Impact of Valve Over-Sizing After Transcatheter Aortic Valve Implantation With a Self-Expanding Valve: A Multislice Computed Tomography Study

Abstract: Background. In transcatheter aortic valve implantation (TAVI), prosthesis over-sizing prevents paravalvular leak (PVL). Strategies of over-sizing for self-expanding bioprostheses are not well established at present. Methods. Patients with aortic valve stenosis scheduled for TAVI underwent preprocedural multislice computed tomography. Based on the degree of over-sizing, a ROC curve was drawn to define the optimal value of valve sizing for reducing PVL after TAVI. Results. A total of 152 consecutive patients were included in the study (mean age, 79.95 ± 7.71 years; log EuroScore: 23.87 ± 8.93%). Based on the ROC curve, sizing of 14% was the optimal that would lead to less moderate/severe PVL (P<.01). Group 1 was defined as sizing <14% (n = 49 patients) and group 2 was defined as sizing ≥14% (n = 103 patients). During a follow-up period of 36 ± 14 months, a total of 9 patients died from group 1 vs 4 patients from group 2 (P<.01). Two of the patients who died had moderate/severe PVL and 11 had no/mild PVL (P=.27). From the population, a total of 49 patients (32%) were found to be in the “borderline” zone. Patients who received the smaller valve had lower mean left ventricular outflow tract diameter (P=.048), higher rate of calcium load (mild: 10 [32%] vs 13 [72%]; moderate: 16 [52%] vs 3 [17%]; severe: 5 [16%] vs 2 [11%]; P=.02) and lower mean of sinus of Valsalva diameter (P=.046) compared with patients who received the bigger valve. Conclusions. In patients undergoing TAVI, over-sizing the prosthesis at least 14% reduces PVL. In borderline cases, taking into consideration additional anatomical parameters may result in low rates of PVL.

J INVASIVE CARDIOL 2019;31(5):E76-E82.

Key words: TAVI, MSCT, over-sizing, self-expandable, PVL

Transcatheter aortic valve implantation (TAVI) has seen a dramatic rise in the past decade, especially in elderly patients who are at increased surgical risk and for those at low surgical risk but not suitable for surgery as per the heart team’s decision.^1,2 The widening of TAVI application to lower-risk patients, however, is hindered by potential TAVI complications, such as paravalvular leak (PVL) and permanent pacemaker (PPM) implantation.

To some extent, these complications can be attributed to inappropriate bioprosthesis size selection. According to current recommendations and based on multislice computed tomography (MSCT) data, the choice of prosthesis size depends on the aortic annulus size, which is evaluated through planimetry after the creation of a projection that is aligned with the basal hinge points of all three cusps that define the virtual basal ring in the true plane of the annulus.³ Preferably, in order to minimize PVL, the sizing of the bioprosthetic aortic valve in patients with tricuspid aortic valve should be larger than the native annulus. Still, significantly over-sized prostheses have been associated with higher rates of complications such as annulus rupture and higher rate of PPM implantation.

Multiple studies have been performed on balloon-expandable valves in order to determine the optimal sizing of the prosthetic valve in various annulus sizes.^4-7 In “borderline” cases, an intentionally reduced volume of fluid within the deployment balloon during the procedure leads to optimal sizing of the bioprosthesis and thus minimizes the risks associated with excessive over-sizing or under-sizing.⁴ On the other hand, data for self-expandable valves are scarce, and even though it is routine clinical practice for self-expandable valves to be over-sized by ~10%, most patients will not meet this target due to the large increments (3 mm) between manufactured prosthesis sizes. This problem may be further exaggerated in self-expanding valves with lower radial force.

Appropriate device sizing in borderline cases may be dependent on additional aortic root anatomical parameters. The aim of this study was to assess in patients undergoing TAVI with a self-expanding valve: (1) the impact of valve over-sizing as calculated by MSCT on postprocedural outcomes; and (2) the consideration of anatomical parameters for valve size selection.

