The introduction of the rapid-deployment (RD) valve into clinical practice has expanded the already rich portfolio of aortic valve substitutes for patients undergoing aortic valve replacement (AVR). Both European [1] and American [2] trials (TRITON and TRANSFORM trials), along with numerous subsequent studies, have demonstrated excellent clinical outcomes following AVR procedures using RD valves.

The RD valve is known to exhibit several advantages over conventional bioprostheses. It decreases aortic cross-clamp time and cardiopulmonary bypass time, thus shortening the overall procedure [2,3]. It also simplifies and facilitates minimally invasive AVR. Furthermore, the hemodynamic performance of RD valve is reported to be superior to that of conventional bioprostheses [4,5].

Despite these advantages of the RD valve, bicuspid aortic valves (BAVs) have been considered a relative contraindication for the implantation of an RD bioprosthesis [6]. BAVs have different anatomic properties of the aortic root with respect to annulus geometry (more elliptic), sinus asymmetry, and leaflet commissures, which often have different heights, especially in the presence of pure BAVs [7]. It has been suggested that RD valves may increase the risk of paravalvular regurgitation and/or potential dislocation due to the asymmetry of the BAV aortic root [8,9].

Our group recently reported excellent early and mid-term outcomes of AVR using RD valves in a real-world all-comers population [10]. In that report, rapid-deployment AVR (RDAVR) using Edwards Intuity was performed for various indications, including bicuspid valve, pure aortic regurgitation, and infective endocarditis. In particular, BAVs represented 52.6% of all cases, and true BAVs (type 0 bicuspid) were identified in 11.2% of the overall cohort. Thus, BAVs may not represent a relative contraindication for RDAVR and might yield comparable outcomes to those of RDAVR for tricuspid aortic valves (TAVs).

The aim of this study was to compare the early echocardiographic and clinical outcomes of RDAVR for BAVs versus TAVs.

Study population

The study protocol was reviewed by the Seoul National University Hospital Review Board and approved as a minimal-risk retrospective study (approval no., H-2301-153-1400), and the requirement for individual consent was waived.

From June 2016 to June 2022, a total of 378 aortic valve procedures were performed by a single surgeon (K.H.K.), 278 of which included AVR using Edwards Intuity (Edwards Lifesciences, Irvine, CA, USA). The surgeon preferentially implanted Intuity if a patient planned to receive a bioprosthetic valve, and only 13 patients received a bioprosthetic valve other than Intuity during the study period. Of 278 patients who underwent AVR using Edwards Intuity, 252 patients were enrolled in this study after excluding 18 pure aortic regurgitations, 5 prosthetic valve failures, 2 cases of infective endocarditis, and 1 quadricuspid valve. In the study population, 147 patients were diagnosed with BAVs and 105 were diagnosed with TAVs, respectively (Fig. 1).

Figure 1. Flow diagram of patient enrollment. AVR, aortic valve replacement.

The diagnosis of BAVs was initially assessed by preoperative echocardiography and computed tomography, and the definitive diagnosis with the specific Sievers classification was made after a thorough intraoperative evaluation for all patients [11].

Operative techniques and strategy

The basic surgical procedures and strategies of RDAVR have been described in the previous study [12]. All operations were performed via median sternotomy, under conventional cardiopulmonary bypass, and. with mild or moderate hypothermia and cardioplegic arrest. After aortotomy, aortic valve excision, and annular decalcification, the valve replica always simulated the annulus; the semilunar design of the valve replica did not completely fit in the native annulus in many cases (in all cases of bicuspid valves), and we focused on these discrepancies. Three guiding sutures and several additional sutures, reflecting our modification of the original instructions for use, were placed at the surgeon’s discretion after careful inspection of the annular geometry to prevent incomplete annular fitting after valve deployment. After parachuting the valve into the annulus, the delivery system was temporarily removed from the valve holder, and a 5-mm videoscope was inserted through the central hole of the valve holder for evaluation of the fit from the inside. The valve position at the left ventricular outflow tract, spatial relationship with the anterior leaflet of the mitral valve, and any loosening or displacement of the guiding sutures were carefully examined under direct vision. After the delivery system was reassembled, balloon expansion was performed with 4.5 or 5.0 atm for 10 seconds, as described in the instructions. After balloon expansion, the videoscope was reinserted, as described above, to check for adequate subannular expansion, correct prosthesis position, and any related abnormalities. After confirmation, the guiding sutures were tied, and the aortotomy was repaired using a typical double-layer technique (or replaced with a graft in cases of concomitant ascending aorta replacement).

