Author + information
- Received November 20, 2015
- Revision received January 4, 2016
- Accepted January 4, 2016
- Published online April 25, 2016.
- Gidon Y. Perlman, MDa,
- Philipp Blanke, MDa,
- Danny Dvir, MDa,
- Gregor Pache, MDb,
- Thomas Modine, MDc,
- Marco Barbanti, MDd,
- Erik W. Holy, MDe,
- Hendrik Treede, MDf,
- Philipp Ruile, MDb,
- Franz-Josef Neumann, MDb,
- Caterina Gandolfo, MDg,
- Francesco Saia, MDh,
- Corrado Tamburino, MDd,
- George Mak, MDa,
- Christopher Thompson, MDa,
- David Wood, MDa,
- Jonathon Leipsic, MDa and
- John G. Webb, MDa,∗ ()
- aDepartment of Cardiology, St. Paul’s Hospital, Vancouver, British Columbia, Canada
- bDepartments of Radiology and Cardiology, University Heart Center Freiburg-Bad Krozingen, Germany
- cDepartment of Cardiovascular Surgery, Hôpital Cardiologique, Lille, France
- dDepartment of Cardiology, Ferrarotto Hospital, University of Catania, Catania, Italy
- eCardiology, University Heart Center, University Hospital Zurich, Zurich, Switzerland
- fDepartment of Cardiovascular Surgery, University Heart Center, Hamburg, Germany
- gOspedale Civico, Palermo, Italy
- hCardiovascular Department, Institute of Cardiology, University of Bologna, Policlinico St. Orsola-Malpighi, Bologna, Italy
- ↵∗Reprint requests and correspondence:
Dr. John Webb, St. Paul’s Hospital, 1081 Burrard Street, Vancouver, British Columbia V6Z 1Y6, Canada.
Objectives This study evaluated the results of transcatheter aortic valve replacement (TAVR) in bicuspid aortic stenosis (AS) using a new-generation TAVR device.
Background A bicuspid AS is often considered a relative contraindication to TAVR. Although initial reports have demonstrated feasibility using early-generation devices, outcomes have not matched those seen with tricuspid AS. Paravalvular aortic regurgitation (AR) has been particularly problematic.
Methods We collected baseline characteristics, procedural data, and 30-day clinical follow-up findings from 8 centers in Europe and Canada that had performed TAVR in bicuspid AS using the SAPIEN 3 valve.
Results 51 patients underwent TAVR using the SAPIEN 3 valve. Patient mean age was 76.2 ± 9.3 years and the Society of Thoracic Surgeons predicted risk of mortality scores were 5.2 ± 3.7%. Bicuspid valve types were: type 0, 11.8%; type 1, 82.3%; and type 2, 1.9%. There were no cases of valve embolization or need for a second valve. Post-dilation was performed in 7.8%. The mean aortic gradient decreased from 49.4 ± 16.0 mm Hg to 11.2 ± 4.7 mm Hg. Post-implantation AR was none/trivial in 63% and mild in 37%. There were no cases of moderate or severe AR. At 30-day follow-up, there were 2 deaths (3.9%), 2 major vascular complications, and 12 patients (23.5%) required pacemaker implantation.
Conclusions TAVR in bicuspid AS using a new-generation device was feasible and effective with favorable valve performance and no cases of moderate or severe AR.
A bicuspid aortic valve is a common congenital heart defect, affecting approximately 1% of the population (1,2). Stenosis of bicuspid valves typically occurs at a younger age than is the case with tricuspid aortic valves. Nevertheless, bicuspid valves remain a significant cause of symptomatic aortic stenosis (AS) in elderly patients considered for transcatheter aortic valve replacement (TAVR) or surgical aortic valve replacement (3).
Bicuspid anatomy has routinely been considered an exclusion in the setting of prospective TAVR trials. However, early reports have shown that TAVR may be a feasible option for bicuspid patients at high surgical risk. These initial reports used early-generation devices with somewhat disappointing outcomes, particularly with regard to paravalvular aortic regurgitation (AR) (4–7). With the predicted expansion of TAVR into lower-risk patient groups, the relative proportion of bicuspid AS is expected to rise, and therefore, there is a need to optimize results in this group of patients as well.
