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 . 9 , N O . 1 1 , 2 0 1 6

ª 2 0 1 6 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 UN 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 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 6 . 0 1 . 0 1 7

IMAGING VIGNETTE

Predicting LVOT Obstruction After TMVR

Dee Dee Wang, MD,a Marvin Eng, MD,a Adam Greenbaum, MD,a Eric Myers, BFA,b Michael Forbes, BFA,b

Milan Pantelic, MD,c Thomas Song, MD,c Christina Nelson, RTRCT,c George Divine, PHD,a Andrew Taylor, MA,a

Janet Wyman, DNP,a Mayra Guerrero, MD,a,d Robert J. Lederman, MD,e Gaetano Paone, MD,a William O’Neill, MDa

EVOLUTION OF CATHETER-BASED STRUCTURAL INTERVENTIONS HAS GIVEN PATIENTS LESS INVASIVE

alternatives to surgery; however, the current generation of transcatheter heart valves (THV) are not specificallydesigned for mitral position implantation and have intrinsic geometry that may make mitral implantationsuboptimal. Operators are faced with unique challenges with valve deliverability, embolism, and notably left

FIGURE 1 CT Guided Sizing of the Mitral Annulus Landing Zone and Fit-Testing CAD-Generated THV into Patient-Specific Mitral

Anatomy to Predict LVOT Obstruction

Pre-procedural imaging utilizes a contrast-enhanced, retrospectively electrocardiogram-gated computed tomography (CT) angiography

acquisitionwithout application of electrocardiogram-dosemodulation (Table 1). Four-dimensional cine clips of themitral annulus (MA)motion are

generated. The MA plane is defined as the basal-most insertion of the mitral leaflets, characterized by a hinge-point motion of the leaflet base

during the mid-to-late diastolic phase of the cardiac cycle, or in the case of a degenerative mitral ring/bioprosthesis, by prosthesis-artifact on CT.

Using the double-oblique method, crosshairs are aligned to the MA plane in sagittal and coronal cross-sections (A and B) allowing for MA

dimensions to be obtained on axial thin sections (C). Once the proposed THV size is identified, computer-aided design (CAD)-generated Sapien,

Sapien XT (SXT) and Sapien 3 valves (Edwards Lifesciences, Irvine, California) are then virtually tested on-screen and again physically in the

patient’s 3DP anatomy to confirm sizing and estimate risk of LVOT obstruction (D). For difficult to-conceptualize computer-based 3D imaging

anatomy, 3DP THVmodels were implanted into the 3DP patient-specific systolic left ventricular outflow tract (LVOT)model to help the structural

heart team visualize the predicted LVOT obstruction to anatomical scale. HU¼Hounsfield unit(s); TAVR¼ transcatheter aortic valve replacement;

3D ¼ 3-dimensional; 3DP ¼ 3-dimensionally printed; TMVR ¼ transcatheter mitral valve replacement.

FIGURE 2 LVOT CAD Prediction Modeling Algorithm

The blood volume of the mid-to-late systolic phase

demonstrating the smallest LVOT surface area between

the basal anteroseptal wall of the LV and the most

inferior/ventricular hinge of the anterior mitral leaflet or

surgical frame strut is segmented out for 3D modeling.

CAD models are generated with the proposed THV

deployed in different LV landing zones (60% LV/40%

atrial to 80% LV/20% atrial) within the patient’s MA/ring

(A). In the setting of surgical bioprosthesis, modeling is

performed with 0% defined within the confines of the

inferior border of the surgical prosthesis strut, and 10% as

inferior to the surgical prosthesis plane into the LV.

The post-TMVR residual LVOT, “neo-LVOT,” surface area is

then calculated for proposed THV by various depths of LV

deployment and angulations of THV delivery system

toward/away from aortic outflow tract (B and C). In order

to predict percentage of LVOT obstruction, the ratio of

neo-LVOT to native LVOT surface area is calculated:

(native LVOT area – neo LVOT area)/native LVOT area.

Abbreviations as in Figure 1.

From the aDivision of Cardiology, Center for Structural Heart Disease, Henry Ford Health System, Detroit, Michigan; bHenry Ford

Innovation Institute, Henry Ford Health System, Detroit, Michigan; cDivision of Radiology, Henry Ford Health System, Detroit,

Michigan; dDivision of Cardiology, Evanston Hospital/NorthShore University Health System, Evanston, Illinois; and theeCardiovascular and Pulmonary Branch, Division of Intramural Research, National Heart, Lung, and Blood Institute, National

Institutes of Health, Bethesda, Maryland. Drs. Wang, O’Neill, Mr. Myers, and Mr. Forbes are co-inventors on a patent

application, assigned to their employer Henry Ford Health System, on software to predict LVOTO. Dr. Greenbaum is a proctor

for Edwards Lifesciences and St. Jude Medical. Dr. Guerrero has received a research grant from and is a proctor for Edwards

Lifesciences. Dr. Paone is a consultant and proctor for Edwards Lifesciences. Dr. O’Neill is a consultant for Edwards

Lifesciences, Medtronic, and St. Jude Medical. All other authors have reported that they have no relationships relevant to the

contents of this paper to disclose.

Manuscript received November 18, 2015; revised manuscript received January 13, 2016, accepted January 14, 2016.

