Everything is Better in 3D: Why Transcatheter Aortic Valve Replacement Should Be Guided By Multidetector Computed Tomography

Keywords:

• 2D echocardiography
• 3D echocardiography
• annulus
• iliofemoral
• multidetector computed tomography
• paravalvular regurgitation
• transcatheter aortic valve replacement

Transcatheter aortic valve replacement (TAVR) now provides hope for patients who would be considered poor candidates for traditional surgical valve replacement. It is a procedure requiring an integrated team approach to ensure proper patient and valve selection. Imaging for valve size selection has traditionally been performed with 2D echocardiography, with good results (as outlined in the Partner A and B trials [1,2]). However, outcomes may be improved further with the more complete integration of multidetector computed tomographic (MDCT) imaging. This editorial will discuss the role MDCT has to play in TAVR planning, and provide evidence to support the claim that ‘everything is better in 3D‘.

TAVR is now available as an option to treat severe symptomatic aortic stenosis for a patient population in which comorbid conditions would have routinely precluded a standard surgical aortic valve replacement [1,2]. More so than with conventional surgery, imaging plays a major role in appropriate patient and valve selection, as well as procedure planning. In the early days of this procedure, this role was performed solely by 2D echocardiography and conventional angiography [3]. The automated and reproducible 3D data sets provided by MDCT may be better suited for anatomic assessment of the aortic annulus and the iliofemoral systems. Importantly, MDCT has the ability to acquire data to perform both these tasks with a single acquisition.

Iliofemoral tortuosity and minimal diameter assessment is important to help avoid peripheral vascular complications when performing TAVR [4]. Assessment of vascular access generally begins with conventional angiography, with assessment of the iliofemoral systems at the time of coronary angiography. However, additional information with regard to vessel tortuosity, burden and pattern of calcification, extent of atherosclerosis, and identification of other high-risk features, including dissection and complex atheroma/ulceration, is gained with performance of MDCT. At our institution, major and minor vascular access complications have been significantly reduced by a more systematic screening process, which included computed tomography angiography. Moderate or severely calcified iliofemoral arteries and vessels with minimal luminal diameters of less than the external delivery sheath diameter, as assessed by MDCT, are strong predictors of vascular complications (29 vs 9%, p = 0.03 and 23 vs 5%, p = 0.01, respectively) [5].

Assessment of annular size and orientation is essential to TAVR success. If the valve is implanted too low there is an increased risk of heart block, paravalvular regurgitation (PAR) and mitral valve dysfunction [6]. Alternatively, if the valve is positioned too high, there is an increased risk of valve embolization, PAR and aortic root injury [7]. Current implantable valves are designed for specific annular sizes, thus accurate preprocedure sizing is critical. Surgical aortic valve replacement prosthesis size is determined under direct visualization during open heart surgery. Clearly, transcatheter heart valve (THV) sizing cannot be performed by direct inspection, but rather relies exclusively on pre- and intra-procedural imaging to do so. Annulus measurements are typically performed using 2D transthoracic echocardiography, transesophageal echocardiography (TEE), or calibrated aortic angiography. Comparison between these methods is fairly limited [8].

The aortic annulus is a complex structure, which has been demonstrated to have a noncircular, often oval, configuration on both 3D TEE and MDCT [9]. The noncircular nature of the aortic annulus confounds methods of assessment that are capable of imaging in 2D only, such as 2D echocardiography or conventional angiography. A study by Messika-Zeitoun et al. demonstrated that measurements taken from MDCT of the annulus could alter TAVR strategy in nearly 40% of patients [10]. These results have caused some confusion with groups performing TAVR, as the success rate has been very good utilizing traditional ultrasound-based measures.

As a result, the clinical utility of MDCT with regards to annulus sizing has remained limited in TAVR assessment, particularly with the balloon expandable prosthesis. Our group has previously determined that the most reproducible MDCT measurements are the mean of the long and short axis of the basal ring and the basal ring area [11]. These measurements are typically 1–1.5 mm larger than TEE measures of the annulus. Given the historical sizing of THVs on the basis of a single annular dimension, Gurvitch et al. provided an initial proposal to develop a MDCT-based sizing scale for TAVR. It entailed formal integration of these differences and an increase in the thresholds for valve sizing by 1–1.5 mm, such that we use the following sizing guidelines when using MDCT-derived mean basal ring: 23 mm valve for annulus ≥19.5 to ≤22.5 mm, 26 mm valve for >22.5 to ≤26.5 mm, and 29 mm valve for >26.5 to ≤29.5 mm [11].