Bicuspid aortic valves differ from trileaflet, tricuspid, stenotic aortic valves, by having annular dimensions that are often larger and leaflets that are more calcified, bulky, and irregular (2,8,9). The aortic geometry and blood flow are also altered in bicuspid AS (10). These characteristics may impede precise location and full apposition of the device with the annulus (11), thus resulting in the relatively greater degree of AR associated with TAVR in the setting of bicuspid AS.
Newer-generation valves may offer advantages over earlier valves. The Edwards SAPIEN 3 valve (Edwards Lifesciences, Irvine, California) is a next-generation device, which incorporates an outer fabric seal designed to prevent paravalvular AR (12–14). This external seal has the potential to adapt better to the irregular annuli shapes and valve orifices of patients with bicuspid valves, thus reducing paravalvular AR and potentially improving outcomes with TAVR in this patient group (15).
This multicenter study aimed to assess the clinical outcomes and valve function associated with implantation of the SAPIEN 3 valve in patients with bicuspid aortic valve stenosis.
Patients and participating centers
Patient data were retrospectively collected from 8 centers in Canada, Germany, Italy, France, and Switzerland (Online Table 1). Patients were treated between June 2012 and July 2015, and were prospectively followed by the performing institutions. Patients were reviewed by respective institutional heart teams, and the study was approved by local institutional boards. All the patients were treated with SAPIEN 3 devices with diameters ranging from 23 to 29 mm. In accordance, vascular access was made through 14-F or 16-F expandable sheaths (eSheath, Edwards Lifesciences). Clinical data were collected and analyzed centrally. Pre-procedural computed tomography (CT) datasets were collected and analyzed in a core lab fashion at St. Paul's Hospital, Vancouver, Canada, by 2 experienced cardiac CT readers. Pre- and post-procedural echocardiographic results were reported by the participating institutions.
Electrocardiogram-gated, contrast-enhanced CT data acquisition was performed using contemporary CT systems and employing institutional CT scan protocols. All available CT datasets were evaluated by a single reader for anatomic quantification (P.B.) and by 2 readers in consensus to determine the valve morphology and extent of calcification (P.B. and J.L.). Datasets were transferred to a dedicated post-processing workstation equipped with Aquarius InTuition (version 188.8.131.52, TeraRecon, Foster City, California). Annular segmentation and annular dimensions were assessed by means of planimetry, yielding long-axis (maximum) and short-axis (minimum) diameters, cross-sectional area, and perimeter. A smoothing algorithm was used to avoid an artificial increase in perimeter by contour irregularities as described previously (16).
The aortic annulus ellipticity index was determined using the formula (maximal diameter/minimal diameter). Degree of oversizing of the device was calculated as: [(device area − annulus area)/annulus area] × 100.
Valve morphology was classified as previously described by Sievers et al. (17) according to the number of cusps and the presence of raphes, as well as spatial position and symmetry of raphes and cusps. Type 0 was assigned to morphologies characterized by the presence of 2 symmetric leaflets/cusps and 1 commissure without evidence of a raphe. Type 1 was assigned to valve morphologies with 1 raphe, and Type 2 when 2 raphes were present. Functional (acquired) bicuspid valves were classified as tricuspid valves with no raphe present but secondarily fused cups, that is, fixation of the commissure between 2 cups in a closed position due to degenerative processes.
To facilitate discrimination of functional bicuspid valves from Type 1 bicuspid valves, in particular to distinguish between a secondarily fused commissure and a raphe, further criteria were applied: A raphe does not extend to the same level on the aortic site as the free margins of the cusps that are forming true commissures. Secondly, diagnosis of a functional bicuspid valve requires symmetry of all 3 cusps, whereas asymmetry is commonly seen in Type 1 bicuspid valves (Figure 1).
The location of the raphe or acquired fusion was reported as R-L, R-N, or N-L, (R = right coronary; L = left coronary; N = noncoronary).
Endpoints and definitions
Procedural, 30-day mortality, and other major clinical endpoints were defined according to the updated Valve Academic Research Consortium (VARC-2) criteria (18). Post-implant AR was defined as the sum of transvalvular and paravalvular regurgitation. The severity of regurgitation was qualitatively assessed and graded using transthoracic echocardiography according to established guidelines (18,19). Aortic regurgitation was categorized as paravalvular, transvalvular, or mixed and was classified as none/trivial, mild, moderate, or severe.