Methods

Study population. A total of 152 consecutive high-risk or inoperable patients with severe symptomatic aortic valve stenosis undergoing TAVI with self-expandable transcatheter heart valves (THVs) were included in the study. CoreValve and EvolutR valves (provided by Medtronic, Inc) were used in all patients. Exclusion criteria were: previous aortic valve replacement (valve-in-valve); bicuspid aortic valve anatomy; and TAVI with a bioprosthesis other than the Medtronic THVs. In addition, patients with deep implantation of the bioprosthesis (>6 mm) were excluded from the study, based on previously published data that optimal implantation depth for the CoreValve is ≤6 mm below the annulus plane.⁸ By excluding patients with deep implantation, we aimed to eliminate this potential causation factor for PPM implantation and/or post-TAVI PVL. All data were prospectively collected and retrospectively analyzed. All patients gave their written consent for the procedure and data acquisition. Decision to perform TAVI was determined by the department’s heart team, and this study was approved by the hospital’s ethics committee.

Echocardiography. Transthoracic echocardiography was performed in all patients as part of the screening process. Severe aortic stenosis was defined based on current recommendations.¹ Transvalvular gradients and effective orifice areas were measured by the continuity equation, while left ventricular outflow tract (LVOT) was measured in the parasternal long-axis view in mid-systole parallel to the aortic valve plane and within 0.5–1.0 cm of the valve orifice, as previously described. Pulmonary artery systolic pressure (PASP) was defined as moderate if ≥31 mm Hg and ≤55 mm Hg and severe if PASP was >55 mm Hg.^9,10 Decision to perform TAVI was based on the severity of the symptoms of aortic valve stenosis, risk evaluation, and consideration of special contraindications to surgery. Transthoracic echocardiograms were performed prior to discharge in all patients and were read by a physician with experience in assessing TAVI patients who was blinded to clinical and procedural parameters. The semiquantitative assessment of PVL severity was based on the estimation of the circumferential extent of the color jet relative to the valve ring in the short-axis view. We consider <10% of the total circumferential extent to be mild, and >30% is considered moderate to severe PVL.¹¹

MSCT protocol and image reconstruction. The MSCT examination protocol has been previously explained in detail.^5,12 Briefly, a dual-source MSCT scanner with prospective electrocardiogram triggering, with a high pitch of 3.4 and a low tube voltage at the level of 100 kV, was used for all measurements. All images and calculations were performed in early systole, when the aortic valve annulus is at its greatest size.¹³ Injection of 80-120 mL of radiocontrast medium at 5 mL/s followed by 30 mL of normal saline was used for all patients.⁴ A commercially available and dedicated postprocessing software (3mensio, Pie Medical Imaging) was used for all measurements, and all studies were evaluated by experienced physicians.

Sizing. For bioprosthesis size selection, previously published literature and the manufacturer-recommended CT-based sizing algorithm (23 mm prosthesis for annuli 18-20 mm; 26 mm prosthesis for annuli 20-23 mm; 29 mm prosthesis for annuli 23-26 mm; 31 mm/34 mm prosthesis for annuli 26-29 mm/26-30 mm) were taken into consideration.

Additional anatomical criteria were particularly taken into consideration for valve selection in borderline-zone patients, such as those with an annulus-derived diameter of 19.5-20.5 mm (borderline zone for 23 mm or 26 mm bioprostheses), 22.5-23.5 mm (borderline zone for 26 mm or 29 mm bioprostheses), or 25.5-26.5 mm (borderline zone for 29 mm or 31 mm/34 mm bioprostheses). These anatomical criteria were evaluated by MSCT and were based on the operators’ experiences, and included the following: (1) semiquantitative calcium deposition of the aortic annulus; (2) LVOT width; and (3) mean diameter of the sinus of Valsalva.

For analysis purposes, in the entire study population, the degree of valve over-sizing was calculated with the following mathematical equation:

Over-sizing = [(device – annulus) / annulus] x 100

where “device” is the bioprosthesis size and “annulus” is the perimeter-derived annulus from the MSCT. Based on this equation, a receiver operating characteristic (ROC) curve was drawn in order to define the optimal value of valve sizing with an aim to reduce PVL after TAVI.

