- Email: [email protected]
Predicting LVOT Obstruction in Transcatheter Mitral Valve Implantation
JACC: CARDIOVASCULAR IMAGING
VOL.
ª 2016 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER
-, NO. -, 2016
ISSN 1936-878X/$36.00 http://dx.doi.org/10.1016/j.jcmg.2016.01.005
IMAGING VIGNETTE
Predicting LVOT Obstruction in Transcatheter Mitral Valve Implantation Concept of the Neo-LVOT Philipp Blanke, MD,a Christopher Naoum, MBBS,a Danny Dvir, MD,a Vinayak Bapat, MD,b Kevin Ong, MD,a David Muller, MBBS,c Anson Cheung, MD,a Jian Ye, MD,a James K. Min, MD,d Nicolo Piazza, MD,e Pascal Theriault-Lauzier, BSC,e John Webb, MD,a Jonathon Leipsic, MDa OUTFLOW TRACT OBSTRUCTION IS A FEARED AND POTENTIALLY LETHAL COMPLICATION OF TRANSCATHETER
mitral valve replacement (TMVR), mitral valve-in-valve (ViV), and valve-in-ring (ViR) procedures as well as implantation of transcatheter heart valves in calcific mitral valve disease. These procedures ultimately lead to elongation of the outflow tract into the left ventricle, whereas the pre-existing “native” left ventricular outflow tract (LVOT) confined by the most basal septum and the intervalvular fibrosa remains unchanged (Figure 1). The newly created elongation of the outflow tract is confined posteriorly by the deflected anterior mitral valve leaflet in TMVR (Figures 1 and 2), ViR, and calcific mitral valve procedures or by the deflected bioprosthetic leaflets and transcatheter heart valve struts in ViV procedures. To better discriminate from the native LVOT, we propose to refer to the newly created elongation as the “neo-LVOT” (1). Although risk factors for narrow neo-LVOT dimensions can be both device- and anatomy-related (Figure 3), we propose to perform a patientspecific assessment by means of computed tomography (CT)-based, virtual, stereolithographic device implantation. This requires segmentation of the native mitral annulus (2), the bioprosthetic sewing ring, or the annuloplasty ring, with subsequent definition of the annular plane and the annular trajectory by means of the least squares plane method to serve as anatomical landmarks for the subsequent device simulation (Figure 4) (1). With integrated device contours, the neo-LVOT can be segmented by means of a centerline technique to finally generate an orthogonal cross-sectional plane for planimetric “neo-LVOT” assessment (Figure 5). Here, we illustrate the concept of the neo-LVOT as well as the simulation and segmentation techniques essential to neo-LVOT prediction using cardiac CT (Figure 6). Transcatheter valve implantation for native mitral valve disease and for failed surgical mitral valve replacement or repair elongate the outflow tract into the left ventricle, referred to as neo-LVOT. Pre-procedural CT-based device simulation may aid in identifying patients at an increased anatomical risk for small neo-LVOT dimensions and ultimately obstruction of the outflow tract.
From the aSt. Paul’s Hospital, University of British Columbia, Vancouver, BC, Canada; bGuy’s and St. Thomas’ Hospital, London, United Kingdom; cSt. Vincent’s Hospital, Sydney, Australia; dNew York-Presbyterian Hospital and the Weill Cornell Medical College, New York, NY; and the eMcGill University Health Centre, Montreal, QC, Canada. Drs. Blanke and Leipsic provide core lab services to Edwards Lifesciences, Tendyne Holdings, and Neovasc; and serve as consultants to Edwards Lifesciences, Neovasc, and Circle Imaging. Drs. Webb, Dvir, Ye, and Cheung are consultants for Edwards. Dr. Piazza is a consultant for Medtronic. Manuscript received October 5, 2015; revised manuscript received January 19, 2016, accepted January 20, 2016.
2
Blanke et al.
JACC: CARDIOVASCULAR IMAGING, VOL.