Continuous variables are presented as mean ± SD and were compared using the Student t test, or paired t test for repeated measures. Categorical variables are presented as frequencies and percentages, and were compared using the chi-square or Fisher exact test. Statistical significance was defined as p < 0.05. Analyses were performed using GraphPad (GraphPad Software, La Jolla, California).
Fifty-one patients with bicuspid aortic valve stenosis were identified. Patients were treated in 8 centers located in 5 countries. The mean age was 76.2 ± 9.3 years, 52.9% women. Average Society of Thoracic Surgeons Predicted Risk of Mortality score was 5.2 ± 3.7%. The baseline clinical, echocardiographic, and CT data are summarized in Table 1.
TAVR was performed with the Edwards SAPIEN 3 valve by the femoral route in 49 patients (96.1%) and by the transcarotid route in 2 patients. The procedure was done with local anesthesia and conscious sedation in 20 patients (39.2%). Pre-implantation balloon valvuloplasty was performed in 52.9% of cases and post-dilation in 4 patients (7.8%). There were no intraprocedural deaths. There were no device malposition events requiring additional interventions. None of the patients required a second valve or conversion to surgical treatment (Table 2).
Transthoracic echocardiographic evaluation at 30 days is presented in Table 3. The mean aortic gradient was reduced from 49.4 ± 16.0 mm Hg to 11.2 ± 4.7 mm Hg, and the aortic valve area was increased from 0.7 ± 0.2 cm2 to 1.7 ± 0.3 cm2. AR was none or trivial in 32 (63%) patients and mild in 19 (37%) patients. There were no patients with moderate or severe AR identified in our cohort.
Patients with devices that were oversized more than 10%, relative to the annular area, had larger post-procedural valve areas (p = 0.01). We observed a 26.9% rate of mild AR when the device was moderately oversized by 10% of the annular area or more. Lesser degrees of oversizing, or even mild undersizing were associated with a 48% rate of mild AR (p = 0.10).
We noted in 38% of the patients a tendency of the valve to expand asymmetrically when viewed by angiography, with either the medial or lateral border of the stent frame being longer than the opposite side. An example of this is shown in Figure 2. We did not note a clinical correlate to this phenomenon; the rates of AR or need for a pacemaker was not higher in patients with a degree of asymmetry >7%.
Table 4 presents the clinical outcomes of the patients in this study. At 30 days, there were 2 deaths. One patient died on day 14 after a procedure-related tamponade that led to gradual deterioration. The second patient died from aspiration following general deterioration on day 27 (second admission). Major or life-threatening bleeding occurred in 5 patients (9.8%). Of these, 3 bleeding events were directly related to the procedure; 1 case of pericardial bleeding and 2 access site hematomas requiring transfusion. Two bleeding events were not directly related to the procedure: bleeding from a gastric ulcer and from a bladder tumor that required transfusions. A new pacemaker was implanted in 12 patients (23.5%) due to: complete heart block in 5 patients, second-degree heart block in 2, alternating right and left bundle branch block in 2, and new bundle branch block associated with other new conduction delays in 3. The need for a new pacemaker was associated with a low implantation position of the valve frame (defined as >8 mm below the annulus level). The rate was 55% in these low positions compared with 10% in the rest of the cohort (p = 0.01) (Figures 3 and 4).
In this study, we evaluated the treatment of patients with stenotic bicuspid aortic valves using a next-generation transcatheter valve, the Edwards SAPIEN 3. To the best of our knowledge, this is the largest report of bicuspid patients treated with the SAPIEN 3 valve, and the only report of patients treated exclusively with this device. This experience demonstrates feasibility with favorable valve performance and minimal paravalvular leaks.
