Characterization of Degenerative Mitral Valve Disease Using

Characterization of Degenerative Mitral Valve Disease Using
Morphologic Analysis of Real-Time Three-Dimensional
Echocardiographic Images
Objective Insight Into Complexity and Planning of Mitral Valve Repair
Sonal Chandra, MD; Ivan S. Salgo, MD, MS; Lissa Sugeng, MD, MPH; Lynn Weinert, BS;
Wendy Tsang, MD; Masaaki Takeuchi, MD; Kirk T. Spencer, MD; Anne O’Connor, MD;
Michael Cardinale, RDCS, RDMS; Scott Settlemier, BA; Victor Mor-Avi, PhD; Roberto M. Lang, MD
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
Background—Presurgical planning of mitral valve (MV) repair in patients with Barlow disease (BD) and fibroelastic
deficiency (FED) is challenging because of the inability to assess accurately the complexity of MV prolapse. We
hypothesized that the etiology of degenerative MV disease (DMVD) could be objectively and accurately ascertained
using parameters of MV geometry obtained by morphological analysis of real-time 3D echocardiographic (RT3DE)
images.
Methods and Results—Seventy-seven patients underwent transesophageal RT3DE study: 57 patients with DMVD studied
intraoperatively (28 BD, 29 FED classified during surgery) and 20 patients with normal MV who were used as control
subjects (NL). MVQ software (Philips) was used to measure parameters of annular dimensions and geometry and leaflet
surface area, including billowing volume and height. The Student t test and multinomial logistic regression was
performed to identify parameters best differentiating DMVD patients from normal as well as FED from BD.
Morphological analysis in the DMVD group revealed a progressive increase in multiple parameters from NL to FED
to BD, allowing for accurate diagnosis of these entities. The strongest predictors of the presence of DMVD included
billowing height and volume. Three-dimensional billowing height with a cutoff value of 1.0 mm differentiated DMVD
from NL without overlap, and billowing volume with a cutoff value 1.15 mL differentiated between FED and BD
without overlap.
Conclusions—Morphological analysis as a form of decision support in assessing MV billowing revealed significant
quantifiable differences between NL, FED, and BD patients, allowing accurate classification of the etiology of MV
prolapse and determination of the anticipated complexity of repair. (Circ Cardiovasc Imaging. 2011;4:24-32.)
Key Words: mitral valve 䡲 echocardiography 䡲 3D imaging 䡲 Barlow disease 䡲 fibroelastic deficiency
䡲 degenerative mitral valve disease
M
cause it relies on qualitative evaluation that requires a high
level of clinical echocardiographic expertise. Quantitative
decision support would be useful in assessing key anatomic
features of DMVD.
itral regurgitation represents a pathophysiological
spectrum of functional and structural defects of the
mitral valve (MV) apparatus. Specifically, degenerative mitral valve disease (DMVD) frequently includes different
degrees of annular dilation, leaflet redundancy, and chordal
dysfunction, which result in variable cardiovascular morbidity and mortality.1 DMVD encompasses 2 broad categories,
fibroelastic deficiency (FED) and Barlow disease (BD).2,3
Accurate diagnosis of these entities along with their specific
location and complexity is important because they require
different surgical planning, which necessitates careful matching of the complexity of reparability with surgical expertise.4 –9 Differential diagnosis in DMVD is challenging be-
Clinical Perspective on p 32
In contrast to 2D echocardiography, real-time 3D echocardiography (RT3DE) using matrix 3D transesophageal technology, with its improved spatial resolution compared with
previously used sparse array transducers, has resulted in
enhanced visualization of the pathomorphology of the mitral
valve, particularly in DMVD.10 –12 We hypothesized that
RT3DE-derived measurements of valvular anatomy could be
Received November 19, 2009; accepted September 21, 2010.
From the University of Chicago Medical Center (S.C., L.S., L.W., W.T., K.T.S., A.O., V.M.-A., R.M.L.), Chicago, Ill; Philips Healthcare (I.S.S., M.C.,
S.S.), Andover, Mass; and the University of Occupational and Environmental Health (M.T.), School of Medicine, Kitakyushu, Japan.
Correspondence to Roberto M. Lang, MD, University of Chicago, MC 5084, 5841 S Maryland Ave, Chicago, IL 60637. E-mail
rlang@bsd.uchicago.edu
© 2011 American Heart Association, Inc.
Circ Cardiovasc Imaging is available at http://circimaging.ahajournals.org
24
DOI: 10.1161/CIRCIMAGING.109.924332
Chandra et al
Table 1.
Morphologic Analysis of Degenerative Mitral Valve Disease
25
Surgical Criteria for the Classification of Degenerative Mitral Valve Disease
FED
BD
Annulus
Normal or near normal valve size (annulus ⬍32 mm)
Severely dilated annulus
Leaflets
Thin transparent w/o excess tissue
Thick w/ excess tissue
Single segment involvement
Multisegmental involvement
Involved segment is thick and distended
Chords
Billowing characteristics
Elongated in the affected segment, w/ or w/o rupture
Thickened, calcified, elongated, restricted w/ or w/o rupture
No billowing of the adjacent segments
Multisegmental billowing
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
used to characterize DMVD objectively. The goals of this
study were (1) to describe the pathomorphology of the MV in
DMVD and (2) to identify specific quantitative 3D echocardiographic parameters that accurately characterize DMVD
and therefore could be used to differentiate patients with
normal, FED, and BD valves, estimate the complexity of
disease, and thus optimize treatment strategy.
Methods
Patients
A total of 80 patients were prospectively studied. Three patients were
excluded because of difficulties encountered with 3D transesophageal echocardiographic (TEE) acquisition and offline postprocessing. Thus, we enrolled into the study a total of 77 patients. This
included 57 consecutive patients with primary DMVD and severe
mitral regurgitation (MR), defined as effective regurgitant orifice
area ⬎0.40 mm2 and/or vena contracta ⬎0.7 cm,13 meeting indication for surgical repair, and 20 control subjects randomly selected
from a pool of patients undergoing TEE for clinical indications other
than MV disease, who did not have any discernible MV pathology.