Procedure. The screening and TAVI procedures have been described previously.^14,15 All procedures were performed under strict sterile conditions in the catheterization lab and a heart surgeon was only present in the non-transfemoral cases when surgical cut was necessary.¹⁶ As part of our institution’s minimal TAVI approach, the majority of cases were performed with local anesthesia and mild sedation, unless a non-transfemoral route had been planned.¹⁷ Stand-by transthoracic echocardiography system with an experienced operator was available in case an emergent scan was necessary.

Definitions and study endpoints. All definitions, measured outcomes, and endpoints were designated according to the Valve Academic Research Consortium-2 criteria.¹⁸ The primary endpoints were all-cause mortality and PVL rates post TAVI. Additional analyses explored cerebrovascular events, bleeding, other catastrophic complications, and new PPM implantation.

Statistical analysis. Continuous variables are presented as mean ± standard deviation and compared with the Student’s t-test. The normality of distribution was assessed using the Shapiro-Wilk test and normality diagrams. Categorical variables are presented as frequencies and percentages and were tested by the Chi-square test. Kaplan-Meier mortality curves were built from the time of the procedure up until follow-up in all patients was complete. ROC curves were generated using moderate/severe PVL post TAVI as the endpoint. P-values <.05 were considered statistically significant. The analysis was performed with SPSS 24 statistical software (SPSS, Inc).

Results

Study population. A total of 152 consecutive patients were included in the study (mean age, 79.95 ± 7.71 years; log EuroScore, 23.87 ± 8.93%). A sizing of 14% was found to be the optimal percentage that would lead to lower rates of moderate/severe PVL in the entire patient cohort. The cut-off value of 14% was derived by including all patients in the ROC analysis in a retrospective fashion, and the area under the curve (AUC) was used in order to measure the predictive accuracy of our values (AUC, 0.806; 95% confidence interval [CI], 0.706-0.905; P<.01) (Figure 1). We attempted to achieve an equal value for sensitivity and specificity, to the extent that it was feasible; the sensitivity was 71% and the specificity was 75% (Youden index J, 0.53; 95% CI, 0.39-0.66; positive predictive value, 97%; negative predictive value, 18%). Based on this value, the cohort was categorized into two groups. Group 1 was defined as sizing <14% (n = 49 patients) and group 2 was defined as sizing ≥14% (n = 103 patients). Tables 1 and 2 provide the patients’ baseline, echocardiographic, and MSCT characteristics according to valve over-sizing groups.

Sizing for patients in a borderline zone. Forty-nine patients (32%) from the entire study population were found to be in the borderline zone. Based on the presence of the prespecified additional anatomical criteria, a smaller valve was selected in 31 patients (64%) while 18 patients (36%) received the bigger valve. Patients who received the smaller valve had lower mean LVOT diameter (23.15 ± 2.86 mm vs 24.77 ± 2.31 mm; P=.048), higher rate of calcium load (mild: 10 [32%] vs 13 [72%]; moderate: 16 [52%] vs 3 [17%], severe 5 [16%] vs 2 [11%]; P=.02) and lower mean sinus of Valsalva diameter (31.16 ± 4.07 mm vs 33.54 ± 3.62 mm; P=.046) compared with patients who received the bigger valve.

Procedural data. All study patients received Medtronic self-expandable valves, with the majority (60%; n = 92) receiving the newer-generation EvolutR THV. The majority of cases were performed from the transfemoral access route; 15 cases were performed from the subclavian artery and 1 case was performed from the direct aortic route. Local anesthesia and mild sedation were the mainstay for the majority of TAVI cases, with general anesthesia performed in the non-transfemoral TAVI cases and in 4 transfemoral cases where the patients were deemed extremely high risk. Predilation before TAVI was similar between the two groups (P=.56) (Table 3).