Evaluation of early clinical outcomes

Operative mortality was defined as any death within 30 days after surgery or during the same hospital admission. Continuous electrocardiography monitoring was applied to all patients until discharge, and the detection of any short runs of atrial fibrillation was regarded as an occurrence of postoperative atrial fibrillation. Low cardiac output was defined as a cardiac index <2.0 L/min/m² or a systolic arterial pressure <90 mm Hg requiring inotropic support (dopamine or dobutamine) of >5 mg/kg/min or mechanical circulatory support (e.g., intra-aortic balloon pump). Acute kidney injury was defined as a 2-fold increase in the serum creatinine level from the preoperative value, a 50% decrease in the glomerular filtration rate, urine output <0.5 mL/kg/hr for 12 hours, or the need for renal replacement therapy regardless of serum creatinine level. Respiratory complications included prolonged ventilation over 48 hours postoperatively, pneumonia, or the need for tracheostomy.

Evaluation of early hemodynamic outcomes

Early postoperative echocardiography was performed in 99.2% (250/252) of the study patients at a median of 6 days (interquartile range [IQR], 5–7 days) after surgery, except for a few mortality cases. The degree of paravalvular leak (PVL) was evaluated according to the guidelines. A mild or greater degree of PVL was considered clinically significant [13]. The echocardiographic parameters of the prosthetic valves included the transvalvular mean pressure gradient (PG) and effective orifice area (EOA). The measurements were performed according to the recommendations for imaging assessments of prosthetic heart valves [14]. The transvalvular mean PG, EOA, and EOA index were compared between the groups stratified by prosthesis size. Follow-up outcomes regarding PVL were also evaluated at postoperative 1 year.

Statistical analysis

Statistical analysis was performed using IBM SPSS ver. 25.0 (IBM Corp., Armonk, NY, USA) and SAS ver. 9.4 (SAS Institute Inc., Cary, NC, USA). Continuous variables are presented as the mean±standard deviation for normally distributed data or median with IQR for data that were not normally distributed, whereas categorical variables are presented as the number and percentage of the subjects. Comparisons of baseline demographics, operative data, early clinical outcomes, and early hemodynamic outcomes between the 2 groups were performed using the chi-square test or the Fisher exact test for categorical variables, and the Student t-test or Mann-Whitney test for continuous variables, as appropriate. Subgroup analysis was performed between type 0 versus type 1 or 2 bicuspid valves in the BAV group. All tests were 2-tailed, and a p-value <0.05 was considered statistically significant.

Preoperative characteristics

There were fewer female patients in the BAV group (36.1% versus 57.1%, p=0.001), and the mean age of the patients was approximately 9 years younger in the BAV group (65.5±9.9 versus 74.5±5.9 years, p<0.001). Risk factors were more prevalent in the TAV group than in the BAV group, including diabetes mellitus (19.7% versus 33.3%, p=0.01), hypertension (46.3% versus 77.1%, p<0.001), chronic kidney disease (12.2% versus 28.6%, p=0.001), and coronary artery disease (15.0% versus 31.4%, p=0.002). In the TAV group, the most common etiology of aortic valve disease was degenerative calcific disease, in 95.2% (100 out of 105) of patients, whereas 4.8% (5 out of 105) of the patients had rheumatic disease (Table 1).

Table 1 . Preoperative characteristics and risk factors.