Patients with MV prolapse associated with connective tissue disorders, Marfan syndrome, rheumatic heart disease, and hypertrophic
cardiomyopathy were excluded. The patients with DMVD underwent intraoperative TEE after induction of general anesthesia and
endotracheal intubation but before cardiopulmonary bypass. At the
time of surgery, these patients were classified by the surgeon as
either BD or FED, based on the criteria listed in Table 1,14 rather
than the 2D TEE findings that were available to the surgeon, but who
was blinded to the 3D TEE images and the results of 3D analysis.
This resulted in 28 patients with BD and 29 with FED. The research
study was approved by the institutional review board, and every
patient signed an informed consent before the TEE study.
Study Design
For the primary aim, 3D MV parameters were measured in all
patients. Additionally, 20 randomly selected DMVD (13 BD; 7 FED)
patients were qualitatively evaluated using 2D and 3D TEE images
by 2 echocardiographers with a combined 15 years of postgraduate
experience, who were blinded to the surgical designations. These
echocardiographers were asked to use the 2D TEE images and then
3D TEE images to identify all the prolapsing and/or flail segments to
determine the etiology of the DMVD (BD versus FED) initially and
overall complexity of the mitral valve lesions. A degenerative MV
was deemed complex on the basis of previously agreed criteria such
as involvement of the central portion of the anterior leaflet, commissural involvement, and bileaflet involvement with multisegmental
prolapsing/billowing or flail components. The 2D and 3D assessment
of segmental lesions was then compared with the surgical findings.
Imaging
After completing the clinical portion of the study, RT3DE imaging
of the MV was performed using an iE33 ultrasound system (Philips
Healthcare, Andover, Mass), equipped with a fully sampled matrix
TEE transducer (X7–2t). Initially, gain settings were optimized using
the narrow-angled acquisition mode, which allows RT3DE imaging
of a pyramidal volume of approximately 30°⫻60° without the need
for ECG gating. Zoomed RT3DE images of the entire MV were then
acquired in a single cardiac cycle, resulting in frame rates between 5
and 18 Hz (mean, 9⫾3 Hz). Acquisition of 3D data sets was repeated
several times to ensure optimal image quality. The optimization and
acquisition required approximately 3 minutes per study.
Image Analysis
Images were reviewed and analyzed offline on an Xcelera workstation (Philips Healthcare) by an echocardiographer blinded to intraoperative designations. To improve visualization, pyramidal datasets
acquired in the zoom mode were cropped along designated X-Y-Z
axes or using a manually positioned cropping plane of choice. The
3D analysis of MV parameters was performed using custom software
(MVQ, QLAB, Philips) as follows. Initially, the end-systolic frame
was defined as the second to last frame before the initiation of the
MV opening. Then, a long-axis view of the mitral apparatus was
used to determine anterior, posterior, anterolateral, and posteromedial annular coordinates. The annulus was then manually outlined by
defining annular points (Figure 1A) in multiple planes rotated around
the axis perpendicular to the mitral annular plane (Figure 1B). The
annulus was then further segmented to identify leaflet geometry and
coaptation points by manually tracing the leaflets in multiple parallel
long-axis planes (Figure 1C) spanning the annulus from commissure
to commissure. The reconstructed MV was subsequently displayed
as a color-coded 3D-rendered surface representing a topographical
map of the mitral leaflets (Figure 1D). The software then automatically generated measurements of key parameters of annular dimensions and geometry and leaflet surface area, including billowing
volume and height. Specifically, these parameters included 2D and
3D annular area, anteroposterior (AP) and commissural (CC) diameters, 3D annular perimeter, intercommissural to anteroposterior
diameter ratio, anterior to posterior nonplanar MV angle, aorticmitral angle, posterior and anterior MV leaflet area, aggregate leaflet
area, billowing height and volume of the anterior segments (A1, A2,
and A3), and posterior scallops (P1, P2, and P3); and aggregate
billowing height and volume. This 3D image analysis took 5 to 10
minutes per data set, depending on image quality and complexity of
the lesion.
Statistical Analysis
Statistical analysis was performed using Microsoft Excel with R 2.8.
Continuous variables were expressed as mean⫾SD and plotted for
each group. Intergroup differences were compared using unpaired
Student t test. Statistical significance was indicated by a probability
value of ⬍0.05. Multinomial logistic regression was performed to
determine which measurements are best predictors of complexity of
DMVD, specifically, which parameters are best at distinguishing
normal from FED and from BD patients. A classification tree
analysis using recursive partitioning was developed to distinguish
normal patients from those with BD and FED and to indicate a
preliminary indication of an optimal cutoff value for the presence of
DMVD and specifically, FED versus BD. In addition, receiver
operating characteristic (ROC) analysis was performed to assess the
superiority of these predictors. The area under the curve (AUC) was
calculated for these parameters. The agreement between the qualitative analysis of 2D and 3D TEE images by 2 experts with surgical
findings, as well as the interobserver agreement, were tested using ␬
26
Circ Cardiovasc Imaging
January 2011
Figure 1. Three-dimensional morphological analysis of a normal mitral valve. Mitral annulus is manually initialized in 1 plane (A) and then repeated in
multiple rotated planes and interpolated (B); MV
leaflets are manually traced from commissure to
commissure in multiple parallel planes (C); the
resultant surface is displayed as a color-coded,
3D-rendered valve surface (D).
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
statistics. The calculated ␬ coefficients were judged as follows: 0 to
0.2, low; 0.21 to 0.4, moderate; 0.41 to 0.6, substantial; 0.61 to 0.8,
good; and ⬎0.8, excellent.
between repeated measurements expressed as a percentage of their
mean.
