Circulation: Cardiovascular Interventions

ORIGINAL ARTICLE
Bicuspid Aortic Valve Anatomy and Relationship
With Devices: The BAVARD Multicenter Registry
A European Picture of Contemporary Multidetector Computed
Tomography Sizing for Bicuspid Valves

BACKGROUND: Sizing for transcatheter aortic valve implantation in Didier Tchetche, MD*
bicuspid aortic valves (BAV) remains controversial. Chiara de Biase, MD*
et al
METHODS AND RESULTS: The aim of the BAVARD (Bicuspid Aortic
Valve Anatomy and Relationship With Devices) retrospective registry is to
capture the sizing ratios used for transcatheter aortic valve implantation
in BAV and analyze the second-generation prostheses geometry
postimplantation. About 101 patients with BAV along with available pre-
and post-transcatheter aortic valve implantation multidetector computed
tomography were compared with 88 tricuspid aortic valves (TAV) patients.
Preprocedural multidetector computed tomography diagnosed type 0
and type 1 BAV in, respectively, 12.9% and 86.1 % of BAV. At baseline,
the ellipticity index was similar between BAV and TAV patients: 1.2±0.1
Downloaded from http://ahajournals.org by on December 9, 2019

versus 1.2±0.1, P=0.09. The mean annular oversizing was, respectively,

1.14±0.04 and 1.04±0.04, P<0.001, in TAV and BAV patients. The mean
prosthesis intercommissural distance, ratio was 1.03±0.1. The mean
diameter of the prostheses at the annulus matched the mean perimeter-
derived diameter of the aortic annulus at baseline with TAV (23.3±2.2
versus 23.6±1.9, P=0.4) and was smaller with BAV (24±2.8 versus
26.8±3.1, P<0.01), confirming 11% underexpansion in BAV. Finally, in
situ, prosthesis diameter and ellipticity followed the same pattern, with
stable values from the distal edge to 12 mm above, in both groups.
CONCLUSIONS: Second-generation prostheses similarly reshape the
aortic annulus in TAV and BAV. Prostheses keep consistent diameters
from distal edge to 12 mm in TAV and BAV. Prosthesis underexpansion
is constantly observed in BAV. Annular-based sizing is accurate in BAV
with minimal oversizing. The intercommissural distance, 4 mm above the
annulus, could be integrated in gray zones.
CLINICAL TRIAL REGISTRATION: URL: https://www.clinicaltrials.gov. *Drs Tchetche and de Biase contributed
Unique identifier: NCT03495050. equally to this work.

The full author list is available on page 9.

VISUAL OVERVIEW: A visual overview is available for this article.
Key Words:  aortic valve ◼ bicuspid
◼ heart valve diseases ◼ multidetector
computed tomography ◼ tricuspid valve

© 2019 American Heart Association, Inc.

