multimodality imaging in the context of transcatheter ... · multimodality imaging in the context...

J A C C : C A R D I O V A S C U L A R I M A G I N G V O L . 8 , N O . 1 0 , 2 0 1 5

ª 2 0 1 5 B Y T H E AM E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N DA T I O N I S S N 1 9 3 6 - 8 7 8 X / $ 3 6 . 0 0

P U B L I S H E D B Y E L S E V I E R I N C . h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j c m g . 2 0 1 5 . 0 8 . 0 0 4

STATE-OF-THE-ART PAPERS

Multimodality Imaging in the Context ofTranscatheter Mitral Valve Replacement
Establishing Consensus Among Modalities and Disciplines
Philipp Blanke, MD,* Christopher Naoum, MBBS,* John Webb, MD,* Danny Dvir, MD,* Rebecca T. Hahn, MD,yPaul Grayburn, MD,z Robert R. Moss, MBBS,* Mark Reisman, MD,x Nicolo Piazza, MD,k Jonathon Leipsic, MD*

JACC: CARDIOVASCULAR IMAGING CME

CME Editor: Ragavendra R. Baliga, MD

This article has been selected as this issue’s CME activity, available online

at http://www.acc.org/jacc-journals-cme by selecting the CME tab on the

top navigation bar.

Accreditation and Designation Statement

The American College of Cardiology Foundation (ACCF) is accredited by

the Accreditation Council for Continuing Medical Education (ACCME) to

provide continuing medical education for physicians.

The ACCF designates this Journal-based CME activity for a maximum

of 1 AMA PRA Category 1 Credit(s) �. Physicians should only claim

credit commensurate with the extent of their participation in the

activity.

Method of Participation and Receipt of CME Certificate

To obtain credit for this CME activity, you must:

1. Be an ACC member or JACC: Cardiovascular Imaging subscriber.

2. Carefully read the CME-designated article available online and in this

issue of the journal.

3. Answer the post-test questions. At least 2 out of the 3 questions

provided must be answered correctly to obtain CME credit.

4. Complete a brief evaluation.

5. Claim your CME credit and receive your certificate electronically by

following the instructions given at the conclusion of the activity.

CME Objective for This Article: After reading this article the reader should

be able to: 1) review the anatomy of the mitral valvular apparatus and its

appearance on noninvasive imaging; 2) delineate the imaging

From the *St. Paul’s Hospital, University of British Columbia, Vancouver, Brit

Center and Cardiovascular Research Foundation, New York, New York; zBaDepartment of Internal Medicine, the Heart Hospital Baylor Plano, Pla

Washington Medical Center, Seattle, Washington; and the kDepartment of

Health Centre, Montreal, Quebec, Canada. Dr. Blanke is a consultant for N

Dr. Webb is a consultant for Edwards Lifesciences. Dr. Hahn has a fina

Neovasc. Dr. Grayburn has received research grants from Abbott Vascular, Me

methodology for the evaluation of the mitral annulus in the context of

transcatheter mitral valvular interventions; 3) provide standard nomen-

clature for the mitral annulus and apparatus on noninvasive imaging

with both echocardiography and MDCT and to introduce the concept

of the D-shaped mitral annulus; 4) provide insight into the salient

measurements to determine patient and device suitability; and 5) high-

light ways in which MDCT and 3-dimensional TEE can help assist both

planning and performance of TMVI.

CME Editor Disclosure: JACC: Cardiovascular Imaging CME Editor

Ragavendra R. Baliga, MD, has reported that he has no relationships to

disclose.

Author Disclosures: Dr. Blanke is a consultant for Neovasc, Edwards

Lifesciences, and Circle Imaging. Dr. Webb is a consultant for Edwards

Lifesciences. Dr. Hahn has a financial relationship with Edwards Life-

sciences and Neovasc. Dr. Grayburn has received research grants from

Abbott Vascular, Medtronic, Edwards Lifesciences, and Boston Scientific;

is a consultant for Abbott Vascular and Tendyne; and has financial re-

lationships with Echo Core Lab, Valtech Cardio, and Tendyne. Dr. Piazza

has financial relationships with Medtronic and HighLife Medical; and is

a HighLife shareholder. Dr. Leipsic is a consultant for Edwards Life-

sciences; and provides institutional core lab services for Edwards Life-

sciences, Neovasc, and Tendyne. All other authors have reported that they

have no relationships relevant to the contents of this paper to disclose.

Medium of Participation: Print (article only); online (article and quiz).

CME Term of Approval

Issue Date: October 2015

Expiration Date: September 30, 2016

ish Columbia, Canada; yColumbia University Medical

ylor Heart and Vascular Institute, Dallas, Texas and

no, Texas; xDivision of Cardiology, University of

Medicine, Division of Cardiology, McGill University

eovasc, Edwards Lifesciences, and Circle Imaging.

ncial relationship with Edwards Lifesciences and

dtronic, Edwards Lifesciences, and Boston Scientific;

http://www.acc.org/jacc-journals-cme

http://crossmark.crossref.org/dialog/?doi=10.1016/j.jcmg.2015.08.004&domain=pdf

http://dx.doi.org/10.1016/j.jcmg.2015.08.004

Blanke et al. J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 8 , N O . 1 0 , 2 0 1 5

Imaging for TMVI O C T O B E R 2 0 1 5 : 1 1 9 1 – 2 0 8

1192

Multimodality Imaging in th
e Context ofTranscatheter Mitral Valve Replacement
Establishing Consensus Among Modalities and Disciplines

ABSTRACT

is

dy

co

All

Ma

Transcatheter mitral valve implantation (TMVI) represents a promising approach to treating mitral valve regurgi-

tation in patients at increased risk of perioperative mortality. Similar to transcatheter aortic valve replacement

(TAVR), TMVI relies on pre- and periprocedural noninvasive imaging. Although these imaging modalities, namely

echocardiography, computed tomography, and fluoroscopy, are well established in TAVR, TMVI has entirely different

requirements. Approaches and nomenclature need to be standardized given the multiple disciplines involved. Herein

we provide an overview of anatomical principles and definitions, a methodology for anatomical quantification, and

perioperative guidance. (J Am Coll Cardiol Img 2015;8:1191–208) © 2015 by the American College of Cardiology

Foundation.

S urgical mitral valve repair or replacement arestandard therapies for mitral regurgitation(MR), however, many elderly patients with clin-

ically significantMR are deemed too high risk for surgi-cal intervention (1) and are often precluded fromsurgical treatment (2,3). Percutaneous repair strate-gies are under investigation as alternative options inhigh-risk populations (4,5). More recently, transcath-eter mitral valve implantation (TMVI) has been pro-posed in a fashion analogous to transcatheter aorticvalve replacement (TAVR) with promising earlyresults (6–9). Building on experience with TAVR,advanced imaging has undergone much earlier inte-gration in TMVI (10,11). Given the complex structureand function of the mitral valve (MV), with itsnonplanar annulus; the lack of a circular, fibrousannular structure; the variability of leaflet and sub-valvular apparatus anatomy and its proximity to theleft ventricular outflow tract (LVOT), the role of imag-ing is possibly even more pertinent (12).

We highlight the important anatomical andnoninvasive imaging issues that are essential forsuccessfully performing TMVI and emphasize thecommon nomenclature for use among various imag-ing modalities and specialties (Tables 1 and 2, CentralIllustration).

a consultant for Abbott Vascular and Tendyne; and has financial relation

ne. Dr. Piazza has financial relationships with Medtronic and HighLife Me

nsultant for Edwards Lifesciences; and provides institutional core lab servi

other authors have reported that they have no relationships relevant to

nuscript received May 26, 2015; revised manuscript received August 10,

OVERVIEW OF CURRENTLY AVAILABLE

TMVI DEVICES

TMVI currently in human trials target treatment ofboth primary MR (i.e., degenerative mitral valve dis-ease [DMVD]) (6,9) and secondary MR (i.e., functionalmitral regurgitation [FMR]) (6–8) with some devicesintended to treat both disease entities (6,9).

An initial attempt at transvenous transseptal im-plantation proved difficult and has not yet beenattempted again in humans (12). Subsequently,transapical access has been successful and is beingpursued, with all valves currently in human trials(6–9), although a transseptal approach is regarded asthe major aim of efforts in innovation.