Reproducibility Analysis
DMVD patients were initially compared with the normal
control subjects. Patient characteristics for the study group
are shown in Table 2. As expected, the FED patients were on
average older than BD patients. Figure 2 shows surgical
views, zoomed RT3DE images, and their respective colorcoded parametric 3D renderings of the MV obtained in 2
patients with different types of DMVD.
Results
Reproducibility of the 3D morphological parameters was assessed in
20 patients, including 10 patients with FED and 10 patients with BD.
Intraobserver variability was assessed using repeated measurements
performed by the same observer a month later, whereas interobserver
variability was evaluated by repeating the analysis by a second
independent observer, blinded to the results of all prior measurements. Variability was expressed in terms of coefficients of variation
Figure 2. Examples of views of the MV from the left atrial perspective obtained in 2 patients with DMVD. Top, FED with a P2 flail leaflet
and ruptured chords; bottom, BD with multisegmental billowing: A, surgical views; B, zoomed 3D echocardiographic views; and C, corresponding 3D-rendered, color-coded images.
Chandra et al
Morphologic Analysis of Degenerative Mitral Valve Disease
Table 2. Clinical Characteristics of the Patients With
Degenerative Mitral Valve Disease (29 Patients With
Fibroelastic Deficiency and 30 Patients With Barlow Disease)
and 20 Control Subjects
Control Subjects
(n⫽20)
FED
(n⫽29)
BD
(n⫽28)
46
49
51
Male sex, %
Mean age, y
Mean BSA, kg/m
63⫾13
61⫾7
53⫾9*‡
1.71⫾0.24
1.76⫾0.31
1.81⫾0.19
77⫾14
75⫾15
Mean heart rate, bpm
73⫾13
Blood pressure, mm Hg
143/83⫾27/16
123/76⫾22/9* 129/67⫾27/11*
significant increase in mean 3D annular area in DMVD
patients versus the normal control subjects (Table 3). In
addition, both AP and CC diameters were increased in the
DMVD patients compared with control subjects. However,
the ratio of CC to AP dimension was reduced to 1.06⫾0.30 in
BD patients, reflecting the more circular shape of the mitral
annulus compared with the more oval shape in control
subjects and FED patients (Table 3). Although no intergroup
differences were noted in 3D annular height and aortic-mitral
angle, a significant increase in the nonplanar angle was
evident in BD patients.
Leaflet Area
BSA indicates body surface area.
*P⬍0.05 versus normal control subjects; ‡P⬍0.05 versus FED.
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
Although data obtained in the control subject showed no
pathology, both DMVD patient groups showed variable
degrees of involvement, and, as expected, larger mitral
annulus and leaflet area, in conjunction with multisegmental
billowing was present in the BD patient group. The summary
of results of the 3D morphological measurements of the
annular geometry, leaflet surface area and prolapse characteristics is shown in Table 3.
Annular Geometry
There was a significant and progressive increase in 3D
annular dimensions from control subjects to FED and even
further in BD patients (Figure 3A and 3B). There was a
Table 3. Summary of Volumetric Measurements of Mitral
Valve Anatomy in Patients With Degenerative Mitral Valve
Disease (29 Patients With Fibroelastic Deficiency and 28
Patients With Barlow Disease) and the 20 Control Subjects
Three-dimensional annular
area, mm2
Two-dimensional annular
area, mm2
AP diameter, mm
CC diameter, mm
CC/AP ratio
Three-dimensional
perimeter, mm
Leaflet area, mm2
Anterior leaflet area, mm2
Posterior leaflet area, mm2
Aortic-mitral angle, degrees
Nonplanar angle, degrees
Billowing height, mm
Billowing volume, mL
A1 billowing volume, mL
A2 billowing volume, mL
A3 billowing volume, mL
P1 billowing volume, mL
P2 billowing volume, mL
P3 billowing volume, mL
27
Control Subjects
FED
BD
797⫾152
1154⫾261*
1834⫾432*†
769⫾148
1122⫾250*
1792⫾426*†
28⫾3
33⫾3
1.21⫾0.1
103⫾10
34⫾5*
40⫾4*
1.21⫾0.1
128⫾17*
44⫾8*†
45⫾9*†
1.06⫾0.3*
158⫾19*†
943⫾218
472⫾109
419⫾102
120⫾10
124⫾11
0.27⫾0.23
0⫾0
0⫾0
0⫾0
0⫾0
0⫾0
0⫾0
0⫾0
1453⫾354* 2302⫾455*†
722⫾178* 1162⫾276*†
735⫾214* 1175⫾306*†
124⫾8
123⫾10
129⫾18
155⫾20*†
4.10⫾2.37* 8.41⫾2.91*†
0.34⫾0.34* 4.71⫾2.52*†
0.03⫾0.04* 0.51⫾0.46*†
0.03⫾0.07* 0.81⫾0.68*†
0.03⫾0.06* 0.67⫾0.52*†
0.05⫾0.11* 0.82⫾0.61*†
0.29⫾0.40* 1.42⫾0.92*†
0.08⫾0.10* 0.54⫾0.45*†
A1, A2, and A3 are segments 1 through 3 of the anterior leaflet; P1, P2, and
P3 are scallops 1 through 3 of the posterior leaflet.
*P⬍0.05 versus normal control subjects; †P⬍0.05 versus FED.
The total leaflet surface area was increased in FED and BD
patients (Table 3). Both anterior and posterior leaflet areas were
increased in DMVD patients (BD significantly greater than
FED) compared with the normal valves; there was no discernible
difference in the degree of enlargement of the anterior versus the
posterior leaflet area in the 2 DMVD subgroups.