https://www.ahajournals.org/journal/
circinterventions

Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107 January 2019 1

 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients
WHAT IS KNOWN
• Bicuspid aortic valves (BAV) remain a challenge for
transcatheter aortic valve implantation.
• There is an ongoing debate about the appropriate
sizing methodology for BAV before transcatheter
aortic valve implantation.
• As there is a continuous decrease in the risk profile
of the patients treated with transcatheter aortic
valve implantation, including younger patients, we
will be confronted by more and more BAV patients.
WHAT THE STUDY ADDS
• The BAVARD (Bicuspid Aortic Valve Anatomy and
Relationship With Devices) registry, apart from
confirming good clinical outcomes with current
second-generation transcatheter aortic valve im-
plantation devices, provides more understanding
about the contemporary sizing methodology and
ratios utilized in these patients.
• It could be one of the first steps for harmonization
of our practices in BAV configurations and repre-
sent a basis for future prospective registries.
T
ranscatheter aortic valve implantation (TAVI) is
an established treatment for intermediate to
high risk patients with symptomatic aortic ste-
nosis (AS).1–3 Bicuspid aortic valves (BAV) are the most
Downloaded from http://ahajournals.org by on December 9, 2019
common congenital valvular abnormality occurring in
0.5 % to 2% of the general population, in 2% to 6%
of patients with severe AS and up to 20% of octo-
nonagenarians undergoing surgery.4 The presence of
BAV has regularly been considered a contraindica-
tion to TAVI because of a high risk of malpositioning,
moderate-severe aortic regurgitation and potentially
accelerated leaflet degeneration.5 BAV may combine
a large annulus, heavily calcified leaflets, and dilated
ascending aorta.6,7 These anatomic features can neg-
atively impact the optimal interaction between the
native valve and the TAVI prosthesis leading to high
rates of paravalvular regurgitation, device underex-
pansion, and the need for a second THV.8 The dilated
aorta may predispose to aortic dissection.9 These find-
ings were mainly related to TAVI with first-generation
devices. New-generation prostheses offer better posi-
tioning and alignment, repositionability, and sealing
properties with more accurate deployment and low
paravalvular leak rate in tricuspid aortic valves (TAV).10
Recently published retrospective registries demon-
strated the safety of TAVI in BAV, using second-gen-
eration prostheses, with clinical outcomes comparable
to TAVI in tricuspid valves.8,11–13 However, despite en-
couraging clinical outcomes, sizing for BAV remains
controversial and debatable. Multidetector computed
tomography (MDCT) is the preferred sizing modality
Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107
for TAVI, including BAV.14–18 Current sizing practices
in BAV patients employ 2 different methodologies:
a standard annular-based sizing and a supra-annular
sizing, at the level of the leaflets or the commissures.
The aim of our retrospective registry is to capture the
sizing ratios used in European and Israeli centers for
BAV patients undergoing TAVI with second-gener-
ation prostheses, when using MDCT as the imaging
modality, and also to analyze by postprocedural MDCT
the prostheses geometry, in situ.
METHODS
The data, analytic methods, and study materials will not be
made available to other researchers for purposes of repro-
ducing the results or replicating the procedure. However, the
methodology for MDCT analysis is described precisely.
We conducted a multicenter retrospective registry in-
cluding all symptomatic patients treated with TAVI for BAV,
using second-generation prostheses, at 8 European and Israeli
centers between January 2016 and December 2017. In total,
184 consecutive BAV patients were identified. Data were ret-
rospectively collected in a dedicated database. MDCT assess-
ment, pre- and post-TAVI, was based on Heart Teams’ decision.
About 101 patients with BAV along with available pre- and
post-TAVI MDCT were included in the current analysis. These
patients were compared with 88 TAV patients, enrolled at 2
centers, in the same period, with available pre- and post-TAVI
MDCT. All patients provided consent for anonymized data ac-
quisition and analysis. The registry was approved by each local
institutional ethical committee. Prosthesis type and size were
left at the discretion of local Heart Teams.
Clinical Outcomes
Clinical outcomes at 30 days were analyzed according to the
Valve Academic Research Consortium-2 criteria.19
Baseline MDCT Analysis
All baseline and post-TAVI ECG-gated MDCT data were ana-
lyzed using the 3Mensio Structural Heart software version
8.0 (Pie Medical Imaging, Maastricht, the Netherlands). Mid-
systolic phases (30%–40% of the R-R interval) were consid-
ered for pre- and post-TAVI reconstructions. MDCT Digital
Imaging and Communications in Medicine data were anony-
mized and analyzed in a corelab-like setting by 3 experienced
specialists. BAV type was characterized according to Sievers et
al.20 Type 0 has no raphe; type 1 has 1 raphe, and type 2 has
2 raphes (Figure 1). The aortic annulus was determined by the
virtual basal ring linking the hinge points of the aortic leaflets.
Annular dimensions were assessed in TAV and BAV patients,
as was the intercommissural distance (ICD) 4 mm above the
aortic annulus (Figure  2; Video I in the Data Supplement).
The left ventricular outflow tract dimensions were arbitrarily
measured at 4 mm below the virtual annulus. Similarly, we
determined the ICD distance at 4 mm above the annulus, for
simplification and standardization. The formula to determine
the annulus ellipticity index was defined by the ratio of max-
imum and minimum annulus diameters (Dmax/Dmin). The
prosthesis-annulus ratio was defined as the labeled prosthesis
January 2019 2
 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients
diameter divided by the perimeter-derived annulus diameter.
The prosthesis-ICD ratio was defined as the labeled prosthesis
diameter divided by ICD measured 4 mm above the annulus.
Post-TAVI MDCT Analysis
The depth of implantation was defined as the maximum pro-
trusion of THV into the left ventricular outflow tract; level 0
being the native aortic annulus. The distance from the left
main coronary artery to the annulus at baseline was used to
locate the annular plane postimplant. The mean perimeter-
derived diameter of THV was measured at the aortic an-
nulus (level 0) and 4, 8, and 12 mm above it. In that region
of interest, representing the 12 distal mm of a given device,
all prostheses have a cylindrical configuration, allowing for
Downloaded from http://ahajournals.org by on December 9, 2019
Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107
Figure 1. Leaflets distribution in the bi-
cuspid cohort.
device comparison and integration of the supra-annular posi-
tion of the leaflets in the Medtronic Evolut R prosthesis. The
prosthetic ellipticity index (Dmax/Dmin) was also measured at
the annulus level and 4, 8, 12 mm above (Figure 2; Video II in
the Data Supplement).
Statistical Analysis
For statistical analysis, quantitative parameters were described
using the following descriptive statistics: mean, SD, median,
Q1, Q3, extreme values (min and max). In the same way,
qualitative parameters were described using the number
and percentage of each modality. Continuous variables were
plotted using boxplots and scatter plots. Comparisons be-
tween groups were computed using the parametric Student
Figure 2. Various multidetector computed
tomography measurements at the level of
the aortic root.
ICD indicates intercommissural distance.
January 2019 3
 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients

Table 1.  Baseline Clinical Characteristics and Aortic Root Dimensions only of exploratory nature, and no adjustment to the α-risk
of the Study Populations
despite multiple variables have been done.
Tricuspid Bicuspid The individual ratio between the prosthesis diameter and
(n=88) (n=101) P Value the mean aortic annular diameter provides information about
Age, y 83.1±5.7 78.2±10.1 <0.01 appropriate sizing. A ratio >1 indicates oversizing while a
ratio <1 relates to undersizing.
Men, n (%) 41 (46) 66 (65) <0.01
Statistical analyzes have been computed using SAS Version
Hypertension, n (%) 79 (90) 64 (63) <0.01
9.4 TS Level 1M4 Copyright ©2002–2012 by SAS Institute
Dyslipidaemia, n (%) 50 (57) 37 (37) <0.01 Inc, Cary, NC. A 2-sided P<0.05 was considered statistically
Diabetes mellitus, n (%) 13 (15) 17 (17) 0.69 significant.
Dialysis, n (%) 2 (2) 1 (1) 0.48
Current smoker, n (%) 10 (11) 7 (7) 0.16
RESULTS
Previous MI, n (%) 8 (9) 5 (5) 0.26
Baseline Clinical and MDCT Data
Previous PCI, n (%) 8 (9) 32 (32) <0.01
Previous CABG, n (%) 26 (29) 2 (2) <0.01
The baseline clinical characteristics of patients with tri-
cuspid and bicuspid AS are reported in Table 1. Several
Atrial fibrillation, n (%) 27 (31) 24 (24) 0.3
differences were identified between both cohorts: BAV
Previous stroke/TIA, n (%) 11 (12) 10 (10) 0.57
patients were younger (78.2±10.1 versus 83.1±5.7
Previous pacemaker, n (%) 10 (11) 7 (7) 0.29 years; P<0.01), more frequently males (65% versus
COPD, n (%) 23 (26) 26 (26) 0.91 46%; P<0.01), with higher risk profiles as expressed
BMI, kg/m2 28.1±18.4 28.7±34.3 0.47 by the Society for Thoracic Surgery score for mortality
BSA, m 2
1.7±0.2 2.2±0.1 0.05 (11.3±8.5 versus 7.6±4.4; P<0.01). Several elements
LVEF, % 53.4±13.3 54.3±14.9 0.48
like chronic immunosuppressive drugs, higher cre-
atinin level, concomitant mitral valvular disease, and
NYHA class I–II, n (%) 33 (37) 38 (38) 0.98
atrial arrhythmias partly explain this difference in the
NYHA class III–IV, n (%) 50 (57) 52 (51) 0.46
Society for Thoracic Surgery score. We did not perform
STS score, % 7.6±4.4 11.3±8.5 <0.01 any matching of both cohorts because of the rela-
Annulus maximal diameter, 26.1±2.3 29.9±3.8 <0.01 tively small sizes and the focus of this study on MDCT
mm
outcomes.
Downloaded from http://ahajournals.org by on December 9, 2019