All devices under investigation incorporate certaincommon features: a transapical access system, anitinol self-expanding frame, bovine pericardial tri-leaflet valve, and a fabric or pericardial sealing cuff(12). Some devices are circular (8,9), whereas othersare D-shaped (6,7), in which the latter has the flatportion of the frame oriented toward the anteriormitral valve leaflet (AML) in order to reduce the po-tential for LVOT obstruction. The bioprostheticvalvular apparatus consists of 3 symmetrical leaflets(6,8,9), with the exception of 1 device, which has

ships with Echo Core Lab, Valtech Cardio, and Ten-

dical; and is a HighLife shareholder. Dr. Leipsic is a

ces for Edwards Lifesciences, Neovasc, and Tendyne.

the contents of this paper to disclose.

2015, accepted August 13, 2015.

AB BR E V I A T I O N S

AND ACRONYM S

AML = anterior mitral valve

leaflet

CT = computed tomography

FMR = functional mitral

regurgitation

LV = left ventricle

LVOT = left ventricular

outflow tract

MA = mitral annulus

MR = mitral regurgitation

PML = posterior mitral leaflet

TEE = transesophageal

echocardiography

TTE = transthoracic

echocardiography

TAVR = transcatheter aortic

valve replacement

TMVI = transcatheter mitral

valve implantation

J A C C : C A R D I O V A S C U L A R I M A G I N G , V O L . 8 , N O . 1 0 , 2 0 1 5 Blanke et al.O C T O B E R 2 0 1 5 : 1 1 9 1 – 2 0 8 Imaging for TMVI

1193

asymmetrical leaflets with a posterior leaflet that isslightly larger than its 2 anterior leaflets (7).

Most devices incorporate a wide atrial skirt orflange, which is in direct apposition to the left atrialwall surrounding the mitral inflow to assist withdevice capture and sealing (7–9). Importantly, noneof the devices relies primarily on radial force forfixation. Fixation is facilitated either by tabs whichengage with the basal myocardium and fibrousskeleton (7), opposing, circumferential anchors toengage with the annulus and leaflets (9), paddles forleaflet capturing at the A2/P2 scallops (8), or anapical tether (6). The anchoring mechanisms are themost distinguishing characteristic among the devices(Figures 1 and 2) and largely determine anatomicalrequirements such as landing-zone characteristics,which may differ significantly between mitral pa-thologies and patients.

MITRAL APPARATUS

The MV complex is composed of the annulus, theanterior and posterior leaflets, chordae tendinae, andpapillary muscles.

MITRAL ANNULUS. Historically, the mitral annulus(MA) is defined by the junction of the left atrium, leftventricle (LV), and mitral leaflets, resulting in a 3-dimensional (3D) saddle-shaped configuration withanterior and posterior peaks (13), the former beingcontinuous with the aortovalvular complex, the latterformed by the insertion of the posterior mitral leaflet(PML), and the nadirs, which are located close to thefibrous trigones.

TABLE 1 Anatomical Nomenclature and Definitions for Annular Quan

Anatomical Entity Alternative Descriptor

Posterior MA PML insertion Ins

Anterior MA Anterior peak, aortic peak Am

Intervalvular fibrosa Aortomitral curtain (surgical),aortomitral continuity,aortomitral junction, fibrouscontinuity

Fib

Left trigone Lateral trigone An

Right trigone Medial trigone An

Trigone-to-trigone distance (TT) Intertrigonal distance Dis

Septal-to-lateral distance (SL) A2-to-P2 distance, anterior-posterior distance

Dis

Commissure-to-commissuraldistance (IC)

Intercommissural distance,medial-to-lateral distance

Dis

Mitral annular trajectory Mitral annular axis Tra

Neo-LVOT Fo

MA ¼ mitral annulus; PML ¼ posterior mitral leaflet; LVOT ¼ left ventricular outflow tr

The junction of the left atrium (LA), LV,and PML insertion typically forms a well-defined, distinct fibrous structure (14–16). Incontrast, the anterior annulus is more diffi-cult to define, having various perspectivesamong specialties and imaging modalities(17–19), primarily due to the continuoustransition of the AML into the intervalvularfibrosa, also referred to as the “aortomitralcurtain” or “continuity.” Surgeons tend toexclude the intervalvular fibrosa from theirMA definition (18) as they can visuallyidentify the distal margin of left atrialmyocardium along the aortomitral curtainintraoperatively. However, the intervalvularfibrosa is often included in cardiac imaging,likely due to the lack of a distinct border onboth computed tomography (CT) and echo-cardiography (Figures 3 and 4). Importantly,the anterior MA does not correspond to thehinge point of the AML, as the latter is located
further toward the ventricle, usually below the fibroustrigones.
MV LEAFLETS. The two MV leaflets are referred toas “anterior” and “posterior.” However, due to theoblique orientation of the mitral apparatus, relativeto the anatomical axes, the leaflets are oriented in amore anterosuperior and posteroinferior position(20). The leaflets are asymmetrical in shape, with theAML being rounded and occupying a third of theannular circumference, whereas the radially narrowerPML occupies the other two-thirds. The coaptation

tification

Definition

ertion of the PML, usually at the junction of the left atrium and leftventricle

biguous definitions, e.g., identical with the insertion of the noncoronaryand left coronary cusp; or excluding the intervalvular fibrosa

rous tissue between both trigones; toward the ventricle transitioninginto the AML; on the atrial side in parts covered by left atrialmyocardium; scaffold for the insertion of parts of the noncoronary andleft coronary cusp

chor for intervalvular fibrosa

chor for intervalvular fibrosa, continuous with semimembranous septum

tance between the left (lateral) and right (medial) trigones

tance between the annulus at A2 to the annulus at P2; for the D-shapedconcept between the trigone-to-trigone distance and P2 in aperpendicular fashion

tance between annulus at P1 and P3 along the bi-commissural view; with3D automated segmentation in a perpendicular fashion to the septal-to-lateral distance through the centroid

jectory perpendicular to the annular plane

rmed by the deflected AML and basal septum after TMVI

act; TMVI ¼ transcatheter mitral valve implantation.

TABLE 2 Role and Contribution of Imaging Modalities in the Context of TMVI

Plan TTE2D

TEE/X-Plane 3D* 3D TEE CT Fluoroscopy

Pre-procedural planning

Quantification of MR þþþ þþ þþþ NA þAnnular dimensions þ þ þþ þþþ NA

Leaflet morphology þþ þþþ þþþ þþ NA

Annular and leafletcalcifications

þþ þþ þ þþþ þ

Chordae þþ þþ þþ þ NA

Papillary muscle anatomy þþ þþ þþ þþþ NA

LV Size and function þþþ þþ NA þþ þþLVOT anatomy þ þþ þþþ þþþ NA

Periprocedural imaging

Localization of ventricularpuncture

NA þþþ* þ NA þ

Guidewire advancementand positioning

NA þþ* þþþ† NA þþ

delivery system advancementand positioning

NA þþþ* þþþ NA þþ

Device deployment NA þþþ* þþþ NA þþRotational alignment NA þ§ þþþ†‡ NA þDevice anchoring NA þþþ* þþ NA þ

Post-TMVR

Valvular competency/para-valvularregurgitation

þþ þþ þþþk þ þ

Trans-mitral gradient þþþ þþþ NA NA NA

LVOT anatomy þþ þþ þþþ þþþ NA

LVOT gradient þþþ þþ NA NA þþþ#

Device apposition/seating þþ þþ þþþ þþþ NA

Device stability þþþ þþþ þþ þþ þþþLeaflet mobility/thrombus þ þþþ þþ þþþ NA

Stent fracture NA NA NA þ þþþ

*X-plane mode. †Live 3D mode. ‡Zoom 3D mode. §Transgastric view. kColor 3D and vena contracta area.#Catheter-based direct gradient measurement.

LV ¼ left ventricle; LVOT ¼ left ventricular outflow tract; MR ¼ mitral regurgitation; NA ¼ not applicable;TMVI ¼ transcatheter mitral valve implantation.



1194

line approximates a semilunar arc, with each endreferred to as a “commissure.” Importantly, theanterolateral and posteromedial commissures do notextend to the annulus, often lacking a distinct sepa-ration of both leaflets. The PML is indented by foldsor clefts, creating 3 frequently unequal scallops, withthe middle scallop being typically the largest. Car-pentier’s nomenclature (21) describes the mostlateral, anterosuperior segment as P1, the centralsegment as P2, and the most medial, posteroinferiorsegment as P3. Although the indentations can be wellvisualized on echocardiography, they are not asdiscrete on CT, owing to partial volume averaging.Thus, on CT the PML can be routinely subdivided into3 equal partitions, P1, P2, and P3. The AML is curtain-like and lacks distinct scallops, although similarlabeling (A1, A2, and A3) is applied to the lateral,middle, and medial segments, respectively.

ECHOCARDIOGRAPHIC IMAGE ACQUISITION. Alth-ough echocardiography is the primary imaging mo-dality for diagnosing and quantifying MR, this paperfocuses on pre-procedural anatomical assessment,intraoperative guidance, and post-operative assess-ment (Table 2).