Characteristic	BAV group (n=147)	TAV group (n=105)	p-value
Sex, female	53 (36.1)	60 (57.1)	0.001
Age (yr)	65.5±9.9	74.5±5.9	<0.001
Body mass index (kg/m2)	24.3±3.2	24.2±3.5	0.86
Body surface area	1.70±0.18	1.62±0.19	<0.001
Risk factors
Diabetes mellitus	29 (19.7)	35 (33.3)	0.01
Hypertension	68 (46.3)	81 (77.1)	<0.001
Dyslipidemia	69 (46.9)	57 (54.3)	0.25
Chronic obstructive pulmonary disease	9 (6.1)	9 (8.6)	0.46
Stroke	10 (6.8)	14 (13.3)	0.08
Chronic kidney disease	18 (12.2)	30 (28.6)	0.001
Renal replacement therapy	3 (2.0)	7 (6.7)	0.10
Coronary artery disease	22 (15.0)	33 (31.4)	0.002
Peripheral arterial occlusive disease	6 (4.1)	7 (6.7)	0.36
Atrial fibrillation	20 (13.6)	10 (9.5)	0.32
Reoperation	5 (3.4)	2 (0.3)	0.70
Left ventricular ejection fraction <35%	7 (4.8)	5 (4.8)	>0.99
EuroSCORE II	2.54±2.60	3.08±3.45	0.16
New York Heart Association class			0.36
I	32 (21.8)	21 (20.0)
II	99 (63.3)	59 (56.2)
III	19 (12.9)	22 (21.0)
IV	3 (2.0)	3 (2.9)
Etiology
Degenerative	0	100 (95.2)
Bicuspid	147 (100.0)	0
Type 0	42 (28.6)	0
Type 1	100 (68.0)	0
Type 2	5 (3.4)	0
Rheumatic	0	5 (4.8)
Emergency operation	2 (1.4)	0	0.51

Values are presented as mean±standard deviation for continuous variables or number (%) for categorical variables..

BAV, bicuspid aortic valve; TAV, tricuspid aortic valve..

Operative data

In the BAV group, a prosthesis size of 23 mm was the most frequently implanted (28.6%), followed by 21-mm and 25-mm valves (23.8% and 23.8%, respectively). However, in the TAV group, a prosthesis size of 21 mm was the most frequently implanted (41.0%), followed by 19-mm and 23-mm valves (22.9% and 21.0%, respectively).

Isolated AVR was performed in 22.4% and 41.0% of patients in the BAV and TAV groups, respectively (p=0.002). In the overall population, concomitant procedures included mitral valve surgery (21.4%), tricuspid valve surgery (7.5%), coronary artery bypass grafting (5.2%), arrhythmia surgery (8.7%), and aorta surgery (47.2%). Concomitant procedures were more frequently performed in the BAV group (77.6% versus 59.0%, p=0.002), and this finding was primarily attributed to concomitant aorta surgery (63.3% versus 24.8%, p<0.001). The number of stitches for AVR was larger in the BAV group than in the TAV group (9 [IQR, 7–11] versus 9 [IQR, 5–10], p=0.024) (Supplementary Table 1).

Early clinical outcomes

The operative mortality rate was 3.2% (8 out of 252), and there was no significant difference between the groups (2.0% versus 4.8% in BAVs versus TAVs, respectively). Common postoperative complications included postoperative atrial fibrillation (39.3%), acute kidney injury (12.3%), and respiratory complications (10.7%). There were no significant differences in the incidence of postoperative complications including permanent pacemaker implantation (0.7% versus 1.9% in BAVs versus TAVs, p=0.57) between the groups (Table 2).

Table 2 . Early clinical outcomes.

Variable	BAV group (n=147)	TAV group (n=105)	p-value
Operative mortality	3 (2.0)	5 (4.8)	0.28
Postoperative complication
Postoperative atrial fibrillation	61 (41.5)	38 (36.2)	0.40
Low cardiac output	4 (2.7)	6 (5.7)	0.33
Permanent pacemaker implantation	1 (0.7)	2 (1.9)	0.57
Acute kidney injury	16 (10.9)	15 (14.3)	0.42
Bleeding reoperation	4 (2.7)	8 (7.6)	0.13
Stroke	4 (2.7)	2 (1.9)	>0.99
Respiratory complication	15 (10.2)	12 (11.4)	0.76
Mediastinitis	1 (0.7)	1 (1.0)	>0.99
Infective endocarditis	0	0	-

Values are presented as number (%)..

BAV, bicuspid aortic valve; TAV, tricuspid aortic valve..

Early hemodynamic outcomes

In the BAV group, no PVL was detected in 87.7% of the patients, whereas trivial or higher PVL (any PVL) and mild or greater PVL were detected in 12.3% and 5.5% of the patients, respectively. In the TAV group, no PVL was detected in 92.3% of the patients, whereas trivial or higher PVL and mild or greater PVL were detected in 7.7% and 1.0% of the patients (p=0.23 and p=0.08), respectively. No patient showed greater than mild PVL in either group. At 1-year follow-up echocardiography, the proportions of patients with trivial or higher PVL increased slightly in both groups. However, no patient still showed greater than mild PVL in either group, and no significant intergroup difference regarding the degree of PVL was observed (Table 3).

