Computed Tomography Planning for Transcatheter Mitral Valve Replacement

Figure 1. Anatomy of the mitral valve and subvalvular apparatus. (a) Anterior mitral leaflet (from lateral to medial, A1, A2, and A3) and posterior mitral leaflet (P1, P2, and P3) in end-systolic cardiac phase. (b) Anterior mitral leaflet (from lateral to medial, A1, A2, and A3) and posterior mitral leaflet (P1, P2, and P3) in end-diastolic cardiac phase. (c) Yellow arrows: anterolateral and posteromedial papillary muscles with chordae tendineae in long-axis 2-chamber view.

Figure 2. TMVR analysis of the mitral annulus by MDCT. (a1 and a2) Determine the center of the mitral valve (yellow dot) and LV apex (green dot) in 2 orthogonal views. This defines an MV centroid axis, which will be used for the next step, which is the identification of the mitral annulus. (b) Interpolated cubic spline is created using a semi-automated way rotating around the mitral annulus centroid point created on the prior step. Sixteen seed points are deposited at the base of the MV leaflets and connected to create the mitral annulus. Each point can be selected and position corrected. The area of the mitral-aortic curtain always appears thicker delineating the transition to a less fibrotic annulus. (c) Define the lateral and septal trigones recognized by the transition point whereby aortic root disappears from the orthogonal plane. (d1) Saddle-shaped mitral annulus. (d2) D-shaped mitral annulus.

Figure 3. Primary and secondary mitral regurgitation seen through multiphasic MDCT. (a) Primary mitral valve regurgitation. Posterior MV leaflet prolapse and a large flail gap (yellow arrow) seen at end-systolic cardiac phase in long-axis 3-chamber view. (b) Secondary mitral valve regurgitation. Both anterior and posterior leaflets are apically tethered along with subvalvular apparatus due to severe dilation and left ventricular remodeling causing restricted MV leaflet motion and incomplete coaptation.

Figure 4. Absolute milliseconds multiphasic MDCT reconstruction for irregular heart rhythm. The shortest R-R interval (fastest heart rate) is identified; in this case, 98 beats per minute (panel a, red dashed box). Divide 60,000 ms by the shortest heart rate (60,000/98 = 612 ms). Then round it to the closest 50 ms increment (i.e., 600 ms) and make a functional reconstruction from 0 to 600 ms at 50 ms increments (panel b).

Figure 5. The methodology for preprocedure Neo-LVOT area assessment using specific TMVR Tendyne Device modeling. (a) Define mitral annulus plane in the midsystolic cardiac phase. (b) Insert the virtual TMVR device model, which has adequate oversizing relative to mitral annulus, and draw centerline for the Neo-LVOT (solid orange line). (c) Measure Neo-LVOT area at the narrowest point through Neo-LVOT centerline axis. A large Neo-LVOT area is predicted at 2.3 cm2 (orange area).

Figure 6. TMVR Neo-LVOT area valve simulation with a Tendyne device modeled before and after alcohol septal ablation. (a1 and a2) Measured neo-LVOT area before alcohol septal ablation was 1.6 cm2 (orange area) at midsystolic cardiac phase. (b1 and b2). MDCT scan was repeated 3 mo after pre-emptive alcohol septal ablation. Same midsystolic phase chosen with an increased neo-LVOT area to 2.6 cm2.

Figure 7. Transapical access planning with MDCT. (a) Volume-rendered images showing virtual handles (in pink through the true LV apex, in blue through the coaxial mitral annular trajectory, and in green the target for apical access) and their position relative to the intercostal space. In this case, as the coaxial annular handle is in the center of the intercostal space, the target (green) handle could be positioned overlapping it, which is ideal but not always possible. (b) Endoluminal volume-rendered views from the LV short and long axis showing the angle (yellow) between the target handle (green) and the true apex (pink).

Figure 8. Transseptal access planning and catheter simulation by MDCT. (a) To help in the transseptal procedural planning, the catheter trajectory (blue) is simulated in the same mid-late systolic phase used for mitral annulus measurement. Using volume-rendered MDCT images, the simulated catheter starts from the inferior vena cava, crosses the fossa ovalis (FO), and gets to the mitral valve (MV) annulus level. Two coaxial lines, one from the center of the FO and the other from the center of the MV, are automatically drawn and connected (dashed line in light blue). The distances from the puncture site to the mitral valve and the angle to achieve the most coaxial orientation for THV deployment are estimated. (b) Fluoroscopic projections derived from MDCT for the FO en-face plane and mitral annulus (MA) en-face plane are provided to guide TEE septal puncture and THV coaxial deployment.

Figure 9. Fluoroscopic projections assessment by MDCT. The 3 main MDCT-derived coplanar fluoroscopic projections used for THV navigation and deployment are provided. They are named based on the directions of mitral annular measurements already described: septal-lateral (SL), trigone-to-trigone (TT), or intercommisural (IC). A compromise view between IC and SL views is usually provided due to C-arm angulation constraints to achieve an IC or TT view.

Figure 10. Valve-in-valve CT planning. (a) The internal dimensions of the degenerated SHV bioprosthesis are measured for optimal THV sizing. On the right side image, the maximal intensity projection demonstrates it to be a #25 ​mm Edwards Magna 3300 (pericardial leaflets). (b) A virtual cylinder with the same specifications provided by the manufacturer regarding width and height of a SAPIEN 3 is simulated and positioned inside the bioprosthesis. The ventricular end of the pre-existent stent posts is used as a limiting reference for the virtual THV (yellow box); if the stent posts are not radiopaque, calcification and hypodense stent posts on contrast CT images can be used as surrogates for proper positioning. Then, in the mid-late systolic phase, a centerline is traced between the virtual THV and the neo-LVOT (orange line), and a perpendicular plane is positioned at the narrowest level (green arrow). This plane, a short-axis plane represented in the last picture, is used for the Neo-LVOT area measurement as demonstrated in orange.

Figure 11. Types and characteristics of mitral surgical heart valves by MDCT (3D maximal intensity projection rendering). On top (orange), bovine valves; on the bottom (blue), porcine valves. Observe that the stent posts and annulus rings of bovine surgical valves are radiopaque and well defined by CT. The porcine Hancock II has 3 circular radiopaque markers on top of the stent post, but the Epic does not have a radiopaque stent post, just a thin radiopaque annulus ring.

Figure 12. Valve-in-ring MDCT planning and complications. (a) Internal dimensions of a semirigid mitral annuloplasty ring (32 mm memo 3D) are measured as demonstrated. A 3D reconstruction shows, in the figure on the left, the extensive mitral annular calcification (MAC) below the ring. (b) As described, a virtual THV is positioned, and the Neo-LVOT is measured at the narrowest distance between the septum and the virtual THV. In this case, 2 factors are potential predictors of LVOT obstruction beyond the Neo-LVOT area, which in theory would be adequate (2.5 cm2). First, the MAC below the annulus may induce THV device canting toward the LVOT and reduce the neo-LVOT area. Second, the anterior mitral leaflet (AML) is preserved, and after device deployment, it may induce dynamic LVOT obstruction because the THV frame will dislocate it anteriorly toward the LVOT. The anterior mitral leaflet has the same length as the ring-to-septal distance, indicating a higher risk for this complication. (c) In the same case, a post valve-in-ring MDCT has shown the 2 potential complications described in panel b. The posterior MAC induced device canting (yellow arrowhead) toward the LVOT, and most importantly, a…