-, NO. -, 2016 - 2016:-–-
Predicting LVOT Obstruction in TMVI
FIGURE 1
Concept of the Neo-LVOT in
TMVR
(Upper row) End-systolic 3-chamber view on cardiac computed tomography. (Lower row) Schematic drawing. Before TMVR, the LVOT is confined by the basal septum, the intervalvular fibrosa, and basal portion of the AML (A and D). The TMVR device deflects the AML “septally,” thereby elongating the outflow tract toward the left ventricle (B and E). This elongation is referred to as neo-LVOT, confined by the basal septum and the septally deflected AML (C and F). AML ¼ anterior mitral valve leaflet; LVOT ¼ left ventricular outflow tract; TMVR ¼ transcatheter mitral valve replacement.
FIGURE 2
Cross-Sectional Imaging
Planes of LVOT and Neo-LVOT
(Left) systolic 3-chamber view on cardiac computed tomography; (middle) schematic drawing; (right) cross-sectional short axis images through LVOT and neoLVOT, with orientation indicated by red lines in the left and middle columns. (A to C) Confinements, orientation, and imaging axis of the native LVOT. After TMVR (Tendyne TMVR system, Tendyne Holdings, Roseville, Minnesota), the native LVOT at the level of the intervalvular fibrosa remains unchanged without encroachment of the device onto the native LVOT crosssectional area (D to E). TMVR results in elongation of the outflow tract into the left ventricle, referred to as neo-LVOT (G to I). Importantly, the neo-LVOT has a different anatomical axis (yellow dashed line, H) than the native LVOT (blue line, E), thus requiring different orientation of the crosssectional imaging planes (C and F, E). Abbreviations as in Figure 1.
JACC: CARDIOVASCULAR IMAGING, VOL.
-, NO. -, 2016
Blanke et al.
- 2016:-–-
Predicting LVOT Obstruction in TMVI
FIGURE 3
Anatomical and Device Related Factors Predisposing to Narrow the
Neo-LVOT Dimension
Greater device protrusion into the left ventricle, device flaring at its left ventricular outflow, larger aorto-mitral angulation, and more pronounced septal bulging will lead to narrower/smaller neo-LVOT dimensions. Abbreviation as in Figure 1.
FIGURE 4
Annular Segmentation and Trajectory for
TMVR, Valve-in-Valve, and Valve-in-Ring Procedures
(Left) Cardiac computed tomography in functional mitral regurgitation; (middle) status post-bioprosthetic mitral valve replacement; (right) status post-mitral valve repair with closed annuloplasty ring. In native anatomy, the mitral annulus is segmented as described recently (A and D) (2). For valve-in-valve and valve-in-ring procedures, the contour of the radiopaque sewing or annuloplasty ring, respectively, are segmented on long- (B and C) and short-axis images (E and F). Segmentations allow for definition of the annular plane (G to I) and subsequently the annular trajectory, which is oriented perpendicular to the annular plane while transecting the geometric center of the annular contour (J, 3-dimensional simulation views; K, L, fluoroscopic simulation). TMVR ¼ transcatheter mitral valve replacement.
3
4
Blanke et al.
JACC: CARDIOVASCULAR IMAGING, VOL.
-, NO. -, 2016 - 2016:-–-
Predicting LVOT Obstruction in TMVI
FIGURE 5
Simulation of a Balloon-Expandable
Transcatheter Heart Valve Into a Failed Surgical Mitral Prosthesis, Calcific Mitral Stenosis, and Simulation of TMVR at End-Systole
Neo-LVOT dimensions change during the cardiac cycle and are typically smallest at end-systole. (A to C) End-systolic 3-chamber views; (D to E), end-systolic cross-sectional area of the neo-LVOT orthogonal to and neo-LVOT centerline (pink and red lines, A to C). (A and D) Simulation of valvein-valve 26-mm balloon-expandable transcatheter heart valve with 20%/80% positioning in regard to the annular plane. The planimetered, predicted neo-LVOT area was 2.5 cm2 at end-systole. (B and E) Simulation of a 26-mm balloon-expandable transcatheter heart valve with 20%/80% positioning into a calcific mitral stenosis at systole and diastole. The simulated neo-LVOT is almost completely obliterated at end-systole, precluding the patient from the procedure. (C and F) Neo-LVOT segmentation for TMVR planning by simulating device specific contours using stereolithographic file integration. Abbreviations as in Figures 1 and 4.