Predictors of Complexity
The t test analysis identified multiple parameters that distinguished DMVD from control subjects. Of these, billowing
height and volume were the strongest predictors, because
other parameters had considerable intergroup overlapping
ranges diminishing their predictive power. Billowing height
demonstrated minimal overlap and therefore was the strongest discriminator between control subjects and DMVD
(Figure 3C). However, this parameter was not a strong
discriminator between the FED and BD patients. Billowing
volume, identified by t test as a statistically significant
parameter, demonstrated minimal overlap between FED and
BD, serving as the strongest identifier of DMVD geometric
complexity in these patients (Figure 3D). Multinomial regression similarly identified leaflet billowing height and volume
as strong predictors for the detection of DMVD and evaluation of its etiology and complexity. Again, billowing height
was found to be paramount in separating control subjects
from DMVD patients, and billowing volume was the strongest parameter distinguishing between the 2 DMVD subgroups. As expected, multisegmental prolapse was evident in
BD patients with all segments billowing to a greater degree
than their counterparts FED patients. In BD patients, the P2
scallop demonstrated the greatest degree of billowing in the
posterior leaflet, followed by P1 and P3 scallops (Table 3).
Optimal Cutoff Values
Classification tree analysis using recursive partitioning demonstrated that patients with billowing volumes of ⬍1.15 mL and
with billowing heights of ⬍1 mm were normal; in contrast,
patients with a billowing volumes of ⬍1.15 mL and prolapse
heights of ⬎1 mm had FED, whereas those with billowing
volumes ⬎1.15 mL had BD (Figure 4). Indeed, the highest
billowing height in the data set for normal patients was 0.9 mm,
and the lowest for a DMVD patient was 1.1 mm. Similarly, the
highest billowing volume among FED patients was 1.1 mL,
while the lowest prolapse volume among BD patients was 1.15
mL. The use of these cutoff values resulted in an algorithm for
objective, quantitative differential diagnosis of DMVD (Figure 5).
28
Circ Cardiovasc Imaging
January 2011
Figure 3. Graphic representation of measurements of MV geometry in the 57
patients with DMVD (29 patients with FED
and 28 patients with BD) and the 20 control subjects. Solid squares with error bars
represent mean⫾ SD; diamonds represent
maximal and minimal values. *P⬍0.05 versus normal; ‡P⬍0.05 versus FED. AP
indicates anteroposterior. See text for
details.
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
ROC Results
ROC curve analysis revealed high AUC values for 2 parameters:
(1) AUC for 3D measurements of billowing height with cutoff
value of ⬎1.0 mm was 0.98; and (2) AUC for 3D measurement
of billowing volume with cutoff of ⬎1.15 mL was 1.0.
Qualitative 2D Versus 3D Evaluation
In the subset of 20 patients tested for diagnostic accuracy of 2D
versus 3D TEE images using surgical inspection as the gold
standard, accurate scallop involvement was identified by 2D
TEE in 76% (46/61, P⬍0.05) and by 3D in 92% (56/61,
P⬎0.05). Surgical agreement with 2D TEE findings was highest
for P2 and A2 segments and lowest for P3, A1, and A3
segments. Commissural involvement was ascertained surgically
in 4 of 20 patients. In none of these 4 patients, the diagnosis was
made by 2D TEE images, whereas with 3D TEE images, the
diagnosis was made in 3 of 4 cases by both observers. Fifteen of
the 20 patients met the criteria for complex mitral valve disease,
based on surgical inspection. Of these, observer 1 diagnosed
using 2D images 10 patients as having complex MV disease
(␬⫽0.50), whereas observer 2 identified 9 patients (␬⫽0.43).
Using 3D images, observer 1 diagnosed 14 patients (␬⫽0.88)
and observer 2 diagnosed 13 (␬⫽0.77) as having complex
disease. In addition, interobserver agreement identifying lesions
by 2D versus 3D morphological analysis was 78% (␬⫽0.56) and
91% (␬⫽0.80), respectively.
Reproducibility of Quantitative 3D Analysis
The mean interobserver and intraobserver variability values
are summarized in Table 4. As expected, the interobserver
variability was greater than the intraobserver variability in
both FED and BD patients. In both groups, for all 4
Figure 4. Scatterplots demonstrating the cutoff
values of 1.0 mm for billowing height (vertical
dashed line) and cutoff value of 1.15 mL for billowing volume (horizontal dashed line) in differentiating
patients with BD versus FED.
Chandra et al
Morphologic Analysis of Degenerative Mitral Valve Disease
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
Figure 5. Clinical paradigm for use of morphological analysis of
real-time 3D echocardiographic images for differential diagnosis
of DMVD to distinguish FED from BD.
parameters the interobserver variability was below 15%,
whereas the intraobserver variability was below 10%, reflecting good reproducibility and comparable to that of most
clinically used echocardiography based measurements.
Discussion
Currently, surgical inspection is the gold standard for making
etiologic differentiation in DMVD patients. However, the
ease of reparability, which is frequently not fully appreciated
until the time of surgery, is determined by complexity of MV
lesions. The lack of reliable preoperative assessment frequently leads to inadequate match between the complexity of
disease and the surgical expertise which becomes obvious
only in the operating room. As a result, difficulties with
presurgical planning and decision support may lead to either
unsuccessful repair or conversion to replacement with poor
outcome in patients with complex valvular disease.15–17 Accordingly, we sought to and successfully demonstrated the
feasibility of using RT3DE-derived volumetric measurements
of valvular anatomy to objectively evaluate MV annular and
leaflet distortion in patients with DMVD before surgery.
Furthermore, we found that this morphological analysis was
able to differentiate between degenerative and normal valves,
and, more importantly, between the 2 different etiologies of
DMVD, namely BD and FED. This finding could support
Table 4. Reproducibility Data for 3D Morphologic Parameters
in Patients With Degenerative Mitral Valve Disease (10 Patients
With Fibroelastic Deficiency and 10 Patients With Barlow
Disease), Expressed as Coefficients of Variation (in %)
Interobserver
Variability
Intraobserver
Variability
FED
BD
FED
BD
Three-dimensional annular surface area
5.5
7.5
3.4
4.4
AP diameter
6.7
6.1
4.6
4.1
Three-dimensional billowing height
12.6
10.9
6.7
5.4
Three-dimensional billowing volume
11.8
12.3
5.9
6.6
29
reliably preoperative automated clinical decision-making and
thus have important surgical implications.
