral points (Figure 1D). In our cohort, patients with prior coronary artery bypass surgery or prior myocardial infarction tended to have a more planar mitral annulus (smaller annular height). Previous studies have indicated that a flatter, more planar mitral annulus is associated with increased MR in functional MR,20,21 and restoration of the normal nonplanar shape of the mitral annulus is an important component of durable mitral valve repair.22,23

Similarly, in our cohort of patients with mixed or functional MR, an annular height > 5.5 mm had superior MR reduction compared with patients with an annular height # 5.5 mm. Taken together with prior data, this indicates that patients with a more planar mitral annulus and mixed or secondary MR may benefit from repair techniques that more directly address the abnormal mitral annulus. Our data indicate that 3D mitral valve modeling could play a role in patient selection before TMVR, particularly as the number of available repair and replacement devices increases. Recurrent moderate to severe or greater MR or subsequent repeat mitral valve procedures (mitral valve replacement or redo TMVR) occurred more commonly in patients with suboptimal anatomy, although this result should be interpreted with caution given the small number of patients requiring repeat mitral valve procedures (three of 59 [5%]). It will be important to validate this finding in larger groups of patients with long-term follow-up, but it is our hope that a more sophisticated understanding of the mechanism of MR and the underlying pathology may allow a more individualized approach to transcatheter repair or replacement strategies in the future. The landscape of transcatheter mitral valve repair and replacement is rapidly evolving. Although the only currently US Food and Drug Administration–approved device for TMVR is the MitraClip, a large number of alternative transcatheter repair techniques and transcatheter replacement devices are currently being studied or in various stages of development. Three-dimensional imaging modalities provide a more robust and quantitative understanding of complex and heterogeneous mitral valve pathology that leads to clinically significant MR. A more precise understanding of the mitral valve pathology will in turn facilitate individualized repair or replacement therapy on the basis of patient-specific anatomy in an effort to enhance patient selection and improve outcomes. Limitations Our study was retrospective, was conducted at a single center, and included a small number of patients. It will be important to confirm these findings in larger groups of patients with long-term follow-up.

8 Oguz et al

Journal of the American Society of Echocardiography - 2019

Figure 3 Residual MR after TMVR according to 3D mitral valve anatomy. Suboptimal anatomy for TMVR was defined as mixed or secondary MR with annular height # 5.5 mm or primary MR with tenting height $ 3 mm. Patients with suboptimal anatomy were more likely to have significant residual MR after the procedure (P < .001).

Figure 2 Three-dimensional anatomic correlates of optimal MR reduction in patients with primary and mixed or secondary MR. Patients with primary MR and concurrent maximal leaflet tenting height of $3 mm were less likely to achieve optimal MR reduction compared with those with tenting height < 3 mm (A). In patients with mixed or secondary MR, mitral annular height was similarly predictive of acute procedural success (B). Patients with annular height #5.5 mm were less likely to have optimal MR reduction compared with those with annular height > 5.5 mm.

Given the small sample size, it is possible that important or significant differences were not recognized because of the small number of patients and end points. Additionally, because the study was performed in a single group of patients, the reported cutoff values and sensitivities and specificities should be validated in larger, prospective cohorts. We also recognize that even though reproducibility was within a reasonable range in our cohort, small differences in annular height, tenting height, and tenting volume within the range of variability could have important implications on the final measurement. In the future it will be important to determine whether more automated measurements can improve the reproducibility of these measurements. We chose the end-systolic frame for our analysis because this typically represents the maximum tenting volume and prolapse volume during the cardiac cycle. Recent software advances allow a dynamic

Figure 4 Recurrent more than moderate MR or repeat mitral valve intervention according to mitral valve 3D anatomy. Patients with suboptimal mitral valve anatomy, defined as primary MR with tenting height $ 3 mm or mixed or secondary MR with annular height # 5.5 mm, were more likely to have recurrent more than moderate MR or require open mitral valve replacement during intermediate-term follow-up (P = .01).

assessment of mitral annular and leaflet motion throughout the cardiac cycle. We acknowledge that a dynamic mitral valve model may provide incremental information beyond a single-frame analysis as performed in our study, and this deserves further investigation. The retrospective nature of this analysis allowed offline analysis of previously acquired 3D volumetric data sets. We acknowledge that for some analyses the process of creating a 3D model can be difficult and/or time consuming, which can make it challenging to integrate into routine clinical practice. There have been recent advances in automation of mitral valve modeling that have the potential to reduce the time needed to model the mitral valve and also improve reproducibility. Advances in automation will likely improve the ability to incorporate 3D mitral valve analyses into future clinical practice.

