Author + information
- Titus Kuehne, MD* (, )
- Sevim Yilmaz, MD,
- Carolin Meinus, MD,
- Phillip Moore, MD,
- Maythem Saeed, DVM, PhD,
- Oliver Weber, PhD,
- Charles B. Higgins, MD,
- Thiemo Blank, PhD,
- Erhard Elsaesser,
- Bernhard Schnackenburg, PhD,
- Peter Ewert, MD,
- Peter E. Lange, MD and
- Eike Nagel, MD
- ↵*Department of Congenital Heart Diseases and Pediatric Cardiology, German Heart Institute, Augustenburger Platz 1, Berlin, 13553, Germany
Successful transcatheter aortic valve implantation in the aortic valve position has been recently reported (1,2). However, implantation remains challenging, and malpositioning of the device can lead to a life-threatening situation. In contrast to conventional X-ray angiography, magnetic resonance imaging (MRI) provides precise three-dimensional information about both soft tissue anatomy and catheter position as well as immediate parameters of cardiovascular function (3). The purpose of this study was to assess the feasibility of MRI to guide transcatheter valve implantation in the aortic valve position.
Self-expanding, non-covered nitinol stents of 20 mm diameter (Angiomed, Karlsruhe, Germany) were used as a carrier for a prosthetic heart valve. A tricuspid valve was fabricated from a 0.1-mm-thick polytetrafluoroethylene membrane (Impra, Phoenix, Arizona) and sutured into the stent. The valved stent was compressed and front-loaded into a 10-F delivery system that comprised a pig tail and a working lumen. For catheter visualization during MRI, small ferromagnetic markers (size=0.5 mm3) were mounted on the delivery system distally and proximally to the loaded valved stent, comparable to radiopaque markers for conventional X-ray angiography.
Interventions were performed in pigs (n = 7, weight = 26 ± 3 kg) under general anesthesia (2% isoflurane). A 12-F dilator sheath was placed, using Seldinger techniques, in the carotid (n = 4) and iliac artery (n = 3). Thereafter, animals were moved into a 1.5-T MR scanner (Intera, Philips Medical Systems, Best, the Netherlands).
Before intervention, phasic aortic blood flow was measured in the ascending aorta using velocity encoded cine (VEC) MRI (repetition time [TR] = 16 ms, echo time [TE] = 9 ms, field of view [FOV] = 200 to 350, matrix = 256 × 256, phase encoding velocity [Venc] = 200 cm/s). A set of balanced fast field echo (bFFE) images was used to evaluate cardiovascular anatomy including the diameter of the aortic valve annulus and its distance to the coronary artery orifices and the mitral valve. Left ventricular (LV) function (ejection fraction) and biventricular wall motion were assessed in a short-axis plane using standard protocols (TR = 2.7 ms, TE = 1.4 ms, flip angle = 50°, FOV = 200 to 350, matrix = 256 × 256, phases = 20 to 25).
For implantation, the delivery system was advanced from the introducer-sheath to the LV under MR fluoroscopy without use of guide wires. The MR fluoroscopy was based on interactive real-time, steady-state, free processing sequences (TR = 2.7 ms, TE = 1.4 ms, FOV = 200 to 350, matrix = 144 × 144, acquisition frame rate = 10 frames/s; images were displayed online on an in-room monitor, and investigators had hearing protection). Coordinates of the susceptibility markers were taken until the distal marker was seen 2 mm below the aortic valve annulus, and a set of multiphase bFFE images was used to confirm final catheter position. The release of the valved stent was finally monitored under MR fluoroscopy. The aim was to implant the annulus of the prosthetic valve on top of the annulus of the native aortic valve.
After implantation, the following parameters were assessed to determine the success of the intervention: 1) patency of the mitral valve and coronary artery orifices; 2) LV function and biventricular wall motion using bFFE; and 3) phasic aortic blood flow within the lumen of the valved stent using VEC MRI. Finally, the position and morphology of the valved stents were determined at autopsy after the animals were euthanized with pentobarbital.
Paired Student ttest was used to compare measurements made before and after the intervention and the MRI- and postmortem-derived measurements.
All seven animals had successful implantation of a prosthetic aortic valve without misplacement or cardiovascular trauma. The average duration of the MRI-guided intervention was 53 ± 11 min. Preoperative, bFFE provided clear images for assessment of cardiovascular anatomy (Fig. 1).The diameter of the aortic valve annulus was measured at systole with 17.4 ± 2.1 mm and its distance to the coronary artery orifices and the upper rim of the mitral valve with 10.5 ± 2.8 mm and 8.2 ± 2.7 mm, respectively. At autopsy, measured diameters/distances were 17.8 ± 3.1 mm, 11.1 ± 3.4 mm, and 8.6 ± 2.4 mm, respectively (p = 0.9). The ferromagnetic markers on the shaft of the delivery system produced a circumscript susceptibility artefact, which was easily perceptible during MRI and enabled clear determination of the position of the loaded valved stent in relation to its surrounding anatomy (Fig. 1).
Susceptibility and radiofrequency shielding effects of the stent allowed quick recognition of the valved stent during and after its delivery but caused no image distortion of the adjacent anatomy (Figs. 1 and 2).⇓After implantation, bFFE images demonstrated the perfusion of the proximal coronary arteries (Fig. 2). Wall motion abnormalities were not observed. The LV ejection fraction as measured before and after valved stent implantation was 61 ± 5% and 60 ± 6%, respectively (p = 0.8). Assessment of quantitative blood flow with VEC MRI showed no aortic regurgitation in five animals and mild central valve incompetence with a regurgitant fraction of 8.2 ± 3.1% in two cases (Fig. 2). There was no evidence for aortic valve stenosis. Peak flow velocities as measured within the lumen of the nitinol stent were 1.1 ± 0.5 m/s compared with 1.0 ± 0.7 m/s measured before the implantation (p = 0.7). Unobstructed mitral valve leaflets were evidenced in all experiments (Fig. 2). Postmortem examination revealed accurate implantation of the prosthetic valve with a precision of 2.3 ± 1.1 mm and confirmed unobstructed coronary arteries (Fig. 2) and mitral valve leaflets.
The current study demonstrates that transcatheter implantation of a prosthetic valve in the aortic valve position is feasible under MRI guidance. The MRI method features the advantage of acquiring images that contain precise three-dimensional information about both soft tissue anatomy and the interventional instrument. Passive MRI catheter tracking by use of susceptibility markers enabled precise position monitoring of the interventional instrument. The 10-F delivery system was easy to control without guide wires. However, the development of guide wires suited for interventional MRI is warranted.
Preoperatively, MRI allowed noninvasive assessment of cardiovascular anatomy and function. During and after implantation, the position of the prosthetic valve was easy to determine. Besides this morphologic information, MRI provided immediate post-intervention physiologic parameters of cardiac function and perfusion of the proximal course of the coronary arteries. In addition, VEC MRI demonstrated optimal function of the implanted prosthetic valve in five cases and mild central valvular insufficiency in two cases.
In conclusion, the results of this study imply that MRI is a promising tool for transcatheter placement of valved stents. However, further validation of safety aspects is needed in chronic studies with larger numbers of animals.
- American College of Cardiology Foundation
- Boudjemline Y.,
- Bonhoeffer P.
- Cribier A.,
- Eltchaninoff H.,
- Bash A.,
- et al.