Although slight bulging of the MV leaflets into the left
atrium during MV closure is normal, in its exaggerated state,
termed “billowing” by Barlow, it is usually associated with a
midsystolic click, excess tissue, and loss of leaflet apposition.18 Prolapse is defined by the failure of MV leaflet edge
coaptation, usually as a result of substantial billowing,
chordal rupture or elongation, frequently resulting in MR.
Two-dimensional echocardiography has no specific criteria
for differentiating normal from pathological billowing. A
severely prolapsed or flail leaflet is relatively easy to identify
on 2D echocardiography by experts with the exception of
certain segments, and even easier with 3D echocardiography.19 RT3DE TEE has been shown to be better than 2D TEE
in identifying lesions in P1, A2, and A3 segments and
bileaflet lesions. However, the severity and magnitude of
prolapse has been difficult to accurately determine preoperatively using 2D TEE, especially in valves with multisegmental lesions.20,21 This is probably the case because 3D TEE is
less operator-dependent than 2D TEE that relies on fine
adjustments of the TEE probe and mental integration of a
limited number of 2D imaging planes to delineate MV
pathology. Analysis in a subset of our patients demonstrated
higher agreement between qualitative assessment of 3D
images and surgical findings when compared with 2D images. Additionally, the interobserver variability for the localization of involved mitral valve scallops was lower with 3D
compared with 2D. These observations further reinforce the
advantages of 3D TEE for the evaluation of complexity and
localization of lesions.
In BD, the MV has complex pathology and consists of
dilated annulus, thick leaflets, excess tissue, chordal elongation, or rupture with billowing and prolapse of leaflets that is
often multisegmental and may involve both leaflets. In
contrast, FED has less complex pathology with thin, frail
chordae, mild annular dilation, and transparent leaflets with
involvement of only 1 segment usually frequently an isolated
P2 prolapse.22
The results of this study suggest that 3D morphological
quantification from TEE images may allow successful verification of these differences in anatomy preoperatively, as
opposed to the current standard of direct intraoperative
diagnosis. In the patients categorized by surgical inspection
as BD, morphological analysis demonstrated larger annuli,
greater leaflet areas, decreased MV nonplanarity, and multisegmental prolapse with increased height and volume, consistent with the billowing nature of these leaflets. In contrast
to BD, FED patients had annular dilation and increase in
leaflet surface area of a lesser magnitude. As expected, in
FED patients, the billowing volume, while slightly increased compared with control subjects, was significantly
reduced compared with BD valves and predominantly
involved the P2 scallop. Multinomial logistic regression
analysis demonstrated billowing height to be the strongest
predictor for the presence of DMVD and billowing volume
to almost perfectly differentiate BD from FED patients. Of
note, the 3D morphological measurements were found to
be highly reproducible.
30
Circ Cardiovasc Imaging
January 2011
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
Our results suggest a systematic approach to differentiating
DMVD from normal valves and importantly BD from FED
preoperatively. In the proposed algorithm, DMVD can be
differentiated from normal valves based on presence of
abnormal billowing height, whereas billowing volume can be
used to distinguish between BD and FED patients. Use of this
quantitative algorithm makes early differentiation feasible
and the diagnosis more objective and less experiencedependent. In addition, the 3D morphological analysis provides adjunctive information on the location and severity of
extent of prolapse or billowing and degree of annular
remodeling.
One might argue that in up to 20% of the patients, a clear
distinction between BD and FED is not possible.23,24 Although there is an overlap between FED and BD in certain
parameters, the billowing volume clearly separated these 2
populations in our group of patients. Though etiologic differentiation alone may not always be sufficient in determining
complexity, precise quantification may help differentiate
surgically straightforward FED patients from those with more
complex pathology, for example, forme fruste. Most importantly, large increase in billowing volume is associated with
more complex repair as is known to be the case in patients
with BD.
Besides obviating the need for anticoagulation and decreasing the risk of endocarditis, MV repair allows preservation of left ventricular function by maintaining the integrity of
the subvalvular apparatus.25 As the procedure of choice,
surgical repair of MV prolapse requires a clear understanding
of the relationship between the etiology of DMVD, the
anatomic and functional features related to the etiology, and,
importantly, the short- and long-term impact of any alterations on the MV anatomy.26 The success and planning of
surgical correction rests on accurate MV anatomic assessment and the detection of those lesions that may predict
unsuccessful repair, such as extensive bileaflet disease or
anterior leaflet pathology.27 Our findings supported a more
accurate assessment of the complexity and localization of
segmental lesions with RT3DE imaging. Furthermore, quantification of prolapsing or billowing volume, in addition to
morphology, can distinguish between more complex lesions
versus secondary, less important lesions.
Addressing the geometric distortion evident especially in
BD patients is indispensable for a successful MV repair.
Reshaping of the mitral annulus via annuloplasty can be
achieved by measuring the length of insertion of the anterior
leaflet, the area of the anterior leaflet, or a combination of
both. Our finding of larger annuli in BD patients was
consistent with previous findings.28 Normally, the AP to CC
diameter ratio is 3:4 during systole.22 Our data demonstrated
an increase in the ratio as well as a change in annular shape
from oval to more circular. Correct sizing and customizing
the shape of the prosthetic annuloplasty ring may be essential
for the restoration of the physiological ratio and preservation
of leaflet mobility. Because the typical size of the annuloplasty ring is between 36 and 40 mm, it may be clinically
helpful to customize the rings in patients with especially large
annuli,28 as evident in some of our patients. Although the
issue of sizing of prosthetic rings remains controversial,29
accurate sizing may be essential toward preventing systolic
anterior MV motion and MR28 and thereby, avoiding resection of the anterior leaflet, which is integral to maintenance of
MV geometric relationships. Conversely, if morphological
analysis reveals a large anterior MV surface area, multisegmental or bileaflet involvement, a longer and more complicated surgical procedure might be anticipated in advance,
thereby steering the patient to centers of expertise in complex
MV repair.