Echocardiographic characteristics post TAVI. Patients from group 1 had similar peak velocity (2.15 ± 0.39 m/s vs 2.13 ± 0.46 m/s; P=.85), peak gradient (19.88 ± 8.48 mm Hg vs 19.95 ± 9.17 mm Hg; P=.96), mean gradient (9.83 ± 3.77 mm Hg vs 10.25 ± 5.08 mm Hg; P=.65), and indexed effective orifice area (1.85 ± 0.35 cm²/m² vs 1.93 ± 0.16 cm²/m²; P=.18) compared with patients from group 2. Patients from group 1 had higher rates of moderate/severe PVL compared with group 2 (none/mild: 40 [81%] vs 100 [97%]; moderate/severe: 9 [19%] vs 3 [3%]; P<.01).

Patients in the borderline zone who received the smaller valve had similar peak velocity (4.39 ± 0.54 m/s vs 4.43 ± 0.48 m/s; P=.78), peak gradient (78.83 ± 18.46 mm Hg vs 80.65 ± 19.55 mm Hg; P=.74), mean gradient (48.73 ± 11.59 mm Hg vs 47.50 ± 10.84 mm Hg; P=.71), and indexed effective orifice area (0.40 ± 0.16 cm²/m² vs 0.34 ± 0.10 cm²/m²; P=.44) compared with patients who received a bigger valve. Interestingly, the rate of PVL did not differ between patients who received the smaller valve compared with patients who received the bigger valve (none/mild: 30 [97%] vs 18 [100%]; moderate/severe: 1 [3%] vs 0 [0%]; P=.33).

Clinical outcomes. During a follow-up period of 36 ± 14 months, a total of 9 patients from group 1 had died compared to 4 patients from group 2 (18% vs 4%, respectively; P<.01) (Table 4 and Figure 2). Among the patients who died, only 1 patient (7.6%) had periprocedural death (due to left ventricular perforation), while 12 patients (92.3%) died during the follow-up period (4 patients [33.3%] due to lower respiratory tract infection, 5 patients [41.6%] due to sepsis, 2 patients [16.6%] due to cancer, and 1 patient [8.3%] due to accident). Among those who died, 2 patients (15.4%) had moderate/severe PVL and 11 patients (84%) had none/mild PVL (16% vs 8%; P=.27). The rate of stroke did not differ between groups.

Patients in the borderline zone who received the smaller valve had similar all-cause mortality rates (2 [6.5%] vs 0 [0%]; P=.17), new PPM implantation rates (12 [39%] vs 5 [28%]; P=.43), stroke (0 [0%] vs 1 [0.05%]; P=.15), and major bleeding events (1 [4%] vs 2 [11%]; P=.27) compared with patients who received a bigger valve.

Discussion

Although adherence to the manufacturer’s recommended annular sizing algorithm has favorable clinical outcomes after placement of a self-expanding aortic valve, this study showed the following: (1) over-sizing the prosthesis by ~14% significantly reduces the risk of PVL, without any additional risk; and (2) in borderline cases, the consideration of additional anatomical parameters may result in lower PVL rates without an increase in overall clinical events. Moreover, even though we found no significant differences in the need for a PPM based on the different levels of over-sizing, larger studies are needed to fully examine the relationship between over-sizing and conduction-system disturbances.

Previous studies have already pointed out several factors implicated in PVL after TAVI, including valve undersizing.^19,20 In our cohort, patients from group 1 had higher moderate/severe PVL rates post TAVI (19% vs 3% in group 2; P<.02). Debry et al demonstrated similar results in terms of PVL rates.⁶ However, meaningful comparison is not possible with this study, as different methodologies were used for the calculation of valve over-sizing and it involved patients receiving both Medtronic and Edwards valves. Similarly, Buzzatti et al used different cut-off values for denoting valve over-sizing and found that 7% over-sizing in general for both the Medtronic and Edwards valves was associated with lower PVL rates.²¹ In addition, they showed over-sizing of 11.5% was necessary for the Medtronic valves in order to reduce significant PVL. In our total cohort of patients, over-sizing of 14% was shown by ROC curve analysis to be the cut-off value for minimizing the risk of moderate/severe PVL post TAVI (AUC, 0.806; 95% CI, 0.706-0.905; P<.01). Although we found no significant differences in the need for PPM based on over-sizing, larger studies are needed to fully examine the relationship between over-sizing and conduction-system disturbances.