Annulus minimal diameter, 20.7±1.9 23.3±3.3 <0.01 Preprocedural MDCT diagnosed type 0 and type
mm
1 BAV in, respectively, 12.9% and 86.1 % of the bi-
Annulus mean perimeter- 23.6±1.9 26.8±3.1 <0.01
cuspid patients. Only 1 patient (1%) had a type 2
derived diameter, mm
BAV (Figure  1). Baseline MDCT analysis, identified a
Annulus mean area derived 23.2±1.9 26.3±3.0 <0.01
diameter, mm
larger annulus and ascending aorta for the BAV group
(26.8±3.1 versus 23.6±1.9; P<0.01; 36.7±5.4 versus
Ellipticity index 1.2±0.1 1.2±0.1 <0.01
30.4±3.4; P<0.01), with a significantly higher calcium
Calcium score, cm3/ 3038.6±3671.1 5143.1±4730.2 0.09
threshold 450 HU
volume (5143±4730 versus 3038±3671 mm3; P=0.02;
Table 1). The ellipticity index was not significantly dif-
ICD at 4 mm above the na 27±3.1 0.02
annulus, mm ferent between BAV and TAV patients: 1.2±0.1 versus
Ascending Aorta, 3 cm 30.4±3.4 36.7±5.4 na
1.2±0.1, P=0.09. Of note, while the mean perimeter-
above the annulus, mm derived diameter of the aortic annulus of BAV patients
Left coronary height, mm 13.4±3.1 15±3.7 <0.01 was 26.8±3.1 mm, the ICD (measured at +4 mm) was
27.0±3.1 mm, P=ns. Thirty-four patients (33.7%) with
Right coronary height, mm 17.1±3.1 18.3±4.2 0.01
BAV had a concordance between the annulus mean
BMI indicates body mass index; BSA, body surface area; CABG, coronary
perimeter-derived diameter and the ICD (ratio 0.9–
artery by-pass grafting; COPD, chronic obstructive pulmonary disease; HU:
Hounsfield Unit; ICD, intercommissural distance; LVEF, left ventricular ejection 1.1), 53 (52.5%) had a flared configuration (annulus
fraction; MI, myocardial infarction; NYHA, New York Heart Association; PCI, smaller than the ICD; ratio <0.9) and only 14 (13.8%)
percutaneous coronary intervention; STS, Society for Thoracic Surgery; and TIA,
had a tapered configuration (annulus greater than the
transient ischemic attack.
ICD; ratio >1.1).
statistics in case of normal distribution and using the non-
parametric Wilcoxon statistics otherwise. Qualitative variables Procedural Details
were described and compared using the χ2 or Fisher statistics
depending on variables distribution. TAVI procedural details are depicted in Table  2.
All statistical analyses were performed at the 0.05 global Overall the femoral route was the predominant
significance level (type I error rate), using 2-sided tests. As this access site. In TAV patients, the Edwards Sapien 3,
is an exploratory study, CIs and statistical tests provided are Medtronic Evolut R, and Boston Lotus devices were

Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107 January 2019 4

 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients

Table 2.  Procedural Details

Tricuspid Bicuspid Evolut R Sapien 3 Lotus

(n=88) (n=101) (n=21) (n=65) (n=10) P Value*
Transfemoral access, n (%) 86 (98) 96 (95) 21 (100) 61 (93.9) 10 (100) na
Transaortic access, n (%) 0 (0) 1 (1) 0 (0) 1 (0.9) 0 (0) na
Transapical access, n (%) 0 (0) 4 (4) 0 (0) 4 (6) 0 (0) na
Subclavian, n (%) 2 (2) 0 (0) 0 (0) 0 (0) 0 (0) na
Device size, mm 26.7±2.1 27.6±2.1 28.1±2.3 27.5±2.0 26.4±1.3 <0.01
Prosthesis-annulus ratio 1.14±0.09 1.04±0.09 1.15±0.1 1.01±0.07 1.00±0.08 <0.01
Prosthesis-ICD ratio na 1.03±0.10 1.16±0.08 1.01±0.10 0.96±0.09 na
Predilation, n (%) 45 (51) 28 (28) 6 (30) 12 (18.2) 5 (50) <0.01
Second valve used, n (%) 1 (1) 8 (8) 3 (15) 4 (6) 0 (0) 0.03
Postdilation, n (%) 16 (18) 20 (20) 6 (30) 11 (15.1) 10 (100) 0.77

*Comparison between tricuspid and bicuspid cohorts. ICD indicates intercommissural distance.

used in, respectively, 19 (21.6%), 26 (29.5%), and 22 Post-TAVI MDCT Data

(25%) of the cases, while 6 (6.8%) patients received
Post-TAVI MDCT findings are summarized in Table 4. The
a Boston Acurate Neo, 6 (6.8%) an Abbott Portico
mean depth of implantation was similar in both groups,
and 9 (10.2%) a Direct Flow prosthesis. In BAV
≈3.4 mm below the aortic annulus. In other words, on
patients, the Edwards Sapien 3, Medtronic Evolut R,
average the prosthesis section of interest (distal edge to
and Boston Lotus devices were used in, respectively,
12 mm above) was located 8.6 mm above the aortic an-
66 (65.3%), 20 (20%), and 10 (9.9 %) of the cases,
nulus in both BAV and TAV patients. Postimplantation, in
while 6 (6.8%) patients received a Boston Acurate
tricuspid patients, the mean diameter of the prosthesis
Neo, 1 (1%) an Abbott Portico, and 1 (1%) a Direct
at the annulus matched the mean perimeter-derived
Flow prosthesis. The mean prosthesis-annulus ratio,
diameter of the aortic annulus at baseline: 23.3±2.2
reflecting the degree of oversizing, was, respectively,
versus 23.6±1.9, P=0.4. In bicuspid patients, prostheses
Downloaded from http://ahajournals.org by on December 9, 2019

1.14±0.04 and 1.04±0.04 (P<0.01) in TAV and BAV

patients. In other words, oversizing was applied in
both groups, but lesser in BAV patients. No under-
sizing was observed. When we integrate the ICD as
a potential reference for sizing in BAV patients, the
mean prosthesis-ICD ratio was even closer to a 1/1
ratio: 1.03±0.1 (Figure  3). The ICD in TAV was not
analyzed, as it is not part of routine practice for stan-
dardized sizing, in tricuspid aortic stenosis.
Balloon predilatation was more frequently carried-
out in TAV patients (51% versus 28%, P<0.01), with
similar rates of postdilatation in both groups (18%
versus 20%, P=0.77). TAVI was a more complex proce-
dure in BAV patients as expressed by a higher need for
a second valve (TAV-in-TAV: 8% versus 1%; P=0.03).
No coronary obstruction was observed in TAV or BAV
patients.
Clinical and Echocardiographic Outcome
Table  3 summarizes the main clinical and echocardio-
graphic outcomes observed at 30 days. Clinical out-
comes were overall similar between groups. The only
significant echographic difference at 30 days was a
smaller indexed orifice area in BAV patients (1.17±0.40
versus 1.33±0.37 cm2/m2; P<0.01). No significant dif-
ference in paravalvular leak and patient-prosthesis mis-
match was found between groups.
were, on average, 11% smaller than the mean annulus
diameter at baseline (24±2.8 versus 26.8±3.1, P<0.01).
This was expressed by the mean prosthesis diameter at
annulus/baseline annulus diameter ratio. Interestingly,
prostheses retained the cylindrical configuration in BAV
and TAV patients with stable diameters from their distal
edge to 12 mm above it. Overall, prosthesis ellipticity
was similar in TAV and BAV patients: 1.18±0.11 versus
1.18±0.14, P=0.5. This was mainly driven by the Evolut
R device, which expressed the greatest consistency in
its ellipticity in TAV and BAV configurations (Table  5).
Finally, prosthesis ellipticity followed the same pattern
as diameters, with stable ratios from the annulus to 8
mm above, (mean depth of implant being 3.4 mm), in
TAV and BAV anatomies.
DISCUSSION
The BAVARD (Bicuspid Aortic Valve Anatomy and
Relationship With Devices) registry is the largest registry
addressing TAVI in BAV with contemporary prosthesis
platforms and pre- and postprocedural MDCT analysis.
This study aimed to capture the sizing ratios used for
TAVI in BAV in contemporary European practice and
analyze the configuration of TAVI prostheses in BAV.
Given the bicuspid type of the study patients, our find-
ings mostly apply to type 0 and type 1 BAV.

Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107 January 2019 5

 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients
Downloaded from http://ahajournals.org by on December 9, 2019

Figure 3. Sizing ratios utilized for different transcatheter aortic valve implantation devices in the bicuspid cohort.
A, Prosthesis labeled diameter/aortic annulus mean perimeter-derived diameter ratio; (B) prosthesis labeled size/intercommissural distance ratio.

BAVARD was not focused on clinical outcomes.
However, we observed similar the Valve Academic
Research Consortium-2 criteria outcomes in TAV and
BAV patients. In a recent report, Yoon et al21 already
demonstrated the improved safety of TAVI with sec-
ond-generation prostheses in bicuspid anatomies as
compared with first-generation devices.
MDCT analysis was the main focus of our registry.
Given the final prosthesis depth of implantation in TAV
and BAV patients, the MDCT region of interest ex-
tended from 4 mm below the annulus to 8 mm above.
At baseline, as compared with TAV patients, the
aortic annulus in BAV patients was larger but not more
elliptical. These findings echo the conclusions from Son
et al.22 Watanabe et al23 compared the outcomes in TAV
and BAV patients post-TAVI with first-generation pros-
theses: patients with BAV had higher gradients, larger
annulus perimeters, and more calcified valves. Higher
postprocedural gradient and valve underexpansion
were frequently observed.
In TAV patients, we observed that the aortic annulus
influenced the final diameter of the TAVI devices. In the
region of interest, the most frequently used second-gen-
eration devices (S3, ER, and Lotus) matched the aortic
annulus mean diameter, remained cylindrical, with con-
stant diameters and ellipticity indexes. With BAV, the

Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107 January 2019 6

 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients

Table 3.  VARC-2 Clinical and Echocardiographic Outcomes at 30 Days Table 4.  Geometry of THV as Assessed in Post-TAVI MDCT

Tricuspid Bicuspid Tricuspid Bicuspid

(88) (101) P Value (N=88) (N=101) P Value
VARC-2 end points Depth of implant, mm 3.4±2.2 3.3±2.9 0.38
 All-cause mortality, n (%) 3 (3.4) 0 (0) 0.61 Mean prosthesis perimeter-derived 23.3±2.2 24.1±2.8 0.04
diameter at annulus, mm
 Myocardial infarction, n (%) 0 1 (1) 1
Prosthesis ellipticity index at annulus 1.18±0.11 1.18±0.14 0.5
 Disabling stroke, n (%) 0 2 (2) 0.5
Prosthesis diameter at annulus/ 0.87±0.07 0.86±0.11 0.5
 Bleeding, n (%) 21 (24) 20 (20) 0.62
prosthesis size ratio
 Life-threating bleeding, 1 (1) 3 (3) 0.62
Prosthesis diameter at annulus/ 0.98±0.06 0.89±0.07 <0.01
n (%)
baseline annulus diameter ratio
 Major bleeding, n (%) 4 (4.5) 11 (11) 0.1
Mean Prosthesis perimeter 23.2±1.9 23.6±2.8 0.43
 Minor bleeding, n (%) 16 (18) 9 (9) 0.06 diameter at 4 mm, mm