The mitral apparatus can be imaged on bothtransesophageal echocardiography (TEE) and trans-thoracic echocardiography (TTE). On TTE, the A2-P2scallops are typically imaged from either parasternalor apical long-axis views. Inferomedial and antero-lateral tilting of the image plane may allow imaging ofthe A1-P1 or A3-P3 coaptation. The apical 4-chamberview illustrates the A3-A2 scallops medially and theP2-P1 scallops laterally, whereas the commissuralview images P3-A2-P1. The basal short axis view ofthe MV shows all 6 mitral valve scallops and bothcommissures, with A1-P1 scallops typically located tothe right of the image. In addition, short-axis viewsmay provide the best imaging planes for identifyingthe location, number, and orientation of the papillarymuscles and attached chordae.

Current TEE guidelines recommend assessmentusing standardized 2D views of the mitral valve(4-chamber, 3-chamber, commissural) from a mid-esophageal position and transgastric views as well as3D imaging planes (22). Generally, imaging protocolsshould be adapted to the individual patient’s heartby using rotational angles that optimize the recordingof important structural and flow information.

3D echocardiography (23) has a variety of 3Dacquisition modes and display options (simul-taneous multiplane imaging, tomographic slices,surface rendering, and volume rendering). In addi-tion, it allows for simultaneous multiplane imaging(“x-plane mode”) with 2D planes in a modifiableangulation to each other. Multiplanar reconstructionof the 3D volume allows for anatomical measure-ments. Importantly, real-time volume rendering per-mits construction of an en face atrial view known asthe surgical view. The ventricular perspective can beimportant in assessing subvalvular structures andLVOT evaluation. 3D TEE usually produces imageswith spatial resolution that is superior to that of3D TTE. For TMVI, 3D TEE is most useful foranatomical assessment, quantification, and intra-procedural guidance, in particular by providing a“surgical” view.

3D TEE images can be acquired using differentacquisition volumes as well as either single-beat ormultibeat datasets (23). Multibeat acquisitionsimprove temporal and spatial resolution but may belimited by “stitch artifacts” due to irregular rhythm orrespiratory variations. Multicycle acquisitions are

CENTRAL ILLUSTRATION Multimodality Imaging for TMVI: Pre-Procedural Screening, Periprocedural Guidance andPost-Procedural Assessment

Overview of the contribution of computed tomography (CT), transesophageal echocardiography (TEE), transthoracic echocardiography (TTE), and fluoroscopy in the

context of transcatheter mitral valve implantation (TMVI). AML ¼ anterior mitral valve leaflet; IC ¼ intercommissural; LVOT ¼ left ventricular outflow tract; PML ¼posterior mitral leaflet; SL ¼ septal-to-lateral; TT ¼ trigone-to-trigone.


1195

FIGURE 1 Schematic of Various TMVI Anchoring Mechanisms

(A) Baseline anatomy. (B) Anchoring with tabs, for example, posteriorly (inferolateral) at a

myocardial shelf and anteriorly at the fibrous trigones. (C) Anchoring by grasping the AML

and PML with paddles. (D) Barbs. (E) Anchoring to the apical myocardium with a tether

(neochord). AML ¼ anterior mitral valve leaflet; LA ¼ left atrium; LV ¼ left ventricle;

PML ¼ posterior mitral leaflet; TMVI ¼ transcatheter mitral valve implantation.

FIGURE 2 Transca

(A) Tiara (courtesy o

FORTIS Transcathete

(C) Tendyne transcat

ings, Roseville, Minn

Mitral Valve Implant



1196

usually needed to present 3D color images withadequate temporal resolution.

The narrow-angle “live” 3D mode allows for “real-time” 3D images, using a single-beat mode anda matrix array transducer display of a relativelynarrow-angle pyramidal volume, which is usuallyinsufficient to visualize the entire mitral apparatus.

theter Mitral Valves Currently in Human Trials

f Neovasc Inc., Richmond, British Columbia, Canada). (B) Edwards

r Mitral Valve (courtesy of Edwards Lifesciences, Irvine, California).

heter mitral valve implantation system (courtesy of Tendyne Hold-

esota). (D) CardiAQ Valve Technologies System for Transcatheter

ation (courtesy of CardiAQ Valve Technologies, Irvine, California).

This is most useful for focusing on a specificabnormality in the mitral valve once orientation isclear. It is extensively used live to guide devicedeployment.

The wide-angle 3D mode with a focused widesector (zoom 3D) permits a wide sector view of the MVapparatus from the annulus to the papillary tips. Thebroader sector mode is associated with inferior spatialand temporal resolution compared to the “live”mode.

The full volume mode has the largest acquisitionsector, which is ideal for imaging the entire mitralapparatus together with the LV. This mode also hasoptimal spatial and temporal resolution, permittingdetailed diagnosis of complex pathologies but isideally used with multiple cycle acquisitions.

CT DATA ACQUISITION AND

RECONSTRUCTION

Given dynamic changes in anatomical configura-tion of the mitral apparatus, LVOT, and LV,contrast-enhanced CT data acquisition for TMVIshould preferably image the entire cardiac cycle(24), for example, by means of retrospectivelyelectrocardiography-gated data acquisition, or byprospective electrocardiography-triggered data ac-quisition with “whole-heart” detector coverage (25).Image acquisition can be limited to the heart; how-ever, imaging of the entire thoracic cage may bebeneficial to determine the intercostal space fortransapical access. Except for annular measure-ments, the mitral apparatus is best evaluated bystandard views mimicking those obtained usingechocardiography (commissural, 3-chamber, 4-chamber,and short-axis views) (Figure 3).

2D ANNULAR ASSESSMENT

On 2D echocardiography, the MA was traditionallyassessed on a “view-based” approach, using 2- and4-chamber views (26), aiming to align the mitralapparatus with the cardiac chambers while uninten-tionally disregarding the noncircular MA geometry.Correct anatomic imaging planes using a truecommissural view showing the P1-A2-P3 scallops anda perpendicular on-axis long-axis view depicting theA2-P2 scallops have been proposed for more adequateassessment (17,27), and can also be mimicked on CT.Although the commissural and long-axis views pro-vide the major and minor MA diameter, respectively,2D measurements overall incompletely describe thecomplex 3D MA geometry, and yielded values arestrongly dependent on their exact orientation (17,27).

FIGURE 3 2D Multiplanar Cardiac CT Views

Short-axis view of the MA region (A) and schematic of the leaflet scallops (B) with dashed

lines indicating the orientation of the views in D to F. (D) Commissural view transecting at

P1-P3 (major MA diameter). (E) Long-axis view transecting through A2-P2, oriented

perpendicularly to the commissural view. The long-axis view is lacking a distinct MA

landmark at A2, resulting in variable measurements (C, inset). (F) Four-chamber view with

a diagonal orientation (not recommended for 2D measurements). 2D ¼ two-dimensional;

CT ¼ computed tomography; LAA ¼ left atrial appendage; LAX ¼ long-axis; MA ¼ mitral

annulus.


1197

For 2D TEE, the variability of imaging planes forany individual is significant, and even with probemanipulation, optimal visualization of the annularplane cannot be guaranteed. Importantly, there iscontroversy regarding the definition of the anteriorMA (Figure 4). Some define it as identical to theinsertion of the non- and left-coronary aortic cuspsincluding the intervalvular fibrosa, whereas othersdeliberately exclude the latter, with the lack of adistinct border to the AML creating further confusion.Overall, these limitations challenge the applicabilityof 2D echocardiographic and CT measurements forTMVI planning (22,27–30).

3D ANNULAR SEGMENTATION

3D segmentation on 3D echocardiography and CTovercomes these aforementioned limitations. 3D MAsegmentation was first described by Levine et al.(13,31) using 2D echocardiography and has been reit-erated with modern imaging technologies (32,33).With contemporary approaches, segmentation isperformed by generating a cubic spline interpola-tion of manually placed seeding points along the3D annular contour. However, until recently, MAsegmentation has largely been of academic interest,with surgical sizing performed at the time of theprocedure in a nonstandardized fashion (18).

Increasing interest in minimally invasive mitralprocedures has triggered development of post-processing solutions for 3D annular segmentation forboth echocardiography and CT, permitting assess-ment of area, perimeter, and other salient measure-ments. The so-called method of “least squares planes”provides the mathematical foundation for deriving 2Dmeasurements from the 3D contour by defining a 2Dplane and the geometrical center, referred to as thecentroid (34,35). Simplified, this can be illustrated byprojecting the 3D contour onto a 2D plane (Figure 5).This method also permits definition of an axis thatis oriented perpendicularly to the 2D plane whiletransecting the centroid.

