Elsevier

International Journal of Cardiology

Volume 249, 15 December 2017, Pages 90-95
International Journal of Cardiology

Image-guidance for transcatheter aortic valve implantation (TAVI) and cerebral embolic protection

https://doi.org/10.1016/j.ijcard.2017.09.158Get rights and content

Abstract

The study was aimed at evaluation of the feasibility and potential benefit of image fusion (IF) of pre-procedural CT angiography (CTA) and x-ray (XR) fluoroscopy for image-guided navigation in transfemoral transcatheter aortic valve implantation (TAVI) with the strong focus on guiding the double-filter cerebral embolic protection device and valve prosthesis placement.

Methods

In 31 patients undergoing TAVI, image registration of CTA-derived 3D anatomical models of the relevant cardiac anatomy and vasculature, and live XR was performed applying a commercially available navigation tool. The approach was evaluated in terms of the accuracy of the overlay. In 27 TAVI patients with IF receiving double-filter cerebral embolic protection device overall procedure time, fluoroscopy time, radiation dose, and total volume of intra-procedural iodinated contrast agent (CA) were registered and compared to those of a control group of prospectively enrolled during the same period of time N = 27 patients receiving the same protection system but without IF.

Results and conclusions

Image co-registration and model-based guidance is feasible in TAVI procedures. The overlay facilitates placement of the embolic protection device, placement of the guide wire in the left ventricle and initial alignment of the valve prosthesis prior to final deployment, thus improving the confidence level of the operators during the procedure without compromising CA or XR dose.

Introduction

Transvascular procedures are commonly performed under x-ray (XR) fluoroscopy guidance. Even though XR provides high temporal and spatial resolution, only two-dimensional (2D) projections with poor soft-tissue contrast can be obtained. The required contrast agents (CA) and relevant doses of ionizing radiation are associated with potential biological risks for patients and the interventional team [1].

Transcatheter aortic valve implantation (TAVI) is an established treatment for patients with severe symptomatic calcific aortic stenosis, who are not eligible for surgery or who are at intermediate or high risk for surgical treatment [2], [3], [4], [5]. During TAVI a biological aortic replacement valve mounted on a stent frame is implanted via a catheter system under x-ray guidance. CA needs to be injected frequently to visualize the anatomy of the aortic root. Crucial for the outcome of the procedure is the correct pre-procedural device selection and correct placement regarding location and orientation relative to the annulus. Misalignment by only a few millimeters of its optimal position may cause perivalvular leakages, valve embolization or obstruction of the coronary arteries [6]. Development of transcatheter devices as well as the valve types has progressed rapidly during the last decade. Cerebral embolic protection devices most commonly consisting of two filters placed in the brachiocephalic artery and the left common carotid artery are used to significantly reduce the number of cerebral lesions and lesion volume [7], [8]. Placing of the protection device usually requires an additional left anterior oblique (LAO) projection angiogram with up to 40 ml additional CA for the respective artery visualization. Repetitive contrast administrations for guidance of placement contain the risk of renal failure in the elderly population with multiple co-morbidities which increases the risk of need for renal replacement therapy, infections leading to sepsis and mortality. In addition, rigid manipulation without contrast agent increases the risk of endothelial damage, dissection or event thromboembolism leading to an acute stroke.

Multimodal intraprocedural three-dimensional (3D) image fusion (IF) has the potential to augment the limited information available from XR with relevant anatomic soft-tissue structures. In recently published studies, initial clinical experience using IF for transvascular procedures showed the feasibility of the method in different applications [9], [10], [11] as well as the potential for reduction in radiation and contrast dose, and overall procedure time [12], [13], [14], [15], [16]. In the published approaches, visualization of anatomical information is achieved by projecting organ shape models or volume-rendered (VR) images on the live fluoroscopy screen. With the recent introduction of commercially available fusion packages on commercial x-ray systems, a wider range of transvascular interventions can potentially be supported by this technique.

During the planning of complex transvascular procedures, high-fidelity pre-interventional data from computed tomography angiography (CTA) is widely used for providing pre-interventional assessment on the target anatomy and respective transvascular access routes. Advanced visualization and guidance technology involving pre-acquired patient-specific 3D models of the relevant anatomic structures can greatly facilitate the relatively complex TAVI procedures by providing a more realistic anatomy of the aortic root and more accurate C-Arm angulation.

Previous work on image-based support for TAVI includes the modeling for procedure planning [17], [18], guidance by tracking the prosthetic valve in fluoroscopic images [19], fusion between ultrasound and fluoroscopic images [20], and a prototype fusion system for a fully automatic 2D–3D fusion of the 3D model provided by the interventional C-Arm cone beam CT (CBCT) and 2D-angiography [21]. Whereas the use of pre-interventional CTA data for image fusion with registration by means of performance of C-Arm CBCT system is intensively investigated in other applications [13], [14], [15], [16], its usage for improving the XR guidance during TAVI is so far only rarely exploited [11]. There are no data regarding its use for placement of embolic protection devices.

Therefore the objective of this study was to investigate the feasibility of x-ray projection-based fusion of pre-interventionally acquired CTA-derived models of the aorta, the carotid arteries, and the left ventricle and XR fluoroscopy during TAVI with special focus on guiding the placement of the cerebral embolic protection device and the aortic valve prosthesis.

Section snippets

Methods

In thirty-one patients presenting with severe aortic valve stenosis 3D models of the aorta, coronary arteries, brachiocephalic artery and the left common carotid artery, as well as the left ventricle (LV) were generated based on the contrast-enhanced 3D cardiac CT data and overlaid onto the XR fluoroscopy (EP Navigator Release 5.1.1.4, Philips Healthcare, Best, The Netherlands) during the implantation procedure. The study protocol conforms to the ethical guidelines of the 1975 Declaration of

Results

No significant difference in terms of patient clinical characteristics was found between IF group and non-IF control group [26 (23–28.3) vs. 25.6 (22.9–31.45), p = 0.503 for BMI and 70 (60–78) vs. 73 (62.5–85) kg, p = 0.261 for weight].

The automatic segmentation of the great vessels and main cardiac anatomic structures was successful in all patients with a segmentation time of less than 60 s. The additional manual segmentation of the brachiocephalic artery with right carotid and right subclavian

Discussion and conclusions

3D CTA has become the single most important imaging modality for examination of the abdominal and iliofemoral arteries [25]. The ability to use CTA-derived 3D images to complement the anatomic information obtained by x-ray fluoroscopy and provide a greater degree of real-time anatomical guidance during structural cardiac interventions in the cathlab holds great potential. Evaluation of the added value of multimodality image co-registration by means of intra-procedural CBCT was exploited

Disclosure of potential conflict of interest

On behalf of all authors, the corresponding authors state that there are no relationships that could be construed as a conflict of interest.

Acknowledgment

The authors would like to thank Dr. Horst Brunner (Dept. Radiology, Ulm University Medical Center) for ongoing support in CT data acquisition and protocol optimization.

References (25)

Cited by (14)

View all citing articles on Scopus
1

This author takes responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation.

2

Shared corresponding authorship.

View full text