The MV annulus has been shown to be nonplanar and
saddle-shaped.30 In our study, 3D morphological analysis
revealed a decrease in nonplanarity as evidenced by the
increased nonplanar angle in BD patients. Restoration of
normal geometry would probably occur with the insertion of
a customized annuloplasty ring that would reinstate normal
geometric AP relationships, elevate the posterior annulus, and
improve leaflet coaptation, all essential toward maintenance
of normal valvular physiology.
One of the consequences of inadequate MV repair is the
development of systolic anterior motion,31–34 resulting in MR.
Known risk factors contributing to recurrent MR include
bileaflet prolapse or billowing, the absence of annuloplasty
ring, chordal shortening, and limited posterior annular dilation.35 By quantifying the extent of the excess anterior and
posterior leaflets’ length, surface area, and billowing volume,
morphological analysis could help identify patient at risk for
developing systolic anterior motion. The long-term freedom
from MR and avoidance of reoperation is less achievable and
variable from center to center in cases of DMVD involving
the anterior leaflet.36,37 One of the goals of MV repair is
restoration of a large surface area of coaptation, which is
challenging when confronted with considerable billowing,
especially when the anterior leaflet is involved.38
Limitations
Measurements performed in this study are time-consuming,
rely on adequate image quality, and require a learning period
for both imaging and data analysis. Although RT3DE zoomed
acquisition can be difficult to perform in patients with an
exceedingly large annulus, only 1 of the 57 DMVD patients
who were screened was excluded from our study for this
reason. Because surgical inspection is performed in an
immobile, flaccid state, as opposed to echocardiographic
assessment, which visualizes the valve in a dynamic state, it
can potentially result in discrepancy between the methodologies,39 one might construe that using surgical findings as the
gold standard reference might be a limitation of our study.
However, this ability of morphological analysis to detect
small, not easily visible areas of billowing, whereas increasing the sensitivity of RT3DE is likely to be diagnostically
inconsequential. In addition, it is known from prior studies
that there are annular changes in geometry during systole40;
however, small multiphasic variations in annular size are not
likely to affect the diagnosis of DMVD. Nevertheless, to
address this potential pitfall, all measurements in our study
were performed at the same phase of the cardiac cycle, that is,
end-systole.
Our methodology using billowing height was able to
distinguish between the 2 forms of DMVD disease, but there
Chandra et al
Morphologic Analysis of Degenerative Mitral Valve Disease
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
may be forme fruste degenerative valves that may be incorrectly classified using our stratified approach based on cutoff
values. However, a congruent approach that incorporates all
the morphological and quantitative findings observed in our
study may still be sufficient to guide the treatment strategy
and timing of surgery. The applicability of this methodology
in patients with other MV pathologies needs to be assessed in
future studies.
In summary, our results show that 3D analysis depicts a
spectrum of pathomorphologic abnormalities that can be used
to accurate differentiate degenerative from normal valves,
and, importantly, FED from BD patients, as reflected in this
study by the close agreement between 3D TEE and the
surgical determinations. Moreover, this analysis facilitates
not only the confirmation of etiology, but highlights quantitative anatomic differences between the groups and provides
a framework for preoperative assessment of the complexity of
repair. As quantification tools become more automated and
less reliant on expertise, they may be used to support clinical
decision-making by a wider group of physicians.
Sources of Funding
This work was supported in part by National Institutes of Health
grant 2R01HL073647.
Disclosures
Drs Salgo, Cardinale, and Settlemier are employees of
Philips Healthcare.
References
1. Avierinos JF, Gersh BJ, Melton LJ III, Bailey KR, Shub C, Nishimura
RA, Tajik AJ, Enriquez-Sarano M. Natural history of asymptomatic
mitral valve prolapse in the community. Circulation. 2002;106:
1355–1361.
2. Barlow JB, Pocock WA. Billowing, floppy, prolapsed or flail mitral
valves? Am J Cardiol. 1985;55:501–502.
3. Carpentier A, Chauvaud S, Fabiani JN, Deloche A, Relland J, Lessana A,
D’Allaines C, Blondeau P, Piwnica A, Dubost C. Reconstructive surgery
of mitral valve incompetence: ten-year appraisal. J Thorac Cardiovasc
Surg. 1980;79:338 –348.
4. Ling LH, Enriquez-Sarano M, Seward JB, Orszulak TA, Schaff HV,
Bailey KR, Tajik AJ, Frye RL. Early surgery in patients with mitral
regurgitation due to flail leaflets: a long-term outcome study. Circulation.
1997;96:1819 –1825.
5. Ling LH, Enriquez-Sarano M, Seward JB, Tajik AJ, Schaff HV, Bailey
KR, Frye RL. Clinical outcome of mitral regurgitation due to flail leaflet.
N Engl J Med. 1996;335:1417–1423.
6. Bonow RO, Carabello BA, Chatterjee K, de Leon AC Jr, Faxon DP, Freed
MD, Gaasch WH, Lytle BW, Nishimura RA, O’Gara PT, O’Rourke RA,
Otto CM, Shah PM, Shanewise JS, Smith SC Jr, Jacobs AK, Adams CD,
Anderson JL, Antman EM, Fuster V, Halperin JL, Hiratzka LF, Hunt SA,
Lytle BW, Nishimura R, Page RL, Riegel B. ACC/AHA 2006 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 developed in collaboration with the Society of
Cardiovascular Anesthesiologists endorsed by the Society for Cardiovascular Angiography and Interventions and the Society of Thoracic
Surgeons. J Am Coll Cardiol. 2006;48:e1– e148.
7. Enriquez-Sarano M, Avierinos JF, Messika-Zeitoun D, Detaint D, Capps
M, Nkomo V, Scott C, Schaff HV, Tajik AJ. Quantitative determinants of
the outcome of asymptomatic mitral regurgitation. N Engl J Med. 2005;
352:875– 883.
