Analysis of Aortic Calcifications in 4 Different Transcatheter Heart Valve Prostheses in Order to Reduce Paravalvular Aortic Leak

Objectives: The wide application of transcatheter aortic valve implantation (TAVI) is bordered also because of paravalvular leak (PVL). Multislice computed tomography (MSCT) analysis of aortic calcifications might lead to better preoperative planning and consequently a reduction of PLV. Aim our study was to investigate PVL after the implantation of 4 different transcatheter heart valve prostheses by means of aortic calcium degree.

Methods: We retrospectively analyzed prospectively collected data from patients receiving a TAVI at our center. Only patients with severe native aortic valve stenosis were included in the study. Aortic calcium volume was calculated in the aortic root, left ventricular outflow tract and both (=device landing zone or “DLZ”) from the preoperative contrast-enhanced MSCT using a dedicated software 3mensio Structural Heart (v7.0SP1, Medical Imaging BV, the Netherlands). Risk factors for PVL were analyzed using a multivariate analysis.

Results: Between October 2009 and 2016, 711 patients underwent TAVI. 539 patients were eligible for the study (SapienXT = 192; Sapien3 = 206; Corevalve = 44; Symetis = 97). All patients underwent TAVI because of high surgical or prohibitive risk. In the study population transfemoral TAVI was performed in 66% of cases. Median calcium volume in DLZ was 757 mm³ (IQR = 734) and distribution was not different between prosthetist groups (ANOVA test p > 0.05). Severe PVL was 0%. Overall incidence of mild-to-moderate PVL was 15.8% (SapienXT = 17.7%; Sapien3 = 6.8%; Corevalve = 31.8%; Symetis = 23.7%). Multivariate analysis based on preoperative clinical and MSCT characteristics identified critical risk factors such as calcium volume of DLZ, transapical access, use of Corevalve and Symetis as independently associated with PVL. Preoperative permanent dialysis, oversizing and use of Sapien3 were independently associated with a lower incidence of PVL. Receiver operating characteristic analysis of calcium volume in DLZ showed an area under the curve of 0.66 (95%CI:0.617–0.699,p = 0.0001) meaning an acceptable discrimination. A cutoff higher than 1079 mm³ has been identified as a limit beyond which significantly increases the risk of PVL.

Conclusions: DLZ's calcium volume influences significantly PVL's occurrence after TAVI. Its measurement in procedural planning should become a routine to reduce PVL. The presence of a DLZ's volume calcification higher than 1079mm³, should influence the strategy, including choice of prosthesis and sizing.