Wang 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 . 9 , N O . 1 1 , 2 0 1 6

Predicting LVOT Obstruction After TMVR N O V E M B E R 2 0 1 6 : 1 3 4 9 – 5 2

1350

FIGURE 3 CT, Computer-Aided-Design, and 3D Print Analysis of the LVOT to Characterize and Quantify the Residual Post-TMVR

“Neo-LVOT” Area Overcomes the Limitations of Traditional 2D Imaging Planes

Preliminary data utilizing 2-dimensional (2D) echo imaging correlated acute angulation of the mitral aorta-outflow-angle (mAOA) with higher

risk of LVOT obstruction compared with that of more obtuse mAOA. On the basis of our single-center experience, the association between

mAOA and risk of LVOT obstruction was not reproducible (A and B). LVOT obstruction is not solely dependent on mAOA. Personalization and

determination of feasibility of TMVR in patient-specific anatomy may not be accounted for if relying on 2D measurements of a 3D anatomy (A).

CT and CAD 3DP analysis of the “neo-LVOT” surface area overcomes the limitations of traditional 2D imaging planes. THV implantation in

the mitral position results in permanent anterior motion/displacement of the surgical/native anterior mitral leaflet. Both THV frame and

permanent anterior motion/displacement can result in severe LVOT obstruction (C). The importance of understanding the application of this

technology is the ability to test devices in patient-specific cardiac anatomy ex vivo for LVOT encroachment. Abbreviations as in Figure 1.

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 . 9 , N O . 1 1 , 2 0 1 6 Wang et al.N O V E M B E R 2 0 1 6 : 1 3 4 9 – 5 2 Predicting LVOT Obstruction After TMVR

1351

TABLE 1 TMVR CT Scanning Protocol

Scan Type Gated CTA Chest

Field of view Start scan acquisition: above the lung apices

Complete scan acquisition: at the bottom of rib cage

Algorithm Standard

Scan helical thickness 0.625 mm

0.16:0.36

0.18:0.36

Pitch:gantry rotation speed, s* 0.20:0.36

0.22:0.36

0.24:0.36

Kilovolts 100 kV if BMI <30 kg/m2

120 kV if BMI >30 kg/m2

Milliamp ECG-modulated mA

Minimum of 200 to maximum of 600 cardiac RR 40–80

Adaptive statistical iterative reconstruction 30%

Recon 2 (for pertinent radiological findings) 2.5 mm

(75% of cardiac cycle)

Recon 3 (for mitral annulus sizing and LVOT evaluation) 1.25 mm

(5%–95% of cardiac cycle by 10% increments for 4Dcine loop of dynamic mitral annulus motion and LVOT evaluation)

Recon 4 (additional LVOT evaluation as needed) 0.625 mm

(25%–55% by 10% increments)

Prep delay SmartPrep at level of left atrium, trigger at 120 HU

Injection rate 4 cc/s

Injection volume 80 cc

Auto transfer PACS

Enter “gated mitral” in exam description

Set maximum mA at phases 35–80 and r-peak center to 75

Injector protocol

4 cc/s 80 cc contrast

3 cc/s 30 cc saline

*Pitch:speed should be adjusted according to the heart rate gating algorithm provided by different vendors as recommended for routine cardiac gating acquisition protocols.

4D ¼ 4-dimensional; BMI ¼ body mass index; CT ¼ computed tomography; CTA ¼ computed tomography angiography; ECG ¼ electrocardiography; HU ¼ Hounsfield unit(s);LVOT ¼ left ventricular outflow tract; PACS ¼ Picture Archiving Communication System; RR ¼ the intervals between 2 R-waves of the ECG; TMVR ¼ transcatheter mitral valvereplacement.

Wang 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 . 9 , N O . 1 1 , 2 0 1 6

Predicting LVOT Obstruction After TMVR N O V E M B E R 2 0 1 6 : 1 3 4 9 – 5 2

1352

ventricular outflow tract (LVOT) obstruction. Therefore, understanding the suitability of prosthesis deliveryand implantation individualized to each heart is of paramount importance.

Successful transcatheter mitral valve replacement (TMVR) depends on accurate sizing of the mitral annulus(Figure 1) and avoidance of LVOT obstruction. Incorporation of computer-aided design and generation of3-dimensional-printed heart models allows for ex vivo device bench testing in patient-specific anatomy(Figures 1 and 2). Modeling of proposed THV at different angles/depths of deployment into the LV allowsestimation of LVOT obstruction of neo-LVOT/LVOT (Figure 3). We now aim to describe the utility of cardiaccomputed tomography (Table 1) and ex vivo THV fit testing with 3-dimensional models to predict LVOTobstruction in TMVR.

ACKNOWLEDGMENTS The authors thank Dr. Scott Dulchavsky, Mark Coticchia, and Lindsay Klee of the HenryFord Innovation Institute for their support toward clinical 3-dimensional printing.

REPRINT REQUESTS AND CORRESPONDENCE: Dr. Dee Dee Wang, Henry Ford Health System, Institute forStructural Heart Disease, 2799 West Grand Boulevard, K14, Detroit, Michigan 48202-2608. E-mail: dwang2@hfhs.org.

KEY WORDS 3D print, computer aided design, LVOT obstruction, transcatheter mitral valve replacement