We propose that the 2D plane derived from the3D contour is referred to as the “MA plane” and thatthe aforementioned axis is referred to as the “MAtrajectory.”

As previously discussed, it has been well estab-lished that the MA is nonplanar (13). However, forTMVI planning, truncation of the saddle-shapedannular contour at a virtual line connecting bothtrigones, referred to as the trigone-to-trigone (TT)distance, has been proposed (10), based on theobservation that the anterior horn of the saddle-shaped contour would otherwise project into the

LVOT (Figure 6). The resulting D-shaped MA is moreplanar and does not project onto the LVOT.

On cardiac CT, 3D segmentation is performed byplacing seeding points for the cubic spline along theinsertion of the PML, using long- and short-axisreformats aligned with the LV long axis (Figure 6).The anterior horn is segmented along the insertion ofthe noncoronary and right coronary aortic cusp intothe intervalvular fibrosa. After identifying the trig-ones, the anterior horn is truncated along the TTdistance to form the D shape (36). Post-processingyields annular area and perimeter, the latter consist-ing of the posterior annulus (PML insertion) and theTT distance. Furthermore, annular geometry is char-acterized by measurement of the septal-to-lateral(SL) distance (A2-to-P2 distance, minor diameter),and the intercommissural (IC) distance (major diam-eter). For standardization and because of the arc-shaped coaptation zone, we recommend that the ICdistance be assessed in a parallel fashion to the TTdistance, while transecting through the centroid ofthe D-shaped annulus, usually yielding the largest

FIGURE 5 3D MA and 2D Projected Contour

The 2D MA area is assessed using the method of least squares,

similar to projecting the contour onto a plane. Orientation of the

plane and MA trajectory are obtained by the least squares plane

calculation. The actual 2D MA plane has the identical orientation

as the projection but transects through the geometrical centroid.

FIGURE 4 2D Cardiac Views and MA Measurements in TEE

Commissural view transecting the annulus at P1-P3, typically

yielding the major MA diameter (A). Perpendicular long-axis view

transecting at A2-P2 (B), yielding the minor MA diameter. In

contrast to P1-P3, there is no distinct landmark at A2 defining the

anterior annulus. TEE ¼ transesophageal echocardiography;

other abbreviations as in Figure 3.



1198

long-axis dimension. Similarly, the SL distanceshould be assessed perpendicularly to TT and IC dis-tances while transecting through the centroid(Figure 6E).

Similar post-processing can be performed off-line,using echocardiographic 3D volume sets (Figure 7).For both CT and echocardiography, multiphasicmeasurements can be performed to characterize thedynamics of the MV complex.

FURTHER ANATOMICAL QUANTIFICATION

Leaflet anatomy and length are important for devicesanchoring to the leaflets themselves (8). Leafletlength can be assessed on a 3-chamber view on bothechocardiography and CT by using mid-diastoliccaliper measurements. Adequate distance of thepapillary muscle tip to the leaflet is important for

devices that anchor behind the leaflets or at thetrigones to ensure sufficient space for the tabs topass (7). Furthermore, anatomy should be assessedfor the presence of direct papillary muscle insertioninto the leaflets, which may interfere with anchoring(Figure 8).

ANNULAR AND LANDING ZONE GEOMETRY

Normative data for mitral annular dimensions varywidely, mainly because of differences in imagingmodalities and segmentation techniques, particularlywith regard to the anterior horn. Early 2D TTE studiesreported relatively smaller MA areas in normalstructures (37) than those in more recent studies us-ing 3D echocardiographic techniques, which reportmean values ranging between 8.4 cm2 and 11.8 cm2

(34,38,39), with comparable values reported in car-diac CT studies among control cohorts (24,39–42). In arecent investigation using the aforementioned defi-nition of the D-shaped annulus, mean MA area was8.9 � 1.5 cm2 in control subjects without significantcardiovascular disease, with broad interindividualvariation noted and larger dimensions observed inmales due primarily to differences in body size(39,43). Mean MA dimensions are increased in TMVIpatients, with mitral valve prolapse being associatedwith larger mean dimensions than FMR (40,41,43,44).In FMR, the saddle height decreases, resulting in amore planar saddle-shaped annular contour (34);however, this does not affect the already more planarD-shaped annular segmentation (10). In regard to in-plane geometry, relatively greater enlargement ofthe SL distance than the IC distance is observed

FIGURE 6 3D-Mitral Annular Segmentation on CT

(A) Saddle-shaped MA segmentation as a cubic spline interpolation. (B) Pink line ¼anterior peak; red line ¼ posterior peak (PML insertion); green and blue dots ¼ fibrous

trigones. Importantly, the anterior peak projects into the LVOT (short-axis view [C] and

long-axis view [D]). The more planar D-shaped annular contour is created by truncating

the saddle-shaped contour at the trigone-to-trigone distance (yellow lines [E and F]).

Important measurements are the projected area septal-to-lateral (SL) and inter-

commissural (IC) distances; the latter is oriented perpendicularly to SL while transecting

through the centroid (F). Abbreviations as in Figures 1 and 3.

FIGURE 7 3D Mitral Annular Segmentation on 3D TEE

Multiplanar reformatting of the 3D TEE dataset with annular contour and segmented

coaptation line. (A) Commissural view. (B) Long-axis view perpendicular to A at A2-P2

(septal-to-lateral). (C) Short-axis view. (D) Surgical view.


1199

in both FMR and mitral valve prolapse patients(40,41,43).

Importantly, landing zone anatomy varies betweenFMR and mitral valve prolapse. In FMR, regional wallmotion abnormalities and/or LV dilation leads to se-vere tethering of mitral leaflets and annular dilation,resulting in not only increased tenting height andreduced coaptation length (43) but also remodeling ofthe basal myocardium with formation of a “myocar-dial shelf,” which can be identified on both echocar-diography and CT (Figures 9C and 9F). In DMVD withfibroelastic deficiency characterized by single-scallopprolapse with other scallops/leaflets often normal orthin and diffuse myxomatous degeneration, withgeneralized valvular thickening, redundant leaflets,and chordal elongations, the insertion of the mitralvalve leaflet may be displaced into the LA (Figures 9Band 9E), referred to as mitral annular disjunction(14,45). In DMVD, a posterior “myocardial shelf” istypically not present, and the basal myocardium maybulge into the lumen with hyperdynamic and hyper-trophied LV anatomy.

MITRAL ANNULAR DYNAMICS

Although minimal MA dimensions are present in earlysystole, MA dimensions increase toward late systole(38,44,46). Importantly, annular dynamics differ be-tween normal subjects and patients with mitral valvedisease. In FMR, the extent of dynamic changes isgenerally diminished (24,46,47), whereas more pro-nounced changes have been described in DMVD(44,46), with loss of systolic area contraction as wellas significant increase in annular area from early tolate systole. However, recent evidence suggests thatrelevant differences exist between the fibroelasticdeficiency and diffuse myxomatous degenerationphenotypes, with abnormal dynamics observed onlyin diffuse myxomatous degeneration (48). The po-tential for changes in MA dimensions emphasizes theimportance of multiphasic annular measurements forTMVI planning.

ANNULAR CALCIFICATION

Mitral annular calcification (MAC) is a commondegenerative process of the fibrous MA associatedwith advancing age (49,50) and end-stage renal dis-ease (51) and is present in approximately 6% of thegeneral population (52). Due to the high prevalence ofboth MR and MAC, these entities may coincidewithout necessarily being causally related to eachother. The extent of MAC can vary from mild andspotty involvement to severe, calcific encasement of

FIGURE 8 Assessment of Papillary Muscle Anatomy

(A) Distance measurement from the anteromedial papillary

muscle tip to the annular plane (yellow line). The red line

indicates the annular trajectory. (B and C) Multiplanar reformat

and endovascular volume rendered image demonstrating direct

insertion of the anteromedial papillary muscle into the AML

(yellow arrow in B, black arrow in C). Abbreviations as in

Figure 1.



1200

the entire annulus, although it is frequently limited tothe posterior part. A rare variant of MAC is caseousannular calcification, which forms bulky, space-occupying lesions predominantly along the posteriorannulus (53). Compared to typical MAC, caseous MACis less echodense and may exhibit homogenousareas of attenuation similar to the contrast-enhancedblood on contrast-enhanced CT but can be welldistinguished on non–contrast-enhanced CT (54). Therelevance of MAC is currently unknown, although it isa contraindication to TMVI if it is severe in most ofthe current feasibility studies.