Table 3 . Early and 1-year echocardiographic outcomes regarding paravalvular leak.

Variable	BAV group	TAV group	p-value
Early echocardiography	146	104
No PVL	128 (87.7)	96 (92.3)	0.23
PVL ≥trivial	18 (12.3)	8 (7.7)	0.23
PVL ≥mild	8 (5.5)	1 (1.0)	0.08
PVL >mild	0	0	>0.99
One-year echocardiography	138	98
No PVL	118 (85.5)	84 (85.7)	0.96
PVL ≥trivial	20 (14.5)	14 (14.3)	0.96
PVL ≥mild	16 (11.6)	8 (8.2)	0.39
PVL >mild	0	0	>0.99

Values are presented as number or number (%)..

BAV, bicuspid aortic valve; TAV, tricuspid aortic valve; PVL, paravalvular leak..

Early hemodynamic performances, including mean PG, EOA, and EOA index, did not significantly differ between the BAV and TAV groups when stratified by the prosthesis size (Fig. 2, Supplementary Table 2).

Figure 2. Comparison of early hemodynamic performance between the groups stratified by prosthesis size. Comparison of transvalvular mean pressure gradient (PG) (A), effective orifice area (EOA) (B), and EOA index (C).

Subgroup analysis for bicuspid aortic valve subtypes

A subgroup analysis was performed between type 0 and type 1 or 2 bicuspid valves in the BAV group. Regarding baseline characteristics, there were more female patients in the type 0 subgroup (52.4% versus 29.5%, p=0.009), whereas the mean age of the patients and the risk factors did not significantly differ between the subgroups. According to the Sievers classification, there were 42 patients (28.6%) with type 0 bicuspid valves, 100 patients (68.0%) with type 1 bicuspid valves, and 5 patients (3.4%) with type 2 bicuspid valves (Supplementary Table 3). In the operative data, 21-mm and 23-mm valves were frequently implanted in the type 0 subgroup, whereas 23-mm and 25-mm valves were frequently implanted in the type 1 or 2 subgroup. Concomitant aorta surgery was more frequently performed in the type 0 subgroup (85.7% versus 54.3%, p<0.001) (Supplementary Table 4). For early clinical outcomes, the operative mortality rate was 2.0% (3 out of 147), and postoperative complications did not significantly differ between the subgroups (Supplementary Fig. 1, Supplementary Table 5).

No PVL was detected in 85.7% and 88.5% of the patients in the type 0 and type 1 or 2 subgroups, respectively. Trivial or higher PVL and mild or greater PVL were detected in 14.3% and 2.4% of the patients in the type 0 subgroup, respectively, and in 11.5% and 6.7% of the patients in the type 1 or 2 subgroups (p=0.65 and p=0.44), respectively. There were no patients with greater than mild PVL in either subgroup. At 1-year follow-up echocardiography, the proportions of patients with trivial or higher PVL increased slightly in both groups. However, no patient still showed greater than mild PVL in either group, and no significant intergroup difference regarding the degree of PVL was observed (Table 4). Early hemodynamic performances, including mean PG, EOA, and EOA index, demonstrated no significant differences between the subgroups (Supplementary Table 6).

Table 4 . Early and 1-year echocardiographic outcomes regarding paravalvular leak in the subgroups of bicuspid aortic valve patients.

Variable	Type 0	Type 1 or 2	p-value
Early echocardiography	42	105
No PVL	36 (85.7)	92 (88.5)	0.65
PVL ≥trivial	6 (14.3)	12 (11.5)	0.65
PVL ≥mild	1 (2.4)	7 (6.7)	0.44
PVL >mild	0	0	>0.99
One-year echocardiography	40	98
No PVL	32 (80.0)	86 (87.8)	0.24
PVL ≥trivial	8 (20.0)	12 (12.2)	0.24
PVL ≥mild	7 (17.5)	9 (9.2)	0.17
PVL >mild	0	0	>0.99

Values are presented as number or number (%)..

PVL, paravalvular leak..