FIGURE 6
Workflow
Overview of basic steps in post-processing for neo-LVOT prediction. (1) Annular segmentation (here, a sewing ring of surgical mitral valve replacement). (2) The annular trajectory is oriented perpendicular to the annular plane and is commonly different form the LV long axis. (3) Valve simulation is performed by superimposing a device shape (cylindrical in this case) with a defined height and radius. The device long axis is oriented parallel to the annular trajectory. The level of implantation in regard to the annular plane (e.g., more ventricular and less atrial implantation) can also be simulated. (4) The anticipated neo-LVOT formed by the septum and the simulated device is segmented using a centerline technique. (5) Finally, neo-LVOT dimensions are planimetrically assessed on a cross-sectional reformat perpendicular to the neo-LVOT centerline. Abbreviation as in Figure 1.
REPRINT REQUESTS AND CORRESPONDENCE: Dr. Philipp Blanke, Centre for Heart Valve Innovation–St Paul’s
Hospital & University of British Columbia, 1081 Burrard Street, Vancouver, Canada V6Z 1Y6. E-mail: phil. [email protected].
REFERENCES 1. Blanke P, Naoum C, Webb J, et al. Multimodality imaging in the context of transcatheter mitral valve replacement: establishing consensus among modalities and disciplines. J Am Coll Cardiol Img
2. Blanke P, Dvir D, Cheung A, et al. Mitral annular evaluation with computed tomography in the context of transcatheter mitral valve implantation. J Am Coll Cardiol Img 2015;8:
2015;8:1191–208.
612–5.
KEY WORDS LVOT obstruction, neo-LVOT, TMVI, TMVR, transcatheter mitral valve implantation, transcatheter mitral valve replacement
VOL.
ª 2016 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION PUBLISHED BY ELSEVIER
-, NO. -, 2016
ISSN 1936-878X/$36.00 http://dx.doi.org/10.1016/j.jcmg.2016.01.005
IMAGING VIGNETTE
Predicting LVOT Obstruction in Transcatheter Mitral Valve Implantation Concept of the Neo-LVOT Philipp Blanke, MD,a Christopher Naoum, MBBS,a Danny Dvir, MD,a Vinayak Bapat, MD,b Kevin Ong, MD,a David Muller, MBBS,c Anson Cheung, MD,a Jian Ye, MD,a James K. Min, MD,d Nicolo Piazza, MD,e Pascal Theriault-Lauzier, BSC,e John Webb, MD,a Jonathon Leipsic, MDa OUTFLOW TRACT OBSTRUCTION IS A FEARED AND POTENTIALLY LETHAL COMPLICATION OF TRANSCATHETER
mitral valve replacement (TMVR), mitral valve-in-valve (ViV), and valve-in-ring (ViR) procedures as well as implantation of transcatheter heart valves in calcific mitral valve disease. These procedures ultimately lead to elongation of the outflow tract into the left ventricle, whereas the pre-existing “native” left ventricular outflow tract (LVOT) confined by the most basal septum and the intervalvular fibrosa remains unchanged (Figure 1). The newly created elongation of the outflow tract is confined posteriorly by the deflected anterior mitral valve leaflet in TMVR (Figures 1 and 2), ViR, and calcific mitral valve procedures or by the deflected bioprosthetic leaflets and transcatheter heart valve struts in ViV procedures. To better discriminate from the native LVOT, we propose to refer to the newly created elongation as the “neo-LVOT” (1). Although risk factors for narrow neo-LVOT dimensions can be both device- and anatomy-related (Figure 3), we propose to perform a patientspecific assessment by means of computed tomography (CT)-based, virtual, stereolithographic device implantation. This requires segmentation of the native mitral annulus (2), the bioprosthetic sewing ring, or the annuloplasty ring, with subsequent definition of the annular plane and the annular trajectory by means of the least squares plane method to serve as anatomical landmarks for the subsequent device simulation (Figure 4) (1). With integrated device contours, the neo-LVOT can be segmented by means of a centerline technique to finally generate an orthogonal cross-sectional plane for planimetric “neo-LVOT” assessment (Figure 5). Here, we illustrate the concept of the neo-LVOT as well as the simulation and segmentation techniques essential to neo-LVOT prediction using cardiac CT (Figure 6). Transcatheter valve implantation for native mitral valve disease and for failed surgical mitral valve replacement or repair elongate the outflow tract into the left ventricle, referred to as neo-LVOT. Pre-procedural CT-based device simulation may aid in identifying patients at an increased anatomical risk for small neo-LVOT dimensions and ultimately obstruction of the outflow tract.