8. Rosenhek R, Rader F, Klaar U, Gabriel H, Krejc M, Kalbeck D,
Schemper M, Maurer G, Baumgartner H. Outcome of watchful waiting in
asymptomatic severe mitral regurgitation. Circulation. 2006;113:
2238 –2244.
31
9. Adams DH, Anyanwu AC. The cardiologist’s role in increasing the rate
of mitral valve repair in degenerative disease. Curr Opin Cardiol. 2008;
23:105–110.
10. Sugeng L, Coon P, Weinert L, Jolly N, Lammertin G, Bednarz JE, Thiele
K, Lang RM. Use of real-time 3-dimensional transthoracic echocardiography in the evaluation of mitral valve disease. J Am Soc Echocardiogr.
2006;19:413– 421.
11. Sugeng L, Shernan SK, Weinert L, Shook D, Raman J, Jeevanandam V,
DuPont F, Fox J, Mor-Avi V, Lang RM. Real-time three-dimensional transesophageal echocardiography in valve disease: comparison with surgical
findings and evaluation of prosthetic valves. J Am Soc Echocardiogr. 2008;
21:1347–1354.
12. Agricola E, Oppizzi M, Pisani M, Maisano F, Margonato A. Accuracy of
real-time 3D echocardiography in the evaluation of functional anatomy of
mitral regurgitation. Int J Cardiol. 2008;127:342–349.
13. Zoghbi WA, Enriquez-Sarano M, Foster E, Grayburn PA, Kraft CD,
Levine RA, Nihoyannopoulos P, Otto CM, Quinones MA, Rakowski H,
Stewart WJ, Waggoner A, Weissman NJ. Recommendations for evaluation of the severity of native valvular regurgitation with twodimensional and Doppler echocardiography. J Am Soc Echocardiogr.
2003;16:777– 802.
14. Anyanwu AC, Adams DH. Etiologic classification of degenerative mitral
valve disease: Barlow’s disease and fibroelastic deficiency. Semin Thorac
Cardiovasc Surg. 2007;19:90 –96.
15. David TE. Outcomes of mitral valve repair for mitral regurgitation due to
degenerative disease. Semin Thorac Cardiovasc Surg. 2007;19:116 –120.
16. Gillinov AM, Cosgrove DM, Blackstone EH, Diaz R, Arnold JH, Lytle
BW, Smedira NG, Sabik JF, McCarthy PM, Loop FD. Durability of
mitral valve repair for degenerative disease. J Thorac Cardiovasc Surg.
1998;116:734 –743.
17. Lee EM, Shapiro LM, Wells FC. Superiority of mitral valve repair in
surgery for degenerative mitral regurgitation. Eur Heart J. 1997;18:
655– 663.
18. Barlow JB, Pocock WA. Mitral valve billowing and prolapse: perspective
at 25 years. Herz. 1988;13:227–234.
19. Grewal J, Mankad S, Freeman WK, Click RL, Suri RM, Abel MD, Oh
JK, Pellikka PA, Nesbitt GC, Syed I, Mulvagh SL, Miller FA. Real-time
three-dimensional transesophageal echocardiography in the intraoperative
assessment of mitral valve disease. J Am Soc Echocardiogr. 2009;22:
34 – 41.
20. Pepi M, Tamborini G, Maltagliati A, Galli CA, Sisillo E, Salvi L, Naliato
M, Porqueddu M, Parolari A, Zanobini M, Alamanni F. Head-to-head
comparison of two- and three-dimensional transthoracic and transesophageal echocardiography in the localization of mitral valve prolapse. J Am
Coll Cardiol. 2006;48:2524 –2530.
21. Delabays A, Jeanrenaud X, Chassot PG, Von Segesser LK, Kappenberger
L. Localization and quantification of mitral valve prolapse using threedimensional echocardiography. Eur J Echocardiogr. 2004;5:422– 429.
22. Filsoufi F, Carpentier A. Principles of reconstructive surgery in degenerative mitral valve disease. Semin Thorac Cardiovasc Surg. 2007;19:
103–110.
23. Fornes P, Heudes D, Fuzellier JF, Tixier D, Bruneval P, Carpentier A.
Correlation between clinical and histologic patterns of degenerative
mitral valve insufficiency: a histomorphometric study of 130 excised
segments. Cardiovasc Pathol. 1999;8:81–92.
24. Adams DH, Anyanwu AC. Seeking a higher standard for degenerative
mitral valve repair: begin with etiology. J Thorac Cardiovasc Surg.
2008;136:551–556.
25. Loop FD. Long-term results of mitral valve repair. Semin Thorac
Cardiovasc Surg. 1989;1:203–210.
26. Salgo IS, Gorman JH III, Gorman RC, Jackson BM, Bowen FW, Plappert
T, St John Sutton MG, Edmunds LH Jr. Effect of annular shape on leaflet
curvature in reducing mitral leaflet stress. Circulation. 2002;106:
711–717.
27. Omran AS, Woo A, David TE, Feindel CM, Rakowski H, Siu SC.
Intraoperative transesophageal echocardiography accurately predicts
mitral valve anatomy and suitability for repair. J Am Soc Echocardiogr.
2002;15:950 –957.
28. Adams DH, Anyanwu AC, Rahmanian PB, Abascal V, Salzberg SP,
Filsoufi F. Large annuloplasty rings facilitate mitral valve repair in
Barlow’s disease. Ann Thorac Surg. 2006;82:2096 –2100.
29. Brown ML, Schaff HV, Li Z, Suri RM, Daly RC, Orszulak TA. Results
of mitral valve annuloplasty with a standard-sized posterior band: is
measuring important? J Thorac Cardiovasc Surg. 2009;138:886 – 891.
32
Circ Cardiovasc Imaging
January 2011
30. Levine RA, Triulzi MO, Harrigan P, Weyman AE. The relationship of mitral
annular shape to the diagnosis of mitral valve prolapse. Circulation. 1987;
75:756–767.