DETERMINATION OF TMVI FEASIBILITY WITH

MULTIMODALITY IMAGING

The relevant anatomy for determination of TMVIfeasibility depends on which device is being

implanted, given the interdevice variations inanchoring mechanisms. TMVI devices require propersizing of the MA as well as a detailed characterizationof the landing zone. Because of dynamic changes,annular dimensions should be assessed at multiplephases throughout the cardiac cycle. Excessive MACor subvalvular calcification could interfere withproper seating and apposition.

Furthermore, anatomical suitability of the an-choring mechanism should be assessed. For devicesanchoring onto the leaflets with paddles (8), forexample, at A2-P2, ensuring sufficient leaflet lengthand ruling out MV prolapse and annular disjunctionat P2 are required. Importantly, papillary muscleand chordae anatomy should be assessed for thepresence of false bands and direct insertion ofpapillary muscles. For devices anchoring with tabs inthe inferolateral basal myocardium (7), the presenceof a “myocardial shelf” that persists throughout thecardiac cycle should be ensured. Hypertrophy of thebasal myocardium with a myocardial bulge or severe,bulky annular calcification may interfere with thisparticular anchoring mechanism. Devices anchoringby an apical tether are not affected by leafletlength or pathology (6). For all TMVI devices, thebasal LV cavity must be able to accommodate thedevice, and small LV cavities with hyperdynamicfunction may not necessarily allow room for theTMVI cage.

PREDICTING LVOT OBSTRUCTION

TMVI devices consist of circumferentially coveredstent struts (7,8,55) that may significantly protrudeinto the LV cavity, interact with the AML, andpotentially encroach upon the LVOT. Because of thisprotrusion, a neo-LVOT is created by the device, theAML, and the interventricular septum. Theoretically,LVOT obstruction can occur due to narrowing of thenative LVOT above the level of the TT line or to for-mation of a narrow neo-LVOT below the level of theTT line toward the LV.

Predisposing factors for LVOT obstruction includeanatomical and device-related factors. LVOT anat-omy exhibits significant interindividual variabilityand is influenced mainly by configuration of theinterventricular septum, LV size, and aortomitralangulation. In particular, a hypertrophied, bulgingseptum reduces the LVOT and neo-LVOT cross-sectional areas.

AORTOMITRAL ANGULATION. Aortomitral angula-tion is the angle between the MA trajectory and theLVOT long axis. Theoretically, a parallel orientationof the MA trajectory and the LVOT long axis would

FIGURE 9 MA Landing Zone in Normal, DMVD, and FMR

(A to C) TEE and (D to F) corresponding multiplanar CT images. Normal 4-chamber-view

annular anatomy is shown (A and D). In mitral valve prolapse MA disjunction (MAd) may be

observed (B andE) in 4-chamber views. In FMRwithLVdilation amyocardial shelf is observed

along the posterior annulus (commissural views [C and F]). DMVD ¼ degenerative mitral

valve disease; FMR¼ functional mitral regurgitation; other abbreviations as in Figures 1 and

3.

FIGURE 10 Prediction of Neo-LVOT Dimensions

End-systolic CT-datasets in FMRwith an anterolateral/lateral myocardial scar (A, C, E) and in

DMVD (B, D, E). (A andB) Three-chamber views and commisural views (C andD) showing the

annular segementation and a simulated cylindrical device (29mm), oriented perpendicularly

to the annular plane. The neo-LVOT formed by the septal myocardium and the device is

segmented (center line technique, orange line). The red bar indicates the position of the

short-axis LVOT view (E and F), which allows for planimetric assessment of the neo-LVOT,

yielding 3.5 cm2 at end-systole (E), indicating low risk for LVOT obstruction, and a slit-like

neo-LVOT (F) suggests high risk for LVOT obstruction. LVOT¼ left ventricular outflow tract.


1201

result in minimal risk of LVOT obstruction, whereas aperpendicular orientation would result in maximalrisk. Aortomitral angulation must be viewed in thecontext of LVOT and LV dimensions and septalanatomy rather than as a single, stand-alone factor.Furthermore, quantification may be limited by sub-jective assessment of the LVOT long axis.

LV SIZE. A small LV cavity is a known risk factorfor LVOT obstruction after surgical valve replace-ment, with a larger LV cavity likely to be able toaccommodate a device without compromising theLVOT. However, the specific LV anatomy needsto be considered. In FMR, the LV may be glo-bally dilated or focally scarred with compensatoryhypertrophy of other myocardial segments. Con-versely DMVD patients with hyperdynamic systolicLV function may exhibit small end-systolic LVdiameters (56).

INTERVENTRICULAR SEPTUM. Similarly, basal septalbulging or basal hypertrophy (>15 mm) may signif-icantly decrease neo-LVOT dimensions and is aknown risk factor for systolic anterior motion ofthe AML in surgical MV repair. Septal hypertrophyis often associated with increased aortomitralangulation, further increasing the LVOT obstructionrisk.

DEVICE-RELATED FACTORS AND SIMULATION OF

DEVICE IMPLANTATION. Device protrusion into theLV and device flaring are determinants of neo-LVOTgeometry, with higher risk associated with moreprotrusion and device flaring.

CT TMVI simulation may predict neo-LVOT geom-etry by embedding a cylindrical or device-specificcontour into the CT dataset, followed by segmenta-tion and planimetrical assessment of the neo-LVOTcross-sectional area (Figure 10). This virtual assess-ment is somewhat limited, however, as there are noestablished cutoff values for minimal neo-LVOT areathat indicate an increased risk of LVOT obstruction.First, studies have been limited to patients with hy-pertrophic obstructive cardiomyopathy, in whom therisk of developing a gradient of >50 mm Hg has beenshown to correspond with LVOT area cutoffs rangingfrom 0.85 to 2.0 cm2 (57,58). Second, the appropriatecardiac phase for assessment is unknown. Neo-LVOTdimensions appear worse at end-systole, but theventricular stroke volume is already ejected atthis time point, suggesting greater importance ofassessment in early or mid-systole. Third, the AMLmay contribute to LVOT obstruction due to displace-ment by the device and systolic anterior motion.Although redundant AML tissue may be noticed on

FIGURE 11 Transapical Access Point

The annular trajectory (light blue) is oriented perpendicularly to the annular plane, which

is discrepant to the line connecting the centroid and true LV apex (purple line). The ideal

access point can be assessed on volume-rendered images. (A) Long-axis view. (B) Volume-

rendered image in true anterior-posterior orientation. (C) Schematic view. (D) Volume-

rendered image viewed from LAO 45� and CRA 25�. CRA ¼ cranial; D2 ¼ second diagonal

branch; LAD ¼ left anterior descending artery; LAO ¼ left anterior oblique.

FIGURE 12 Assess

The annular trajecto

can be assessed in re

Volume-rendered im

the relationship to th

surement between t



1202

echocardiography and CT, device simulation does notfactor in AML behavior post-implantation. Fourth, LVand neo-LVOT dimensions will significantly dependon the hemodynamic state and can be altered byexertion and LV remodeling.

ment of Transapical Approach

ry (light blue) and connection line to the true LV apex (purple line)

lation to the ribs, intercostal spaces, sternum, and midline. (A)

age shows trajectories and rib cage. (B) Surface rendering illustrates

e sternum and midline. (C) Axial view with curved distance mea-

he access point and midline.

DETERMINING ACCESS LOCATION

The MA trajectory oriented perpendicularly to the MAplane represents the ideal delivery path to facilitatecoaxial device deployment. By extending the MAtrajectory beyond the epicardium on post-processing,the ideal LV access point can be identified, whichis commonly located laterally or anteriorly to thetrue apex (Figure 11). Once identified, CT simulationshows the trajectory’s relationship to the coronaryarteries, papillary muscles, or myocardial scars andprovides the distance to the MA plane for intra-procedural guidance. Further extension of the tra-jectory beyond the body surface can help determinethe ideal intercostal space and distance from thesternal midline (Figure 12). By creating an en faceview, the orientation of the MA trajectory, commonlypointing to the left anterior oblique and caudally, canbe provided to facilitate alignment of the deliverysystem with the trajectory for coaxial deployment(Figure 13C).