Journal of the American Society of Echocardiography Volume - Number -

Oguz et al 9

Figure 5 Mixed prolapse and tenting in a patient with primary MR and suboptimal MR reduction after TMVR. (A) Three-dimensional TEE mitral valve modeling displayed from the left atrial perspective demonstrating prolapsing scallops laterally (shaded red) and tented scallops medially (shaded blue) in a patient with primary MR and bileaflet prolapse. A maximal prolapse height of 5 mm was present laterally, and a maximal tenting height of 5 mm was seen medially in the same patient. (B) Three-dimensional TEE zoomed image of the mitral valve displayed from the left atrial perspective, showing prolapsed (red arrow) and tented (blue arrow) scallops. (C) Counterclockwise rotation in the long-axis view by 2D TEE imaging shows bileaflet prolapse laterally (red arrow). (D) Clockwise rotation in the long-axis view by 2D TEE imaging shows focal tethering of the posterior leaflet (blue arrow). (E) Preprocedural echocardiographic imaging demonstrated severe MR. (F) Postprocedural imaging demonstrated significant residual MR. A, Anterior; AL, anterolateral; Ao, aortic annulus; P, posterior; PM, posteromedial.

10 Oguz et al

CONCLUSION Three-dimensional quantitative analysis of the mitral valve and apparatus is more closely associated with procedural efficacy for TMVR than available 2D echocardiographic measurements. In primary MR, leaflet tenting volume and tenting height are important correlates of optimal MR reduction, while in mixed or secondary MR, a parameter of mitral annular planarity, annular height, appears to most accurately correlate with optimal MR reduction. These findings warrant further investigation in larger groups of patients but could have important implications in patient selection for TMVR in the future. ACKNOWLEDGMENT We thank the Turkish Society of Cardiology for a research grant supporting this work.

SUPPLEMENTARY DATA Supplementary data related to this article can be found at https://doi. org/10.1016/j.echo.2019.06.014.

REFERENCES 1. Nkomo VT, Gardin JM, Skelton TN, Gottdiener JS, Scott CG, EnriquezSarano M. Burden of valvular heart diseases: a population-based study. Lancet 2006;368:1005-11. 2. Iung B, Baron G, Butchart EG, Delahaye F, Gohlke-B€arwolf C, Levang OW, et al. A prospective survey of patients with valvular heart disease in Europe: the Euro Heart Survey on Valvular Heart Disease. Eur Heart J 2003;24:1231-43. 3. Mirabel M, Iung B, Baron G, Messika-Zeitoun D, Detaint D, Vanoverschelde JL, et al. What are the characteristics of patients with severe, symptomatic, mitral regurgitation who are denied surgery? Eur Heart J 2007;28:1358-65. 4. Dziadzko V, Clavel MA, Dziadzko M, Medina-Inojosa JR, Michelena H, Maalouf J, et al. Outcome and undertreatment of mitral regurgitation: a community cohort study. Lancet 2018;391:960-9. 5. Feldman T, Foster E, Glower DD, Kar S, Rinaldi MJ, Fail PS, et al. Percutaneous repair or surgery for mitral regurgitation. N Engl J Med 2011;364: 1395-406. 6. Obadia JF, Messika-Zeitoun D, Leurent G, Iung B, Bonnet G, Piriou N, et al. Percutaneous repair or medical treatment for secondary mitral regurgitation. N Engl J Med 2018;379:2297-306. 7. Stone GW, Lindenfeld J, Abraham WT, Kar S, Lim DS, Mishell JM, et al. Transcatheter mitral-valve repair in patients with heart failure. N Engl J Med 2018;379:2307-18. 8. Mackensen GB, Lee JC, Wang DD, Pearson PJ, Blanke P, Dvir D, et al. Role of echocardiography in transcatheter mitral valve replacement in native mitral valves and mitral rings. J Am Soc Echocardiogr 2018;31:475-90. 9. Nyman CB, Mackensen GB, Jelacic S, Little SH, Smith TW, Mahmood F. Transcatheter mitral valve repair using the edge-to-edge clip. J Am Soc Echocardiogr 2018;31:434-53.