Prior TAVR experience with early-generation balloon-expandable and self-expanding valves reported higher than usual rates of paravalvular leaks (Table 5). In a recent report by Mylotte et al. (4), a 28.4% rate of grade 2 or more AR was observed in a large series of 139 patients. In our cohort of patients, we did not observe clinically significant (>mild) paravalvular leaks in any patient. The low rate of paravalvular leaks may reasonably be attributed to the improved sealing properties of the external sealing layer of the inflow portion of the SAPIEN 3 valve and improvements in accuracy of valve positioning with an improved valve delivery system. Importantly, CT annular sizing was routinely used to select an appropriately sized implant, this has been shown to be associated with a reduction in AR in patients with bicuspid valves (4). Patients with devices that were oversized >10% of the annular area had a particularly low rate of AR, consistent with previous reports in tricuspid patients (14,20). Importantly, moderate oversizing guided by CT annular measurements appeared safe and was not associated with annular rupture.
The technical success of the valve implantation process was very high in our cohort, despite the adverse anatomic characteristics typical of bicuspid patients. The device success rate of 98% reported here is comparable to that reported with the same next-generation valve in patients with tricuspid aortic valves (12,13). Only 4 patients (7.6%) required post-implantation balloon dilation due to residual paravalvular regurgitation, a low rate compared with other recently reported bicuspid series where post-dilation was required in 18% to 30% of patients (4,21,22). This high rate of success might be attributable to the low profile and improved accuracy of valve positioning with the SAPIEN 3 delivery system, the improved sealing with the outer skirt, as well as increased operator experience and case planning. The mechanism for the asymmetric expansion of some valves that was noted in this study is not known at this stage, and it is unclear whether this is unique to bicuspid patients, whether this occurs with other devices, and whether this has long-term clinical relevance.
Previous reports of bicuspid AS have also included patients treated with self-expanding valves, unlike our study. Self-expanding valves might be more capable of conforming to the irregular orifice of the bicuspid valve but less capable of achieving a circular formation after implantation (8,23). Studies of bicuspid patients that compared results with self-expanding versus balloon-expandable valves did not find conclusive evidence for an advantage of a certain type of valve in preventing AR or improving clinical outcomes (4,8,24). Admittedly, these studies were all underpowered to assess these differences between valve types. Our study of patients treated exclusively with a new balloon-expanded valve shows excellent sealing properties and no evidence of aortic injury, which has become a rare event with modern pre-procedural imaging (13).
A recent analysis of outcomes in bicuspid patients noted a higher device success rate and lower incidence of significant AR in patients with type 1L-R bicuspid valves (fusion of the left and right cusps) (21). Type 1L-R was by far the most common form of bicuspidity in our cohort, possibly contributing to the excellent device performance reported. The mechanism for this association is not clear.
New pacemaker implantation was performed in 12 patients (23.5%), a rate higher than that generally observed with balloon-expandable devices in patients with tricuspid aortic valve stenosis (25). Prior reports with earlier balloon-expandable and self-expanding devices in bicuspid patients also documented relatively high pacemaker implantation rates, ranging from 14% to 50% (4,7,21,22,24). The reasons for the high pacemaker rates following TAVR in bicuspid patients are not clear. More aggressive oversizing or lower implantation resulting from the irregular anatomy of bulky bicuspid valves might be associated with more compression of the conduction system, but this remains an area of uncertainty. In our cohort, a greater degree of oversizing (>10%) was not associated with a higher rate of pacemaker implantation (26.9% vs. 20.0%; p = 0.74) Recent experience with the SAPIEN 3 valve in patients with tricuspid aortic valves suggests that pacemaker rates may be higher than seen with earlier-generation balloon-expandable valves and similar to the rates observed in our study of bicuspid valves (13,26,27). It has been suggested this is largely due to the greater length of the SAPIEN 3 valve combined with positioning recommendations that led to low implantation positions (13,26). We observed a similar higher rate of need for a new pacemaker in patients with a very low implantation position. Newer sizing and positioning recommendations may lead to lower pacemaker rates, although this is unproven (28).
Other rates of complications in our cohort were not unexpected from the population of TAVR patients and bicuspid patients (Table 4), including 5 (9.8%) patients with major or life-threatening bleeding events. The bleeding rates we report are not unique to the bicuspid subset of patients, and the small size of our cohort cannot assess these thoroughly. There were 2 major vascular complications in our cohort; this low rate is probably due to the low-profile sheaths used for SAPIEN 3 valves, which are associated with less vascular trauma.