PRE-OPERATIVE PREDICTIONS OF

FLUOROSCOPIC ANGULATIONS

Device advancement, unsheathing, and release aremonitored by fluoroscopy (Figure 14). Similar to TAVR,coplanar fluoroscopic projections facilitate coaxialdevice deployment. MA plane segmentation on CT canprovide projection angulations yielding an optimalviewing curve, displaying the corresponding cranial-caudal angulation for a given LAO/RAO angulation(Figure 13). Secondary to relatively vertical position ofthe MA, these curves exhibit a steep slope with pro-nounced changes in cranial/caudal angulation forsmall changes in RAO angulation. In addition tobeing orthogonal to the MA, TMVI requires C-armprojections aligned with specific anatomical andanchoring structures. Given the asymmetrical MA,2 views are intuitive: the septal-to-lateral viewparallel to the SL line (A2-P2 view) and the TT viewparallel to the TT line. However, projection angula-tions are limited by physical restraints of the C-armand required procedural access. Angulations requiredto achieve the SL view are generally in the rangeof practical C-arm working angles, whereas angula-tions for the TT view are not (11). Alternatively, acompromise view between the TT view and SLview has been shown to be effective during devicedeployment (11).

The TMVI procedure is performed without angi-ography, and the noncalcified annulus lacks a fluo-roscopically identifiable anatomical landmark. Anindirect landmark can be created by using a coronary

FIGURE 13 Prediction of Fluoroscopy Angulation

(A) Short-axis view. (B) Optimal viewing curve. (C) En face view. (D and G) TT view. (E and H) Compromise view. (F and I) SL and A2-P2 views.

SL ¼ septal-to-lateral; TT ¼ trigone-to-trigone.

FIGURE 14 Intraoperative Fluoroscopy Guidance

Intraoperative fluoroscopy shows partial deployment of a Tiara

device (Neovasc Inc.) with coplanar depiction of the transapical

delivery system, evident by the linear appearance of the circular

marker. A transjugular coronary sinus wire serves as an

anatomical landmark.


1203

sinus guidewire (11). Using a transjugular approach,the guide wire usually follows the floor of the coro-nary sinus at P1 to then align with the outer curvatureat P2. Given the significant interindividual variability,with the sinus often located distant to the annularplane along the left atrium (59,60), this individualpath can be predicted pre-procedurally by using theCT dataset (Figure 15), allowing for guidewire simu-lation in relation to the annular plane in any fluoro-scopic view (11).

GENERAL PRINCIPLES OF

DEVICE IMPLANTATION AND

INTRAPROCEDURAL IMAGING

Intraprocedural imaging is performed with TEE(Online Videos 1 and 2 [videos are courtesyof Tendyne Holdings, Roseville, Minnesota]) andfluoroscopy (Table 2). The location of the ventricularpuncture can be confirmed on TEE by “poking” theLV with a finger at the intended cannulationsite (Figure 16), ideally using standardized views,

http://jaccimage.acc.org/video/2015/vol8/issue10/0578_VID1-vol8iss10.wmv


FIGURE 16 Localiz

(A) Simultaneous mu

intended cannulation

FIGURE 15 Coronary Sinus Segmentation

Spline segmentation along the outer curvature beginning at P2. (A and B) Short- and long-

axis views, respectively, and subsequent angiographic simulation (C), illustrating the

relationship between the potential coronary sinus wire and the annular plane.



1204

comparable to the previous CT simulation, aiming atfollowing the MA trajectory but staying away from thepapillary muscles, septum, and right ventricularapex. Continuous imaging of the guidewire is per-formed to determine the correct placement across the

ing the Apical Cannulation Site

ltiplane image showing the surgeon’s finger (red star) poking the

site. (B) Initial wire path (yellow dotted line).

MA and positioning in the right pulmonary vein, us-ing fluoroscopy and TEE. To ensure that the wire doesnot pass through the chordae, an inflated ballooncatheter may be advanced into the left atrium andpulled back under TEE, fluoroscopic, and tactile sur-veillance. At this stage, C-arm angulation alreadyprovides either a SL/A2-P2 or a compromise viewalong the optimal viewing curve, as proposedby prior CT-analysis, allowing for a coplanar depic-tion of the delivery system if aligned with the MAtrajectory. The delivery system is introduced underboth fluoroscopic and TEE guidance, again confirmingfree passage from the apex to the left atrium andexcluding entanglement of the device in the sub-valvular apparatus by moving the device in the MVorifice.

Depending on device design, device unsheathingand unfolding begins either above or at the annularlevel. Centering of the delivery system in the mitralorifice at A2-P2 is guided by TEE, using either a 3Den face surgical view or a multiplane 2D view, suchas simultaneous long-axis and commissural views,or deep gastric short-axis views, especially if 3Dviews are suboptimal. The x-plane function is mostuseful for centering and determining the degree ofadvancement with respect to the annulus, whereasthe en face 3D view is helpful for judging rotationalalignment, although in practice, both functions areoften interchanged rapidly. Unfolding of the atrialflange/skirt is monitored on both fluoroscopy andechocardiography. Continuous monitoring of theorientation with appropriate rotational adjust-ment can be performed throughout the deploymentprocess, ensuring alignment of the flat portion ofD-shaped devices or specific anchoring mechanismwith the mitral apparatus.

A coronary sinus wire may aid estimation of thedevice’s position in relation to the annular plane onfluoroscopy by mentally integrating the distance ofthe coronary sinus to the MA plane from prior CTanalysis. Furthermore, fluoroscopy can show changesin the atrial skirt configuration, when the partiallyunfolded device is lowered toward the annular plane,supported further by tactile feedback. Atrial skirtapposition to the atrial wall is documented on TEE,typically using the x-plane mode with the long-axisview as the primary view and the commissural viewas the secondary view. These views and the ability tochange the orientation of the secondary view mayallow rapid imaging of anchoring mechanisms prior torelease of the device. Devices anchoring to the MVleaflets with paddles (8) require synchronous captureof the AML and PML at A2 and P2 before the mainbody is unsheathed. Correct paddle orientation and

FIGURE 17 Periprocedural TEE Imaging During TMVI With the Fortis Device

The treatment of severe mitral regurgitation (A) with TMVI (B-G). Anchoring paddles are initially positioned outside the leaflets (C), aligned at

A2-P2 using a short-axis gastric view (D). Leaflets are captured between the paddles and valve body, and the atrial flange is released (E),

followed by deployment of the valve and sealing of the atrial skirt (F and G), leading to resolution of mitral regurgitation (H). Abbreviations as

in Figures 1 and 4.


1205

centered position at A2-P2 must be confirmed onx-plane or 2D transgastric short-axis views. Appro-priate leaflet capture and paddle insertion areconfirmed on the long-axis view. Finally, furtherunsheathing of the main body is monitored on bothTEE and fluoroscopy (intraprocedural examples areshown in Figures 17 and 18).

POST-TMVI ECHOCARDIOGRAPHIC

ASSESSMENT

Immediately following deployment, 2D and 3D im-aging confirm appropriate seating, stability, radialorientation, relationship to the captured leaflets, andprosthetic valve function. Comprehensive 2D and 3Dassessments of the LVOT are performed by using co-lor, pulsed and continuous wave Doppler to excludepotential LVOT obstruction. The transgastric windowcan be used to measure LVOT velocities.

Color Doppler is used to assess central or para-valvular mitral regurgitation (Figure 19). Qualitative

and semiquantitative methods of assessing para-valvular mitral regurgitation have been reportedpreviously (61). Assessment may be complicatedby device-related acoustic shadowing and irregularor atypical regurgitant jets. Here, TEE is likelyto be more sensitive than TTE. However, adjudicationof paravalvular mitral regurgitation severity maybe difficult due to the variability and complexityof orifice geometry and absence of a true gold stan-dard. Therefore, incorporation of other methods,such as pulmonary venous flow pattern (“systolicblunting/reversal”) and LVOT-to-transmitral velocitytime integrals ratio (as a surrogate for mitral regur-gitant volume) may be helpful (61). Although data arepresently lacking, assessment of paravalvular mitralregurgitation severity by 3D vena contracta areaseems likely to have an increasing role.

Mitral valve orifice area can be quantified by directplanimetry or by using Doppler and the continuityequation (usually by TTE), which is preferred in theabsence of significant mitral regurgitation. Although

FIGURE 18 Deployment of Tendyne Valve Using 3D Zoom Surgical Views

(A) Sheath (arrow) is seen in the LA above native leaflets. (B) Valve flange (arrow) is

released and begins to appear in LA. (C) Valve flange is rotated, aligning the flat part of the

D-shaped mitral annulus with the aortic-mitral curtain (arrows). (D) Flange is fully opened,

and the bioprosthesis is seen in the center. Online Videos 1 and 2. Abbreviations as in

Figure 1.

FIGURE 19 TEE Images Immediately After Implantation of Tendyne Valve

X-plane view shows mid-commissural (A) and long-axis (B) views. Valve leaflets are in

closed position (mid-systole). Color Doppler images in same views showing LVOT pres-

ervation and no paravalvular leakage (C and D). Abbreviations as in Figures 4 and 10.