From the aSt. Paul’s Hospital, University of British Columbia, Vancouver, BC, Canada; bGuy’s and St. Thomas’ Hospital, London, United Kingdom; cSt. Vincent’s Hospital, Sydney, Australia; dNew York-Presbyterian Hospital and the Weill Cornell Medical College, New York, NY; and the eMcGill University Health Centre, Montreal, QC, Canada. Drs. Blanke and Leipsic provide core lab services to Edwards Lifesciences, Tendyne Holdings, and Neovasc; and serve as consultants to Edwards Lifesciences, Neovasc, and Circle Imaging. Drs. Webb, Dvir, Ye, and Cheung are consultants for Edwards. Dr. Piazza is a consultant for Medtronic. Manuscript received October 5, 2015; revised manuscript received January 19, 2016, accepted January 20, 2016.
2
Blanke et al.
JACC: CARDIOVASCULAR IMAGING, VOL.
-, NO. -, 2016 - 2016:-–-
Predicting LVOT Obstruction in TMVI
FIGURE 1
Concept of the Neo-LVOT in
TMVR
(Upper row) End-systolic 3-chamber view on cardiac computed tomography. (Lower row) Schematic drawing. Before TMVR, the LVOT is confined by the basal septum, the intervalvular fibrosa, and basal portion of the AML (A and D). The TMVR device deflects the AML “septally,” thereby elongating the outflow tract toward the left ventricle (B and E). This elongation is referred to as neo-LVOT, confined by the basal septum and the septally deflected AML (C and F). AML ¼ anterior mitral valve leaflet; LVOT ¼ left ventricular outflow tract; TMVR ¼ transcatheter mitral valve replacement.
FIGURE 2
Cross-Sectional Imaging
Planes of LVOT and Neo-LVOT
(Left) systolic 3-chamber view on cardiac computed tomography; (middle) schematic drawing; (right) cross-sectional short axis images through LVOT and neoLVOT, with orientation indicated by red lines in the left and middle columns. (A to C) Confinements, orientation, and imaging axis of the native LVOT. After TMVR (Tendyne TMVR system, Tendyne Holdings, Roseville, Minnesota), the native LVOT at the level of the intervalvular fibrosa remains unchanged without encroachment of the device onto the native LVOT crosssectional area (D to E). TMVR results in elongation of the outflow tract into the left ventricle, referred to as neo-LVOT (G to I). Importantly, the neo-LVOT has a different anatomical axis (yellow dashed line, H) than the native LVOT (blue line, E), thus requiring different orientation of the crosssectional imaging planes (C and F, E). Abbreviations as in Figure 1.
JACC: CARDIOVASCULAR IMAGING, VOL.
-, NO. -, 2016
Blanke et al.
- 2016:-–-
Predicting LVOT Obstruction in TMVI
FIGURE 3
Anatomical and Device Related Factors Predisposing to Narrow the
Neo-LVOT Dimension
Greater device protrusion into the left ventricle, device flaring at its left ventricular outflow, larger aorto-mitral angulation, and more pronounced septal bulging will lead to narrower/smaller neo-LVOT dimensions. Abbreviation as in Figure 1.