While clinical outcomes of patients receiving percutaneously placed aortic valves sized preoperatively based on echocardiographic measures have been good, there is still room for improvement. Even with routine oversizing of the implanted valve by 1–2 mm, moderate-to-severe PAR is observed in 10–15% of patients postprocedure [1,2]. The clinical importance of PAR post-TAVR is gaining recognition. The recently published 2-year PARTNER outcome data not only suggest that 40% of patients alive at 2 years have at least mild PAR, but perhaps more importantly that those patients with PAR have a significantly worse prognosis than those without [12]. While PAR may not be exclusively related to undersizing, there is growing awareness that the limitations of 2D imaging does not allow for a true understanding of annular geometry and thus can lead to inappropriate THV selection. Recently, Willson et al. have shown that 3D acquired area measures of the annulus can in fact predict PAR [13]. When the nominal area of the THV is not larger than the area of the annulus, the likelihood of moderate/severe PAR is quite high [13]. Based on these findings, a number of groups, including ours, have begun integrating area and perimeter measurements into THV selection and annular sizing [13,14]. While there is a growing integration of MDCT data in valve selection for TAVR, it is important to recognize that this information is now one of many pieces of data that are included in the decision-making process.

Additionally, given limited sizes of aortic valves available full integration of 3D computed tomography data are not yet possible.

MDCT can also aid in determination of the orientation of the aortic root relative to the body axis. While the aortic valve and annulus are most commonly directed cranially and anteriorly, with a slight degree of angulation to the right, there is significant variability between patients. Standard practice with implantation is to perform arch aortograms in a number of orthogonal planes to ensure coaxial deployment, perpendicular to the aortic root. At our center, we employ the following method utilizing images derived from MDCT acquisitions, to help guide placement prior to procedure. Double oblique transverse projections of the aortic root and basal ring are created. Using this projection a triangle formed by three points deposited on the inferior most aspect of the aortic cusps is created. A volume-rendered image with the triangle superimposed is generated. The reconstruction can then be manipulated through a series of any angles, with perpendicularity assured by superimposition of the triangle formed at the inferior most aspect of the aortic cusps. Computer software yields angiographic projections perpendicular to the native valve plane in three axes: cranial–caudal without right anterior oblique or left anterior oblique (LAO) angulation, straight right anterior oblique to LAO as needed without cranial or caudal angulation and LAO 30° with cranial or caudal angulation based on manipulation of the volume-rendered image. The aforementioned axes have been chosen based on past experiences; in addition, they are considered reasonable working angles given the physical constraints of the catheterization laboratory. The ability to provide this kind of information preprocedure can help to limit the number of aortograms performed before implantation and thus, decrease procedural time, radiation dose and reduce the volume of iodinated contrast administered. The limitation of contrast volume is an importance consideration given the prevalence of renal dysfunction and subsequent risk of contrast-induced nephropathy in a TAVR cohort [15].

MDCT for TAVR is not without drawbacks. As mentioned previously, iodinated contrast administration is a necessity for TAVR planning with MDCT. Given the prevalence of renal dysfunction in this patient population, this can be a problematic issue. Techniques including direct injection of contrast via an arterially placed catheter have been used with some success to help limit contrast volume [16]. Radiation dose is obviously higher with MDCT compared with TEE or transthoracic echocardiography, although given the age and life expectancy of patients forming the bulk of TAVR candidates, this is felt to only be a theoretical concern. Incidentally, detected findings on MDCT for TAVR planning also pose an interesting problem. They are common and are often significant enough to warrant a change in patient management [17]. How such findings are handled in a patient population with a limited life expectancy will therefore require a thoughtful approach.

While the purpose of this article was to extol the virtues of MDCT for TAVR planning, in reality, this field requires a team approach with the best features of various imaging modalities utilized to optimize patient outcome. MDCT is an attractive tool given its ability to image the complex structure of the aortic annulus with great anatomic detail and provide accurate assessment of the iliofemoral system all in a single examination. It is almost certain MDCT will play a prominent and more defined role in the planning of TAVR procedures in the future.

Financial & competing interests disclosure

J Leipsic is a consultant, advisory board member and speaker‘s bureau member for General Electric Healthcare. J Leipsic is a member of the speaker‘s bureau and consultant for Edwards Lifesciences Inc. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.