1206

a significantly prolonged pressure half-time mayindicate valve stenosis, this method has limitations,given the potential impact of variable LV and LAcompliance on the pressure decay slope and shouldnot be used to report valve areas (61).

FOLLOW-UP ECHOCARDIOGRAPHIC

IMAGING

TTE is convenient for evaluation of TMVI devicesover time. Apical views allow assessment of LVvolumes, LVEF, and global longitudinal strain andstrain rate to assess reverse remodeling and im-provement of LV systolic function after TMVI. Theyalso allow accurate assessment of mitral valve gra-dients and calculation of the LVOT/mitral inflowVTI ratio. In contrast, TTE imaging of the left atriummay be challenging due to acoustic shadowing,making it difficult to evaluate changes in LA volumesor paravalvular leaks, especially from the apicalwindows.

FUTURE DIRECTIONS

Pre-procedural imaging aids primarily in deter-mining patient eligibility and device sizing. Inparticular, prediction of LVOT obstruction mayrequire image analysis beyond merely device simu-lation, especially in borderline situations. Here,computational fluid dynamics may provide moreinsight but must also integrate dynamic changes ofLVOT anatomy.

Similar to that from TAVR, information derivedfrom CT is currently mentally integrated into theprocedure, in particular for optimizing fluoroscopicprojections and access of the delivery system.Although attempts have been made to fuse CT dataand fluoroscopic images, major limitations willconsist of discrepant patient positioning andanatomical distortion during apical access. Instead,fusion of TEE and fluoroscopy appears moreappealing and is of greater relevance, in particularin regard to device positioning within the mitralapparatus.

CONCLUSIONS

TMVI is an evolving treatment strategy for patientswith mitral regurgitation, requiring elaborate pre-procedural and periprocedural imaging. CT and 3Dechocardiography aid in determining patient suit-ability by 3D anatomical quantification, landing zonecharacterization, and TMVI simulation to identifypatients at increased risk for LVOT obstruction.




1207

Additionally, CT provides fluoroscopy angulation andthe most suitable ventricular access site to allow forcoaxial deployment. Both 2D and 3D TEE guide theprocedure in conjunction with fluoroscopy. Involve-ment of multiple imaging modalities and specialtiesrequires common nomenclature and understandingof the anatomy involved.

REPRINT REQUESTS AND CORRESPONDENCE: Dr.Jonathon Leipsic, Department of Radiology, Univer-sity of British Columbia, Centre for Heart ValveInnovation-St. Paul’s Hospital, 1081 Burrard Street,Vancouver V6Z 1Y6, Canada. E-mail: [email protected].

RE F E RENCE S

1. Iung B, Baron G, Butchart EG, et al.A prospective survey of patients with valvularheart disease in Europe: The Euro Heart Survey onValvular Heart Disease. Eur Heart J 2003;24:1231–43.

2. Enriquez-Sarano M, Akins CW, Vahanian A.Mitral regurgitation. Lancet 2009;373:1382–94.

3. Mehta RH, Eagle KA, Coombs LP, et al. Influ-ence of age on outcomes in patients undergoingmitral valve replacement. Ann Thorac Surg 2002;74:1459–67.

4. Feldman T, Young A. Percutaneous approachesto valve repair for mitral regurgitation. J Am CollCardiol 2014;63:2057–68.

5. Herrmann HC, Maisano F. Transcatheter therapyof mitral regurgitation. Circulation 2014;130:1712–22.

6. Moat N, Duncan A, Lindsay A, et al. Trans-catheter mitral valve replacement for the treat-ment of mitral regurgitation: in-hospital outcomesof an apically tethered device. J Am Coll Cardiol2015;65:2352–3.

7. Cheung A, Webb J, Verheye S, et al. Short-termresults of transapical transcatheter mitral valveimplantation for mitral regurgitation. J Am CollCardiol 2014;64:1814–9.

8. Bapat V, Buellesfeld L, Peterson MD, et al.Transcatheter mitral valve implantation (TMVI)using the Edwards FORTIS device. EuroInter-vention 2014;10 Suppl U:U120–8.

9. Sondergaard L, Brooks M, Ihlemann N, et al.Transcatheter mitral valve implantation via trans-apical approach: an early experience. Eur J Car-diothorac Surg 2015. Feb 3 [E-pub ahead of print].

10. Blanke P, Dvir D, Cheung A, et al. A simplifiedD-shaped model of the mitral annulus to facilitateCT-based sizing before transcatheter mitral valveimplantation. J Cardiovasc Comput Tomogr 2014;8:459–67.

11. Blanke P, Dvir D, Naoum C, et al. Prediction offluoroscopic angulation and coronary sinus loca-tion by CT in the context of transcatheter mitralvalve implantation. J Cardiovasc Comput Tomogr2015;9:183–92.

12. De Backer O, Piazza N, Banai S, et al. Percu-taneous transcatheter mitral valve replacement:an overview of devices in preclinical and earlyclinical evaluation. Circulation CardiovascularInterv 2014;7:400–9.

13. Levine RA, Triulzi MO, Harrigan P, Weyman AE.The relationship of mitral annular shape to thediagnosis of mitral valve prolapse. Circulation1987;75:756–67.

14. Hutchins GM, Moore GW, Skoog DK. The as-sociation of floppy mitral valve with disjunction ofthe mitral annulus fibrosus. N Engl J Med 1986;314:535–40.

15. Berdajs D, Zund G, Camenisch C, Schurr U,Turina MI, Genoni M. Annulus fibrosus of themitral valve: reality or myth. J Card Surg 2007;22:406–9.

16. Angelini A, Ho SY, Anderson RH, Davies MJ,Becker AE. A histological study of the atrioven-tricular junction in hearts with normal and pro-lapsed leaflets of the mitral valve. Br Heart J1988;59:712–6.

17. Foster GP, Dunn AK, Abraham S, Ahmadi N,Sarraf G. Accurate measurement of mitral annulardimensions by echocardiography: importance ofcorrectly aligned imaging planes and anatomiclandmarks. J Am Soc Echocardiogr 2009;22:458–63.

18. Bothe W, Miller DC, Doenst T. Sizing for mitralannuloplasty: where does science stop andvoodoo begin? Ann Thorac Surg 2013;95:1475–83.

19. Anderson RH, Kanani M. Mitral valve repair:critical analysis of the anatomy discussed. Mul-timed Man Cardiothorac Surg 2007;2007:mmcts2006.002147.

20. Theriault-Lauzier P, Andalib A, Martucci G,et al. Fluoroscopic anatomy of left-sided heartstructures for transcatheter interventions: insightfrom multislice computed tomography. J Am CollCardiol Intv 2014;7:947–57.

21. Carpentier AF, Lessana A, Relland JY, et al. The“physio-ring”: an advanced concept in mitral valveannuloplasty. Ann Thorac Surg 1995;60:1177–85;discussion 1185–6.

22. Hahn RT, Abraham T, Adams MS, et al.Guidelines for performing a comprehensive tran-sesophageal echocardiographic examination: rec-ommendations from the American Society ofEchocardiography. J Am Soc Echocardiogr 2013;26:921–64.

23. Lang RM, Badano LP, Tsang W, et al. EAE/ASErecommendations for image acquisition anddisplay using three-dimensional echocardiogra-phy. Eur Heart J Cardiovasc Img 2012;13:1–46.

24. Alkadhi H, Desbiolles L, Stolzmann P, et al.Mitral annular shape, size, and motion in normalsand in patients with cardiomyopathy: evaluationwith computed tomography. Invest Radiol 2009;44:218–25.

25. Bolen MA, Popovic ZB, Dahiya A, et al. Pro-spective ECG-triggered, axial 4-D imaging of theaortic root, valvular, and left ventricular struc-tures: a lower radiation dose option for

preprocedural TAVR imaging. J Cardiovasc ComputTomogr 2012;6:393–8.

26. Quinones MA, Otto CM, Stoddard M, et al.Recommendations for quantification of Dopplerechocardiography: a report from the DopplerQuantification Task Force of the Nomenclatureand Standards Committee of the American Societyof Echocardiography. J Am Soc Echocardiogr2002;15:167–84.

27. Hyodo E, Iwata S, Tugcu A, et al. Accuratemeasurement of mitral annular area by using singleand biplane linear measurements: comparison ofconventional methods with the three-dimensionalplanimetric method. European heart journal car-diovascular Imaging 2012;13:605–11.

28. Grewal J, Mankad S, Freeman WK, et al. Real-time three-dimensional transesophageal echocar-diography in the intraoperative assessment ofmitral valve disease. J Am Soc Echocardiogr 2009;22:34–41.