When the BAV patients were grouped into type 0 and 2 BAVs together and compared to the patients with type 1 BAVs, the incidence of PVL also showed no significant differences between the groups (Supplementary Table 7).

The present study demonstrated 3 main findings. First, the incidence of PVL after RDAVR on early postoperative echocardiograms demonstrated no significant differences between the BAV and TAV groups. Second, the early clinical outcomes including operative mortality, morbidities, and the incidence of permanent pacemaker implantation were also comparable between the groups. Third, in the subgroup analysis of BAVs, early clinical outcomes and the incidence of PVL showed no significant differences between type 0 and type 1 or 2 bicuspid valves (Fig. 3).

Figure 3. This study was conducted to compare early echocardiographic and clinical outcomes after rapid-deployment aortic valve replacement (AVR) for bicuspid versus tricuspid aortic valves. Rapid-deployment AVR for bicuspid aortic valves demonstrated comparable early echocardiographic and clinical outcomes to those for tricuspid aortic valves, and the outcomes were also satisfactory for type 0 bicuspid aortic valves. Values are presented as number (%).

In patients with BAVs, the use of RD valves is not preferred, or even contraindicated, by many surgeons due to concerns that uneven alignment of the cusps and aortic root asymmetry in BAVs may result in PVL [15]. The SURD-IR (Sutureless and Rapid-Deployment Aortic Valve Replacement International Registry), one of the largest prospective registries for RDAVR, reported that only 4.1% of the patients had BAVs, which implies that many surgeons remain concerned and hesitate to implant RD valves in BAVs [15,16]. The CADENCE-MIS trial, a prospective randomized multicenter trial, included patients with Sievers type 1 BAVs, but excluded those with type 0 BAVs in the study design [17]. The TRITON trial, a European prospective multicenter trial that reported excellent safety and hemodynamic performance of RD valves for up to 5 years, reported that congenital BAVs were an exclusion criterion of the trial [1,18].

Despite concerns about RDAVR for BAVs, convincing outcomes have been reported in several publications. A single-center retrospective study demonstrated that 107 BAV patients received RD valves with acceptable implant success rates and long-term outcomes [19]. However, a trend toward an increased frequency of moderate–severe paravalvular regurgitation was observed at long-term follow-up. Furthermore, only 9 patients with type 0 BAVs were included in this study. Another study enrolled 191 patients with BAVs from a prospective multicenter registry and demonstrated that early outcomes and hemodynamic performances were excellent [15]. However, the outcomes did not discriminate between sutureless and RD valves, and the Sievers classification of BAVs was not detailed in the report.

In our institution, a single surgeon (K.H.K.) used Edwards Intuity exclusively for patients who planned to receive a bioprosthetic valve, regardless of the etiology of aortic valve disease [10]. The fact that BAVs were present in 52.6% of the entire cohort and that 11.2% of the study population had type 0 BAVs demonstrates that our series was indeed a real-world all-comers setting, because more than 50% of all AVRs for aortic stenosis are estimated to be due to BAVs according to a previous report [20]. Our series of this all-comers population is unique and invaluable for evaluating the feasibility of RDAVR in BAVs.

According to the instructions for the use of RD valves, the manufacturer recommends placing only 3 anchoring sutures. In BAVs, however, 3 anchoring sutures placed at the nadir of each sinus are usually not equally distributed in the annular circumference because the widths of the coronary sinuses are different in BAVs. The discrepancy between the native annulus and the sewing ring of the RD valve would be maximized in cases of type 0 BAVs. Furthermore, we experienced some patients, even with TAVs, whose aortic annulus did not perfectly fit with the sewing ring of the RD valve when only 3 anchoring sutures were used and the discrepancy persisted. To overcome this discrepancy, we adopted a technical modification of placing more than 3 additional anchoring sutures, which enables complete fitting of the native annulus to the sewing cuff of the RD valve (Fig. 4) [12]. Based on this modification, the procedures used for BAVs, even type 0 BAVs, have been uneventful with satisfactory results. By investing several extra minutes in additional anchoring sutures, we were able to achieve complete annulus fitting and substantially reduce the occurrence of PVL. We also believe that this complete annulus fitting contributed to the lower incidence of pacemaker implantation in our series (1.4% before discharge and 2.3% during follow-up) [10].