31. Grossi EA, Galloway AC, Parish MA, Asai T, Gindea AJ, Harty S,
Kronzon I, Spencer FC, Colvin SB. Experience with twenty-eight cases of
systolic anterior motion after mitral valve reconstruction by the Carpentier technique. J Thorac Cardiovasc Surg. 1992;103:466 – 470.
32. Jebara VA, Mihaileanu S, Acar C, Brizard C, Grare P, Latremouille C,
Chauvaud S, Fabiani JN, Deloche A, Carpentier A. Left ventricular
outflow tract obstruction after mitral valve repair: results of the sliding
leaflet technique. Circulation. 1993;88:II30 –II34.
33. Lee KS, Stewart WJ, Lever HM, Underwood PL, Cosgrove DM.
Mechanism of outflow tract obstruction causing failed mitral valve repair:
anterior displacement of leaflet coaptation. Circulation. 1993;88:
II24 –II29.
34. Rey MJ, Mercier LA, Castonguay Y. Echocardiographic diagnosis of left
ventricular outflow tract obstruction after mitral valvuloplasty with a
flexible Duran ring. J Am Soc Echocardiogr. 1992;5:89 –92.
35. Flameng W, Herijgers P, Bogaerts K. Recurrence of mitral valve regurgitation after mitral valve repair in degenerative valve disease. Circulation.
2003;107:1609–1613.
36. David TE, Ivanov J, Armstrong S, Christie D, Rakowski H. A comparison
of outcomes of mitral valve repair for degenerative disease with posterior,
anterior, and bileaflet prolapse. J Thorac Cardiovasc Surg. 2005;130:
1242–1249.
37. Mohty D, Orszulak TA, Schaff HV, Avierinos JF, Tajik JA, EnriquezSarano M. Very long-term survival and durability of mitral valve repair
for mitral valve prolapse. Circulation. 2001;104:I1–I7.
38. Carpentier A. Cardiac valve surgery: the “French correction.” J Thorac
Cardiovasc Surg. 1983;86:323–337.
39. Patel V, Hsiung MC, Nanda NC, Miller AP, Fang L, Yelamanchili P,
Mehmood F, Gupta M, Duncan K, Singh A, Rajdev S, Fan P, Naftel DC,
McGiffin DC, Pacifico AD, Kirklin JK, Lin CC, Yin WH, Young MS,
Chang CY, Wei J. Usefulness of live/real time three-dimensional transthoracic echocardiography in the identification of individual segment/
scallop prolapse of the mitral valve. Echocardiography. 2006;23:
513–518.
40. Pai RG, Tanimoto M, Jintapakorn W, Azevedo J, Pandian NG, Shah PM.
Volume-rendered three-dimensional dynamic anatomy of the mitral
annulus using a transesophageal echocardiographic technique. J Heart
Valve Dis. 1995;4:623– 627.
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
CLINICAL PERSPECTIVE
Degenerative mitral valve disease (DMVD) frequently includes different degrees of annular dilation, leaflet redundancy,
and chordal dysfunction, which result in variable cardiovascular morbidity and mortality. DMVD encompasses 2 broad
categories, fibroelastic deficiency and Barlow disease. Accurate diagnosis of these entities along with their specific location
and complexity is important because they require different surgical planning, which necessitates careful matching between
the complexity of reparability and surgical expertise. Differential diagnosis in DMVD is challenging because it relies on
qualitative evaluation that requires a high level of expertise. Quantitative decision support would be useful in assessing key
anatomic features of DMVD and therefore guide appropriate surgical strategy. This is the first comprehensive study to
characterize DMVD objectively using real-time 3D echocardiography. We sought to describe the pathomorphology of the
MV in DMVD and to identify real-time 3D echocardiography parameters that can accurately characterize DMVD. Our
results have yielded a systematic approach to differentiating DMVD from normal valves and importantly Barlow disease
from fibroelastic deficiency before surgery. In the proposed algorithm, DMVD can be differentiated from normal valves
based on billowing height, whereas billowing volume can be used to distinguish between Barlow disease and fibroelastic
deficiency patients. Use of this quantitative algorithm makes early differentiation feasible and the diagnosis more objective
and less experience-dependent, highlights quantitative anatomic differences between the groups, and provides a framework
for preoperative assessment of the complexity of repair. As quantification tools become more automated and less reliant
on expertise, they may be used to support clinical decision-making by a wider group of physicians.
Downloaded from http://circimaging.ahajournals.org/ by guest on November 19, 2016
Characterization of Degenerative Mitral Valve Disease Using Morphologic Analysis of
Real-Time Three-Dimensional Echocardiographic Images: Objective Insight Into
Complexity and Planning of Mitral Valve Repair
Sonal Chandra, Ivan S. Salgo, Lissa Sugeng, Lynn Weinert, Wendy Tsang, Masaaki Takeuchi,
Kirk T. Spencer, Anne O'Connor, Michael Cardinale, Scott Settlemier, Victor Mor-Avi and
Roberto M. Lang
Circ Cardiovasc Imaging. 2011;4:24-32; originally published online September 30, 2010;
doi: 10.1161/CIRCIMAGING.109.924332
Circulation: Cardiovascular Imaging is published by the American Heart Association, 7272 Greenville Avenue,
Dallas, TX 75231
Copyright © 2010 American Heart Association, Inc. All rights reserved.
Print ISSN: 1941-9651. Online ISSN: 1942-0080
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circimaging.ahajournals.org/content/4/1/24
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Circulation: Cardiovascular Imaging can be obtained via RightsLink, a service of the Copyright Clearance
Center, not the Editorial Office. Once the online version of the published article for which permission is being
requested is located, click Request Permissions in the middle column of the Web page under Services. Further
information about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation: Cardiovascular Imaging is online at:
http://circimaging.ahajournals.org//subscriptions/