29. Fattouch K, Castrovinci S, Murana G, et al.Multiplane two-dimensional versus real timethree-dimensional transesophageal echocardiog-raphy in ischemic mitral regurgitation. Echocardi-ography 2011;28:1125–32.

30. Ben Zekry S, Nagueh SF, Little SH, et al.Comparative accuracy of two- and three-dimensional transthoracic and transesophagealechocardiography in identifying mitral valve pa-thology in patients undergoing mitral valve repair:initial observations. J Am Soc Echocardiogr 2011;24:1079–85.

31. Levine RA, Handschumacher MD, Sanfilippo AJ,et al. Three-dimensional echocardiographic re-construction of the mitral valve, with implicationsfor the diagnosis of mitral valve prolapse. Circu-lation 1989;80:589–98.

32. Lee AP, Hsiung MC, Salgo IS, et al. Quantita-tive analysis of mitral valve morphology in mitralvalve prolapse with real-time 3-dimensionalechocardiography: importance of annular saddleshape in the pathogenesis of mitral regurgitation.Circulation 2013;127:832–41.

33. Jassar AS, Vergnat M, Jackson BM, et al.Regional annular geometry in patients withmitral regurgitation: implications for annulo-plasty ring selection. Ann Thorac Surg 2014;97:64–70.

34. Flachskampf FA, Chandra S, Gaddipatti A, et al.Analysis of shape and motion of the mitralannulus in subjects with and without cardiomyop-athy by echocardiographic 3-dimensional re-construction. J Am Soc Echocardiogr 2000;13:277–87.

mailto:[email protected]

mailto:[email protected]

http://refhub.elsevier.com/S1936-878X(15)00590-2/sref1































































































































































1208

35. Legget ME, Bashein G, McDonald JA, et al.Three-dimensional measurement of the mitralannulus by multiplane transesophageal echo-cardiography: in vitro validation and in vivodemonstration. J Am Soc Echocardiogr 1998;11:188–200.

36. Blanke P, Dvir D, Cheung A, et al. Mitralannular evaluation with CT in the context oftranscatheter mitral valve replacement. J Am CollCardiol Img 2015;8:612–5.

37. Ormiston JA, Shah PM, Tei C, Wong M. Sizeand motion of the mitral valve annulus in man. I. Atwo-dimensional echocardiographic method andfindings in normal subjects. Circulation 1981;64:113–20.

38. Mihaila S, Muraru D, Piasentini E, et al.Quantitative analysis of mitral annular geometryand function in healthy volunteers using trans-thoracic three-dimensional echocardiography.J Am Soc Echocardiogr 2014;27:846–57.

39. Sonne C, Sugeng L, Watanabe N, et al. Age andbody surface area dependency of mitral valve andpapillary apparatus parameters: assessment byreal-time three-dimensional echocardiography.Eur J Echocardiogr 2009;10:287–94.

40. Delgado V, Tops LF, Schuijf JD, et al. Assess-ment of mitral valve anatomy and geometry withmultislice computed tomography. J Am Coll Car-diol Img 2009;2:556–65.

41. Beaudoin J, Thai WE, Wai B,Handschumacher MD, Levine RA, Truong QA.Assessment of mitral valve adaptation with gatedcardiac computed tomography: validation withthree-dimensional echocardiography and mecha-nistic insight to functional mitral regurgitation.Circ Cardiovasc Imaging 2013;6:784–9.

42. Gordic S, Nguyen-Kim TD, Manka R, et al.Sizing the mitral annulus in healthy subjects andpatients with mitral regurgitation: 2D versus 3Dmeasurements from cardiac CT. Int J CardiovascImg 2014;30:389–98.

43. Naoum C, Leipsic J, Cheung A, et al. Mitralannular dimensions and geometry in normalsubjects and patients with mitral regurgitation:implications for transcatheter mitral valve im-plantation. J Am Coll Cardiol Img 2015. In press.

44. Grewal J, Suri R, Mankad S, et al. Mitralannular dynamics in myxomatous valve disease:new insights with real-time 3-dimensional echo-cardiography. Circulation 2010;121:1423–31.

45. Carmo P, Andrade MJ, Aguiar C, Rodrigues R,Gouveia R, Silva JA. Mitral annular disjunction inmyxomatous mitral valve disease: a relevant ab-normality recognizable by transthoracic echocar-diography. Cardiovasc ultrasound 2010;8:53.

46. Levack MM, Jassar AS, Shang EK, et al. Three-dimensional echocardiographic analysis of mitralannular dynamics: implication for annuloplastyselection. Circulation 2012;126:S183–8.

47. Kaplan SR, Bashein G, Sheehan FH, et al.Three-dimensional echocardiographic assessmentof annular shape changes in the normal andregurgitant mitral valve. Am Heart J 2000;139:378–87.

48. Clavel MA, Mantovani F, Malouf J, et al. Dy-namic phenotypes of degenerative myxomatousmitral valve disease: quantitative 3-dimensionalechocardiographic study. Circ Cardiovasc Img2015;8:e002989.

49. Savage DD, Garrison RJ, Castelli WP, et al.Prevalence of submitral (anular) calcium and itscorrelates in a general population-based sample(the Framingham Study). Am J Cardiol 1983;51:1375–8.

50. Fox CS, Vasan RS, Parise H, et al. Mitralannular calcification predicts cardiovascularmorbidity and mortality: the Framingham HeartStudy. Circulation 2003;107:1492–6.

51. Maher ER, Young G, Smyth-Walsh B, Pugh S,Curtis JR. Aortic and mitral valve calcification inpatients with end-stage renal disease. Lancet1987;2:875–7.

52. Movahed MR, Saito Y, Ahmadi-Kashani M,Ebrahimi R. Mitral annulus calcification is associ-ated with valvular and cardiac structural abnor-malities. Cardiovasc Ultrasound 2007;5:14.

53. Deluca G, Correale M, Ieva R, Del Salvatore B,Gramenzi S, Di Biase M. The incidence and clinicalcourse of caseous calcification of the mitralannulus: a prospective echocardiographic study.J Am Soc Echocardiogr 2008;21:828–33.

54. Plank F, Al-Hassan D, Nguyen G, et al. Caseouscalcification of the mitral annulus. CardiovascDiagn Ther 2013;3:E1–3.

55. Lutter G, Lozonschi L, Ebner A, et al. First-in-human off-pump transcatheter mitral valvereplacement. J Am Coll Cardiol Intv 2014;7:1077–8.

56. Scampardonis G, Yang SS, Maranhao V,Goldberg H, Gooch AS. Left ventricular

abnormalities in prolapsed mitral leaflet syn-drome. Review of eighty-seven cases. Circulation1973;48:287–97.

57. Fukuda S, Lever HM, Stewart WJ, et al. Diag-nostic value of left ventricular outflow area inpatients with hypertrophic cardiomyopathy: areal-time three-dimensional echocardiographicstudy. J Am Soc Echocardiogr 2008;21:789–95.

58. Qin JX, Shiota T, Lever HM, et al. Impact of leftventricular outflow tract area on systolic outflowvelocity in hypertrophic cardiomyopathy: a real-time three-dimensional echocardiographic study.J Am Coll Cardiol 2002;39:308–14.

59. del Valle-Fernandez R, Jelnin V,Panagopoulos G, Ruiz CE. Insight into the dy-namics of the coronary sinus/great cardiac veinand the mitral annulus: implications for percuta-neous mitral annuloplasty techniques. Circ Car-diovasc Interv 2009;2:557–64.

60. Tops LF, Van de Veire NR, Schuijf JD, et al.Noninvasive evaluation of coronary sinus anatomyand its relation to the mitral valve annulus: im-plications for percutaneous mitral annuloplasty.Circulation 2007;115:1426–32.

61. Zoghbi WA, Chambers JB, Dumesnil JG, et al.Recommendations for evaluation of prostheticvalves with echocardiography and Doppler ultra-sound: a report From the American Society ofEchocardiography’s Guidelines and StandardsCommittee and the Task Force on ProstheticValves. J Am Soc Echocardiogr 2009;22:975–1014.

KEY WORDS anterior mitral valve leaflet,functional mitral regurgitation, leftventricular outflow tract, mitralregurgitation, TMVI, TMVR, transesophagealechocardiography, transcatheter mitral valvereplacement, transcatheter mitral valveimplantation

APPENDIX For sequential video clipsshowing advancement of delivery system,release of atrial skirt, rotational alignment, andcomplete deployment of a Tendyne valve,please see the online version of this paper.

Go to http://www.acc.org/jacc-journals-cme to take the CMEquiz for this article.






































































































































multimodality imaging in the context of transcatheter ... · multimodality imaging in the context...

Documents