In balloon-expandable valves utilized in borderline cases, it has been shown that an intentionally reduced volume of fluid within the deployment balloon during the procedure leads to optimal sizing of the bioprosthesis, thus minimizing the risks associated with excessive over-sizing or under-sizing.⁴ In contrast, data regarding the concept of over-sizing in self-expanding bioprostheses are very limited, and an option such as the fluid method does not exist for balloon-expandable valves; thus, alternative solutions must be sought.

Using a variety of techniques (echocardiography, MSCT, angiography), multiple parameters have already been related to PVL, particularly large annulus dimensions, valve calcifications, LVOT-ascending aorta angle, and depth of prosthesis implantation.^22-24 In cases in the borderline zone, wherein either a larger or a smaller device can be selected, we have found that considering additional anatomical parameters as calculated by MSCT may result in low rates of PVL. More specifically, we calculated the mean LVOT diameter, the semiquantitative calcium load of the aortic annulus, and the mean sinus of Valsalva diameter. We found that patients in the borderline zone who received a smaller valve had lower mean LVOT diameter, higher rate of calcium load, and lower mean sinus of Valsalva diameter compared with patients who received a bigger valve. Indeed, among 49 patients who were in the borderline zone, only 1 patient (2.0%) had moderate/severe PVL after TAVI. This finding may have important implications for valve selection in patients with borderline perimeters. For example, if the mean LVOT diameter and the mean sinus of Valsalva diameter are small with a high calcium load in a patient with an annular-perimeter based diameter of 23 mm, the selection of a 26 mm bioprosthesis may be considered instead of a 29 mm bioprosthesis. Similarly, if the mean LVOT diameter and the mean sinus of Valsalva diameter are big with a low calcium load in a patient with an annular-perimeter based diameter of 26 mm, the selection of a 31 mm bioprosthesis may be considered instead of a 29 mm bioprosthesis. Thus, our data support the hypothesis that additional anatomical parameters may be utilized in conjunction with the degree of over-sizing as depicted by MSCT to help direct the size of the prosthesis in order to achieve optimal clinical outcomes after TAVI. The consideration of the aforementioned three anatomical factors (especially in borderline cases) is based on our experience. Specifically, aortic annulus calcium deposition, LVOT width, and mean sinus of Valsalva diameter should be considered when planning a TAVI for patients with borderline sizes.

At mid-term follow-up, all-cause mortality was 9%, with a higher incidence among patients in group 1 vs patients in group 2 (18% vs 4%, respectively; P<.01). Overall, only 1 patient died periprocedurally due to left ventricular perforation, while the rest died due to non-cardiac causes. Debry et al used a different methodology for grouping their population; however, they showed that patients with normal sizing had marginally higher mortality rates compared with moderate or severe over-sizing patients at 30 days (15% vs 7% vs 5%, respectively; P=.08).⁶ We found no aortic root rupture in either group – a finding that further reinforces the idea that over-sizing is a safe approach in TAVI patients.^25,26 We observed that more than moderate PVL had an unfavorable impact on mortality, although non-cardiac deaths were encountered in this study. The current analysis demonstrated that over-sizing of 14% and consideration of additional anatomical parameters improved PVL post TAVI. Although previous studies have demonstrated that more than moderate PVL plays a significant role in cardiac mortality, we did not confirm the aforementioned finding, as the primary aim of this study was to investigate the procedural factors that may improve PVL rates.