As TAVR is being offered to increasingly younger and lower risk populations, the proportion of patients with bicuspid valves may be expected to rise. Our results show that using a next-generation device, the SAPIEN 3 valve, can overcome the main limitations of TAVR in bicuspid patients and offer a viable treatment option with results comparable to those seen in patients with tricuspid aortic valves.
The main limitation of our study is its retrospective nature, necessitated by the routine exclusion of bicuspid patients from prospective studies. The relative weight of each center in a small study might introduce variability in the results, which is not generalizable. However, short-term device performance was excellent despite the potential variability of procedural and imaging techniques and expertise amongst centers. We did not include a control group of bicuspid patients treated with previous generation valves to avoid a bias stemming from older techniques and earlier expertise. The size of our cohort makes it underpowered for the identification of subgroups of patients (e.g., bicuspid types) with different outcomes.
Treating patients with stenotic bicuspid aortic valves with a next-generation balloon-expandable valve is feasible with promising early clinical and hemodynamic results.
WHAT IS KNOWN? Bicuspid aortic valve stenosis is often considered a relative contraindication to transcatheter aortic valve implantation. Initial reports have shown feasibility, but higher rates of paravalvular regurgitation than observed for tricuspid aortic valves.
WHAT IS NEW? Implantation of a new-generation device was associated with minimal paravalvular regurgitation and good clinical outcomes.
WHAT IS NEXT? Rates of pacemaker implantation after TAVR in bicuspid AS were relatively high and require further study to understand the mechanism.
For a list of the participating centers and Online Table 1, please see the online version of this article.
Drs. Blanke, Dvir, Pache, Barbanti, and Webb are consultants for Edwards Lifesciences. Dr. Neumann's institution has received research grants, speaker honoraria, and travel support from Edwards Lifesciences. Dr. Gandolfo is a proctor for Edwards Lifesciences. Dr. Saia has received consulting fees from Abbott Vascular, Eli Lilly and Company, AstraZeneca, The Medicines Company, St. Jude Medical, and Medtronic; and speaker’s fees from Abbott Vascular, Eli Lilly and Company, AstraZeneca, St. Jude Medical, Terumo, Biosensors, Edwards Lifesciences, Sorin, and Boston Scientific. Dr. Tamburino has received honoraria from Abbott, Medtronic, and St. Jude Medical. Dr. Thompson has received travel support from Edwards Lifesciences. Dr. Wood is a consultant for Edwards Lifesciences; and has received clinical trials/grant support from Edwards Lifesciences. Dr. Leipsic is a consultant and provides the core laboratory for Edwards Lifesciences. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- aortic regurgitation
- aortic stenosis
- computed tomography
- transcatheter aortic valve implantation
- Received November 20, 2015.
- Revision received January 4, 2016.
- Accepted January 4, 2016.
- 2016 American College of Cardiology Foundation
- Steinberger J.,
- Moller J.H.,
- Berry J.M.,
- Sinaiko A.R.
- Siu S.C.,
- Silversides C.K.
- Roberts W.,
- Janning K.,
- Ko J.,
- Filardo G.,
- Matter G.
- Mylotte D.,
- Lefevre T.,
- Søndergaard L.,
- et al.
- Wijesinghe N.,
- Ye J.,
- Rodés-Cabau J.,
- et al.
- Mahadevia R.,
- Barker A.J.,
- Schnell S.,
- et al.
- Binder R.K.,
- Rodés-Cabau J.,
- Wood D.A.,
- et al.
- Webb J.,
- Gerosa G.,
- Lefèvre T.,
- et al.
- Yang T.H.,
- Webb J.G.,
- Blanke P.,
- et al.
- Kappetein A.P.,
- Head S.J.,
- Généreux P.,
- et al.
- Zoghbi W.A.,
- Chambers J.B.,
- Dumesnil J.G.,
- et al.
- Willson A.B.,
- Webb J.G.,
- Labounty T.M.,
- et al.
- Hayashida K.,
- Bouvier E.,
- Lefèvre T.,
- et al.
- Schymik G.,
- Lefèvre T.,
- Bartorelli A.L.,
- et al.
- Dvir D.,
- Walther T.,
- Wood D.,
- et al.