FIGURE 4
Annular Segmentation and Trajectory for
TMVR, Valve-in-Valve, and Valve-in-Ring Procedures
(Left) Cardiac computed tomography in functional mitral regurgitation; (middle) status post-bioprosthetic mitral valve replacement; (right) status post-mitral valve repair with closed annuloplasty ring. In native anatomy, the mitral annulus is segmented as described recently (A and D) (2). For valve-in-valve and valve-in-ring procedures, the contour of the radiopaque sewing or annuloplasty ring, respectively, are segmented on long- (B and C) and short-axis images (E and F). Segmentations allow for definition of the annular plane (G to I) and subsequently the annular trajectory, which is oriented perpendicular to the annular plane while transecting the geometric center of the annular contour (J, 3-dimensional simulation views; K, L, fluoroscopic simulation). TMVR ¼ transcatheter mitral valve replacement.
3
4
Blanke et al.
JACC: CARDIOVASCULAR IMAGING, VOL.
-, NO. -, 2016 - 2016:-–-
Predicting LVOT Obstruction in TMVI
FIGURE 5
Simulation of a Balloon-Expandable
Transcatheter Heart Valve Into a Failed Surgical Mitral Prosthesis, Calcific Mitral Stenosis, and Simulation of TMVR at End-Systole
Neo-LVOT dimensions change during the cardiac cycle and are typically smallest at end-systole. (A to C) End-systolic 3-chamber views; (D to E), end-systolic cross-sectional area of the neo-LVOT orthogonal to and neo-LVOT centerline (pink and red lines, A to C). (A and D) Simulation of valvein-valve 26-mm balloon-expandable transcatheter heart valve with 20%/80% positioning in regard to the annular plane. The planimetered, predicted neo-LVOT area was 2.5 cm2 at end-systole. (B and E) Simulation of a 26-mm balloon-expandable transcatheter heart valve with 20%/80% positioning into a calcific mitral stenosis at systole and diastole. The simulated neo-LVOT is almost completely obliterated at end-systole, precluding the patient from the procedure. (C and F) Neo-LVOT segmentation for TMVR planning by simulating device specific contours using stereolithographic file integration. Abbreviations as in Figures 1 and 4.
FIGURE 6
Workflow
Overview of basic steps in post-processing for neo-LVOT prediction. (1) Annular segmentation (here, a sewing ring of surgical mitral valve replacement). (2) The annular trajectory is oriented perpendicular to the annular plane and is commonly different form the LV long axis. (3) Valve simulation is performed by superimposing a device shape (cylindrical in this case) with a defined height and radius. The device long axis is oriented parallel to the annular trajectory. The level of implantation in regard to the annular plane (e.g., more ventricular and less atrial implantation) can also be simulated. (4) The anticipated neo-LVOT formed by the septum and the simulated device is segmented using a centerline technique. (5) Finally, neo-LVOT dimensions are planimetrically assessed on a cross-sectional reformat perpendicular to the neo-LVOT centerline. Abbreviation as in Figure 1.
REPRINT REQUESTS AND CORRESPONDENCE: Dr. Philipp Blanke, Centre for Heart Valve Innovation–St Paul’s
Hospital & University of British Columbia, 1081 Burrard Street, Vancouver, Canada V6Z 1Y6. E-mail: phil. [email protected].
REFERENCES 1. Blanke P, Naoum C, Webb J, et al. Multimodality imaging in the context of transcatheter mitral valve replacement: establishing consensus among modalities and disciplines. J Am Coll Cardiol Img
2. Blanke P, Dvir D, Cheung A, et al. Mitral annular evaluation with computed tomography in the context of transcatheter mitral valve implantation. J Am Coll Cardiol Img 2015;8:
2015;8:1191–208.
612–5.
KEY WORDS LVOT obstruction, neo-LVOT, TMVI, TMVR, transcatheter mitral valve implantation, transcatheter mitral valve replacement