Study limitations. This is a retrospective study and all limitations that may apply should be taken into consideration. Our results can be considered hypothesis-generating for future studies. The study included a small number of patients and is subject to confounding bias. As a consequence of the small cohort, there are few cases of moderate to severe PVL, with the overall incidence in accordance with previously published studies. The 31 mm CoreValve was not available until the latter part of the study and was later replaced by the 34 mm EvolutR bioprosthesis. In addition, PVL estimation was not performed during the procedure and postdilation was performed according to the discretion of the operator based on the leak noticed intraprocedurally. It cannot be excluded that the final PVL assessment may have been influenced by postdilation in a subgroup of patients. Still, overall the rates of postdilation were non-significant between the two groups (13% in group 1 vs 17% in group 2; P=.62).

There are some sizing differences between two valve generations. However, subgroup analysis was not performed due to the small size of the study population. Still, when the analysis is limited to patients who received an EvolutR valve, only small deviations are found for the optimal sizing (oversizing of ~15% decreases the rate of moderate/severe PVL [AUC, 0.751; P=.03; 95% CI, 0.607-0.896; Youden index J, 0.48; 95% CI, 0.32-0.57]).

Conclusion

In patients undergoing TAVI with a self-expanding valve, over-sizing the prosthesis by at least 14% as measured by MSCT reduces the risk of PVL. Moreover, in borderline cases when either a large or a smaller device can be selected, taking into consideration additional anatomical parameters may result in low PVL rates.

References

1. Baumgartner H, Falk V, Bax JJ, et al. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J. 2017;38:2739-2791.

2. Cribier A, Eltchaninoff H, Bash A, et al. Percutaneous transcatheter implantation of an aortic valve prosthesis for calcific aortic stenosis: first human case description. Circulation. 2002;106:3006-3008.

3. Piazza N, de Jaegere P, Schultz C, Becker AE, Serruys PW, Anderson RH. Anatomy of the aortic valvar complex and its implications for transcatheter implantation of the aortic valve. Circ Cardiovasc Interv. 2008;1:74-81.

4. Binder RK, Webb JG, Willson AB, et al. The impact of integration of a multidetector computed tomography annulus area sizing algorithm on outcomes of transcatheter aortic valve replacement: a prospective, multicenter, controlled trial. J Am Coll Cardiol. 2013;62:431-438.

5. Blanke P, Willson AB, Webb JG, et al. Oversizing in transcatheter aortic valve replacement, a commonly used term but a poorly understood one: dependency on definition and geometrical measurements. J Cardiovasc Comput Tomogr. 2014;8:67-76.

6. Debry N, Sudre A, Elquodeimat I, et al. Prognostic value of the ratio between prosthesis area and indexed annulus area measured by multiSlice-CT for transcatheter aortic valve implantation procedures. J Geriatr Cardiol. 2016;13:483-488.

7. Dvir D, Webb JG, Piazza N, et al. Multicenter evaluation of transcatheter aortic valve replacement using either SAPIEN XT or CoreValve: degree of device oversizing by computed-tomography and clinical outcomes. Catheter Cardiovasc Interv. 2015;86:508-515.

8. Petronio AS, Sinning JM, Van Mieghem N, et al. Optimal implantation depth and adherence to guidelines on permanent pacing to improve the results of transcatheter aortic valve replacement with the Medtronic CoreValve system: the CoreValve prospective, international, post-market ADVANCE-II study. JACC Cardiovasc Interv. 2015;8:837-846.

9. Galie N, Hoeper MM, Humbert M, et al. Guidelines for the diagnosis and treatment of pulmonary hypertension: the task force for the diagnosis and treatment of pulmonary hypertension of the European Society of Cardiology (ESC) and the European Respiratory Society (ERS), endorsed by the International Society of Heart and Lung Transplantation (ISHLT). Eur Heart J. 2009;30:2493-2537.

10. Baumgartner H, Hung J, Bermejo J, et al. Echocardiographic assessment of valve stenosis: EAE/ASE recommendations for clinical practice. J Am Soc Echocardiogr. 2009;22:1-23.

11. Pibarot P, Hahn RT, Weissman NJ, Monaghan MJ. Assessment of paravalvular regurgitation following TAVR: a proposal of unifying grading scheme. JACC Cardiovasc Imaging. 2015;8:340-360.