Journal of the American Society of Echocardiography - 2019

10. Attizzani GF, Ohno Y, Capodanno D, Cannata S, Dipasqua F, Imme S, et al. Extended use of percutaneous edge-to-edge mitral valve repair beyond EVEREST (Endovascular Valve Edge-to-Edge Repair) criteria: 30-day and 12-month clinical and echocardiographic outcomes from the GRASP (Getting Reduction of Mitral Insufficiency by Percutaneous Clip Implantation) registry. JACC Cardiovasc Interv 2015;8:74-82. 11. Thaden JJ, Malouf JF, Nkomo VT, Pislaru SV, Holmes DR Jr., Reeder GS, et al. Mitral valve anatomic predictors of hemodynamic success with transcatheter mitral valve repair. J Am Heart Assoc 2018;7:e007315. 12. Sorajja P, Mack M, Vemulapalli S, Holmes DR Jr., Stebbins A, Kar S, et al. Initial experience with commercial transcatheter mitral valve repair in the United States. J Am Coll Cardiol 2016;67:1129-40. 13. Lesevic H, Karl M, Braun D, Barthel P, Orban M, Pache J, et al. Long-term outcomes after MitraClip implantation according to the presence or absence of EVEREST inclusion criteria. Am J Cardiol 2017;119:1255-61. 14. Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. J Am Soc Echocardiogr 2015;28:1-39.e14. 15. Zoghbi WA, Adams D, Bonow RO, Enriquez-Sarano M, Foster E, Grayburn PA, et al. Recommendations for noninvasive evaluation of native valvular regurgitation: a report from the American Society of Echocardiography developed in collaboration with the society for cardiovascular magnetic resonance. J Am Soc of Echocardiogr 2017;30:303-71. 16. Shanks M, Delgado V, Ng AC, van der Kley F, Schuijf JD, Boersma E, et al. Mitral valve morphology assessment: three dimensional transesophageal echocardiography versus computed tomography. Ann Thorac Surg 2010;90:1922-9. 17. Otani K, Takeuchi M, Kaku K, Haruki N, Yoshitani H, Eto M, et al. Evidence of a vicious cycle in mitral regurgitation with prolapse: secondary tethering attributed to primary prolapse demonstrated by three-dimensional echocardiography exacerbates regurgitation. Circulation 2012;126(11 suppl 1):S214-21. 18. Kim J, Alakbarli J, Palumbo MC, Xie LX, Rong LQ, Tehrani NH, et al. Left ventricular geometry predicts optimal response to percutaneous mitral repair via MitraClip: integrated assessment by two- and threedimensional echocardiography. Catheter Cardiovasc. Interv 2019;93: 1152-60. 19. Lavall D, Mehrer M, Schirmer SH, Reil JC, Wagenpfeil S, Bohm M, et al. Long-term hemodynamic improvement after transcatheter mitral valve repair. J Am Soc Echocardiogr 2018;31:1013-20. 20. Watanabe N, Ogasawara Y, Yamaura Y, Wada N, Kawamoto T, Toyota E, et al. Mitral annulus flattens in ischemic mitral regurgitation: geometric differences between inferior and anterior myocardial infarction: a real-time 3-dimensional echocardiographic study. Circulation 2005;112:I458-62. 21. El Sebaie MH, Abdelatti MN, Zarea AA, Farag AM, Hashem AA, Fadel AM. Assessment of mitral valve geometric deformity in patients with ischemic heart disease using three-dimensional echocardiography. Egypt Heart J 2017;69:13-20. 22. Guenzinger R, Schneider EP, Guenther T, Wolf P, Mazzitelli D, Lange R, et al. Three-dimensional valve repair—the better care? Midterm results of a saddle-shaped, rigid annuloplasty ring in patients with ischemic mitral regurgitation. J Thorac Cardiovasc Surg 2014;148:176-82. 23. Chen TE, Ong K, Suri RM, Enriquez-Sarano M, Michelena HI, Burkhart HM, et al. Three-dimensional echocardiographic assessment of mitral annular physiology in patients with degenerative mitral valve regurgitation undergoing surgical repair: comparison between early- and latestage severe mitral regurgitation. J Am Soc Echocardiogr 2018;31:1178-89.

Journal of the American Society of Echocardiography Volume - Number -

Oguz et al 10.e1

APPENDIX

Supplemental Figure 1 Pearson correlation and Bland-Altman analysis. Comparison of annular height, tenting height, and tenting volume measured by 3D Mitral Valve Navigator QLAB using Pearson correlation (top) and Bland-Altman (bottom).