Sizing the annulus for transcatheter aortic valve implantation: more than a simple measure?

Abstract

Accurate preoperative assessment of the aortic annulus dimension is crucial for the success of transcatheter aortic valve implantation (TAVI). All the available imaging modalities, however, have some weak points, and the ‘gold standard’ to be used is still under debate. In this context, the recent work by Dr Blanke and coworkers is a commendable attempt to develop a standardized, easy to reproduce method for the identification and measurement of the aortic annulus [1, 2]. A very interesting finding of this study is the close correlation found between the preoperative cross-sectional area (CSA) of the virtual ring passing through the nadir of the three aortic cusps and the postoperative CSA. This observation would suggest that preoperative determination of the CSA could help in predicting the final form and size of the aortic annulus in most TAVI patients [1]. However, the behaviour of the aortic root and aortic annulus is not always predictable based on the preoperative exams [3]. This is partly due to intrinsic anatomic properties of the aortic root—the ‘virtual ring’ is largely inhomogeneous, coursing through the muscular septum, the membranous septum and the mitro-aortic curtain. Moreover, leaflet and annular calcifications are often asymmetrical, adding further complexity to any attempt to predict the final shape of the prosthetic valve, the displacement of the native calcified leaflets and the sealing of the annulus around the prosthesis. A very good exemplification of this problem is given by the bicuspid aortic valve, which is usually ovoidal and heavily calcified, and may fail to adapt to the circular transcatheter valve [4]. For this reason, we have developed a simple method to analyse the reaction of the aortic root to the balloon inflation, and to obtain a dynamic measure of the CSA. In the uncertain cases (‘borderline’ annulus, important discrepancy between TEE and CT measurements, massive and/or eccentric calcifications), we proceed to a ‘calibrated’ balloon valvuloplasty. A conventional valvuloplasty balloon (Cristal Balloon, Balt, Montmorency, France) is inflated with a diluted contrast agent through a threeway stopcock. When the waist generated on the balloon profile by the aortic annulus is stretched, the stopcock is closed and an aortography is performed to exclude the presence of aortic regurgitation. The amount of contrast agent used to inflate the balloon is carefully recorded, and the balloon is deflated and withdrawn. On the bench, the balloon is inflated again to the same volume, and is then sized with a sizing plate (Amplatzer sizing plate, AGA Medical Italia, Milan, Italy). This simple manoeuvre allows us to accurately measure the aortic annulus, to estimate the risk of postoperative paraprosthetic leakage and to check the relationships between the dislocated aortic valve leaflets and the coronary arteries ostia. In our experience, this procedure has led to a change in the size of the percutaneous prosthesis in three patients, and to the conversion to transapical approach in two patients needing a 29 mm prosthesis. In conclusion, we believe that the CT protocol described by Blanke could greatly enhance our ability to evaluate the anatomy of the aortic root preoperatively. In doubtful cases, however, calibrated balloon valvuloplasty could add fundamental dynamic information, allowing us to ‘feel’ the aortic annulus while measuring it [3].