12. Latsios G, Spyridopoulos TN, Toutouzas K, et al. Multi-slice CT (MSCT) imaging in pretranscatheter aortic valve implantation (TAVI) screening. How to perform and how to interpret. Hellenic J Cardiol. 2018;59:3-7. Epub 2017 Oct 28.

13. Jurencak T, Turek J, Kietselaer BL, et al. MDCT evaluation of aortic root and aortic valve prior to TAVI. What is the optimal imaging time point in the cardiac cycle? Eur Radiol. 2015;25:1975-1983.

14. Toutouzas K, Latsios G, Stathogiannis K, et al. One-year outcomes after direct transcatheter aortic valve implantation with a self-expanding bioprosthesis. A two-center international experience. Int J Cardiol. 2016;202:631-635.

15. Stathogiannis K, Synetos A, Toutouzas K, et al. Medtronic CoreValve: achieving optimal outcomes. Hellenic J Cardiol. 2015;56(Suppl A):4-8.

16. Toutouzas K, Synetos A, Latsios G, et al. The requirement of extracorporeal circulation system for transluminal aortic valve replacement: do we really need it in the catheterization laboratory? Catheter Cardiovasc Interv. 2018;91:E43-E48. Epub 2015 May 6.

17. Melidi E, Latsios G, Toutouzas K, et al. Cardio-anesthesiology considerations for the trans-catheter aortic valve implantation (TAVI) procedure. Hellenic J Cardiol. 2016;57:401-406.

18. Kappetein AP, Head SJ, Genereux P, et al. Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document. Eur Heart J. 2012;33:2403-2418.

19. Kodali SK, Williams MR, Smith CR, et al. Two-year outcomes after transcatheter or surgical aortic-valve replacement. N Engl J Med. 2012;366:1686-1695.

20. Wilczek K, Bujak K, Regula R, Chodor P, Osadnik T. Risk factors for paravalvular leak after transcatheter aortic valve implantation. Kardiochir Torakochirurgia Pol. 2015;12:89-94.

21. Buzzatti N, Maisano F, Latib A, Cioni M, et al. Computed tomography-based evaluation of aortic annulus, prosthesis size and impact on early residual aortic regurgitation after transcatheter aortic valve implantation. Eur J Cardiothorac Surg. 2013;43:43-50.

22. Gripari P, Ewe SH, Fusini L, et al. Intraoperative 2D and 3D transoesophageal echocardiographic predictors of aortic regurgitation after transcatheter aortic valve implantation. Heart. 2012;98:1229-1236.

23. Sherif MA, Abdel-Wahab M, Stocker B, et al. Anatomic and procedural predictors of paravalvular aortic regurgitation after implantation of the Medtronic CoreValve bioprosthesis. J Am Coll Cardiol. 2010;56:1623-1629.

24. Abdel-Wahab M, Comberg T, Buttner HJ, et al. Aortic regurgitation after transcatheter aortic valve implantation with balloon- and self-expandable prostheses: a pooled analysis from a 2-center experience. JACC Cardiovasc Interv. 2014;7:284-292.

25. Ando T, Takagi H, Telila T, Afonso L. Comparison of outcomes in new-generation versus early-generation heart valve in transcatheter aortic valve implantation: a systematic review and meta-analysis. Cardiovasc Revasc Med. 2018;19:186-191. Epub 2017 Jul 13.

26. Ruparelia N, Latib A, Kawamoto H, et al. A comparison between first-generation and second-generation transcatheter aortic valve implantation (TAVI) devices: a propensity-matched single-center experience. J Invasive Cardiol. 2016;28:210-216.

From the First Department of Cardiology, Medical School of Athens University, Hippokration Hospital, Athens, Greece.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Toutouzas reports proctor income from Medtronic CoreValve. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript submitted October 1, 2018, provisional acceptance given November 20, 2018, final version accepted January 9, 2019.

Address for correspondence: Konstantinos Toutouzas, MD, Hippokration Hospital, 26 Karaoli and Dimitriou Streets, 15562 Holargos, Athens, Greece. Email: ktoutouz@gmail.com