Figure 4. A technical modification of placing more than 3 additional anchoring sutures which enables complete fitting of the native annulus to the sewing cuff of the rapid deployment (RD) valve. (A) A type 0 bicuspid aortic valve (BAV). (B) Excision of the native valve and decalcification of the native annulus was completed. (C) Because the shape of the annulus of a type 0 BAV and the sewing cuff of the RD valve have a substantial discrepancy, placing only 3 anchoring sutures would result in potential gap between them. Thus, several additional anchoring sutures were placed to achieve complete annulus fitting. (D) The RD valve was finally implanted to the type 0 BAVs. The 2 commissures of the type 0 BAVs are pointed with white arrows, one of which is not in accordance with the commissures of the tricuspid bioprosthesis.

In addition, checking the fitness of annulus to the sewing ring using a videoscope is not a standard method for surgical AVR. This procedure was initiated at our institution based on the belief that confirming the achievement of complete annulus fitting with direct vision is of paramount importance in the prevention of PVL and conduction abnormalities. Using a videoscope to confirm annulus fitting is a simple yet effective method, particularly for assessing areas that are difficult to visualize with the naked eye. We experienced several cases that were initially supposed to be well-deployed just after the landing. However, videoscopic examinations detected significant incomplete deployment of the RD valves, and re-deployment was required in these patients [12].

By these modifications from the original instructions for use, the purported benefit of RD valves—namely, reduced procedural times—might be significantly compromised. However, the authors believe that secure deployment with additional sutures is more important than rapid deployment per se because it would reduce the incidence of PVL and permanent pacemaker implantation.

Limitations

There are several limitations of this study that should be noted. First, this was a retrospective single-center study with a small sample size. Second, only early clinical and echocardiographic outcomes were evaluated in this study. Because the severity of PVL may worsen during follow-up, further investigation with longer-term follow-up would be needed. Third, our series included patients who underwent surgery when the operating surgeon had insufficient experience in the initial period of RDAVR at our institution. Most cases of significant PVL occurred in the early stage of the series, and PVL became very rare after getting past the learning curve. Fourth, although the patient characteristics of the BAV group were quite different from those of TAV group, we did not adopt analytic methodologies, including propensity score matching, to adjust these differences for fear that these methodologies would not properly represent the real-world population. The discrepancies in the prevalence of baseline comorbidities should be considered in the interpretation of the comparative clinical outcomes between the two groups. Fifth, a comparison in time economics between Intuity versus conventional bioprostheses was not performed in this study because only a few patients received a bioprosthetic valve other than Intuity during the study period by the surgeon. Sixth, the incidence of PVL after RDAVR in this study might be suboptimal compared with those after conventional AVR in previous studies, which requires further investigation.

RDAVR for BAVs demonstrated comparable early echocardiographic and clinical outcomes to those for TAVs, and the outcomes were also satisfactory for type 0 BAVs.

Author contributions

Conceptualization: KHK, SHS. Data curation: KHK, SHS, YK, JSK, JWC. Formal analysis: SI, SHS. Methodology: KHK, SHS. Project administration: SI. Visualization: SI. Writing–original draft: SI, SHS. Writing–review & editing: KHK. Final approval of the manuscript: all authors.

Conflict of interest

Kyung Hwan Kim is an official proctor for Edwards Lifesciences. Jae Woong Choi is an editorial board member of the journal but was not involved in the peer reviewer selection, evaluation, or decision process of this article. No other potential conflict of interest relevant to this article was reported.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Supplementary materials

Supplementary materials can be found via https://doi.org/10.5090/jcs.23.070. Supplementary Fig. 1. Comparison of early hemodynamic performance between the subgroups of bicuspid valve patients stratified by prosthesis size. Supplementary Table 1. Operative data. Supplementary Table 2. Comparison of early hemodynamic performance between the groups. Supplementary Table 3. Preoperative characteristics and risk factors for the subgroups of bicuspid valve patients. Supplementary Table 4. Operative data in the subgroups of bicuspid aortic valve patients. Supplementary Table 5. Early clinical outcomes in the subgroup of bicuspid aortic valve patients. Supplementary Table 6. Comparison of early hemodynamic performance between the subgroups of bicuspid aortic valve patients. Supplementary Table 7. Early and 1-year echocardiographic outcomes regarding paravalvular leak in the subgroups of bicuspid aortic valve patients.