 Vascular complications, 24 (27) 16 (16) 0.05 Prosthesis ellipticity index at 4 mm 1.16±0.1 1.17±0.13 0.94
n (%)
Prosthesis diameter at 4 mm/ICD na 0.88±0.11 na
 Major vascular 0 5 (5) 0.06 ratio
complications, n (%)
Mean Prosthesis perimeter 23.4±1.9 23.4±2.8 0.78
 Minor vascular 22 (26) 11 (11) <0.01 diameter at 8 mm, mm
complications, n (%)
Prosthesis ellipticity index at 8 mm 1.17±0.12 1.18±0.13 0.75
 Pacemaker implantation, 12 (14) 13 (13) 0.87
Mean Prosthesis perimeter 23.7±2.2 23.9±2.8 0.55
n (%)
diameter at 12 mm, mm
Echographic outcomes
Prosthesis ellipticity index at 12 mm 1.15±0.1 1.17±0.12 0.23
 Effective orifice area, cm2 2.1±0.5 1.9±0.6 0.07
ICD indicates intercommissural distance; MDCT, multidetector computed
 Indexed effective orifice 1.33±0.37 1.17±0.4 <0.01 tomography; TAVI, transcatheter aortic valve implantation; and THV,
area, cm2/m2 transcatheter heart valve.
 Mean gradient, mm  Hg 9.4±4.9 10.7±4.9 0.15
 Mild-moderate aortic 11(12.5) 21(20.8) 0.11
conserve stable diameters and ellipticity when they
regurgitation, n (%) meet a point of high resistance, with similar patterns
in both TAV and BAV. In TAV patients that point of re-
Downloaded from http://ahajournals.org by on December 9, 2019

 Severe aortic regurgitation, 0 (0) 0 (0) …

n (%) sistance is usually located in the aortic annulus, while
 Moderate patient-prosthesis 5 (5.7) 15 (14.8) 0.09 in BAV patients it could be positioned above the aortic
mismatch, n (%) annulus, at the level of the commissures and leaflets.
 Severe patient-prosthesis 2 (2.3) 8 (7.9) 0.14 In a MDCT analysis of 41 BAV patients treated with S3,
mismatch, n (%) Kawamori et al24 found lesser expansion and greater
VARC-2 indicates Valve Academic Research Consortium-2.
Table 5.  Ellipticity Index of the Most Frequently Used Second-
Generation THV
prostheses trended to be slightly more elliptical but
N Tricuspid Bicuspid P Value
overall exerted the same cylindrical pattern. The only and
Evolut R 47 26 21 0.33
important difference with BAV was prostheses evident
underexpansion, underscored by mean diameters being  Ellipticity index at annulus 1.2±0.1 1.3±0.2
constantly smaller than the mean aortic annulus diam-  Ellipticity index at 4 mm 1.2±0.07 1.2±0.1 0.93
eter and the ICD. This finding highlights potential points  Ellipticity index at 8 mm 1.2±0.1 1.2±0.1 1
of constraints throughout the aortic root. This underex-  Ellipticity index at 12 mm 1.21±1.13 1.24±0.16 0.07
pansion in BAV is of utmost importance as it may po-
Sapien 3 84 19 65 0.01
tentially hamper prosthesis durability or even promote
 Ellipticity index at annulus 1.09±0.05 1.12±0.06 0.01
leaflet thrombosis. It stresses the need for refined siz-
 Ellipticity index at 4 mm 1.09±0.05 1.13±0.07 0.06
ing policies, to select the appropriate prosthesis size, and
procedural technique modification to obtain the max-  Ellipticity index at 8 mm 1.09±0.03 1.14±0.09 0.05
imum expansion achievable in a given bicuspid anatomy.  Ellipticity index at 12 mm 1.09±0.04 1.13±0.08 0.02
The stability of both prosthesis diameters and ellip- Lotus 31 21 10 0.01
ticity from 0 to 12 mm attests to the high radial force  Ellipticity index at annulus 1.11±0.07 1.2±0.1 0.01
and the ability of second-generation devices to reshape
 Ellipticity index at 4 mm 1.11±0.06 1.19 0.01
the surrounding structure in both TAV and BAV. In some
 Ellipticity index at 8 mm 1.14±0.1 1.21±0.09
patients, with self-expanding or mechanically expanded
 Ellipticity index at 12 mm 1.1±0.06 1.21±0.09
devices, that reshaping may be facilitated by an ade-
quate postdilatation. Second-generation prostheses, THV indicates transcatheter heart valve.

Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107 January 2019 7

 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients
ellipticity in BAV patients as compared with TAVI. One
possible explanation for that discrepancy with our find-
ings, could be difference in sizing, with bigger prosthe-
ses used in their series and devices potentially failing
to achieve their maximal diameter and circularity in a
relatively too constrained landing zone. In our series,
minimal oversizing was applied (3%–4%), when using
the mean perimeter-derived aortic annulus diameter,
for sizing in BAV.
In patients with TAV, the landing zone usually inte-
grates the aortic annulus and the left ventricular out-
flow tract, 4 mm below it. As an analogy, given the
location of constraint points, the landing zone in BAV
patients could run from the aortic annulus to 4 mm
above it. That explains, in an effort of simplification,
our proposal of integration of the ICD at 4 mm above
the annulus for sizing in BAV patients, at least type 0
and type 1 variations. Several configurations can be
identified. In a tubular configuration, the mean aortic
annulus diameter matches the ICD and can be used
for sizing with an average oversizing of 3% in our
cohort. In a flared configuration, in which the mean
aortic annulus diameter is smaller than the ICD, it could
also be used as the reference for sizing. In a tapered
configuration (mean perimeter-derived diameter of
the annulus greater than ICD), the ICD could be inte-
grated, with a 0.9–1/1 ratio because prostheses were
systematically smaller than the ICD in our BAV patients
(Figure 4). Importantly, annulus-based sizing was appli-
Downloaded from http://ahajournals.org by on December 9, 2019
cable to 88% of our BAV patients (Table I in the Data
Supplement). Even though rare, it remains important to
identify a tapered configuration. In such anatomy, an
Figure 4. Various configuration of the landing zone in bicuspid patients and simplified sizing algorithm.
ICD indicates intercommissural distance.
Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107
annular-based sizing would result in selecting a device
potentially too large for the patient, with inherent risks
of aortic root rupture or greater device underexpansion.
The calcium burden is likely a major player in the final
expansion of TAVI prostheses. It should be quantified
and integrated in the sizing process.
From our registry, we can hypothetize that in type
0 and type 1 BAV, (1) annulus-based sizing is valid, (2)
minimal oversizing (3%–4%) ensures good clinical out-
comes, (3) in gray zones and when the ICD is smaller
that the mean annular diameter (taper), selecting a
smaller THV size could avoid excessive oversizing and its
consequences, and (4) landing THVs 3 to 4 mm below
the aortic annulus could be safe and associated with
good clinical outcomes. Our findings and simplified siz-
ing algorithm need to be validated in larger prospective
registries, ideally evaluating the different types of pros-
theses separately. This sizing algorithm will be part of
the BIVOLUT X prospective registry.
Limitations
Our study has several limitations. It is a retrospective
registry with selection bias in THV type and sizing from
one center to the other. As not all consecutive patients
had pre- and post-TAVI MDCT assessment, our find-
ings are exposed to interpretation bias but the number
of patients with interpretable data may strengthen our
conclusions. The distribution of the different types of
second-generation prostheses was not homogeneous
between BAV and TAV patients. Various sizing algo-
rithms were used and only a prospective registry could
January 2019 8
 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients
validate our proposed simplified sizing methodology for
BAV. Differences in baseline clinical characteristics be-
tween groups also represent a limitation of our registry.
Given the relatively small size of the study population, we
liberally did not perform any comparison of devices nor
paired analysis, so, our findings have to be analysed with
caution. The small size of our cohort may explain the
lack of mortality observed at 30 days in bicuspid patients.
Larger prospective registries are required, exploring the
appropriate sizing algorithm for each prosthesis type.
Conclusions
Second-generation TAVI prostheses similarly reshape the
aortic annulus in tricuspid and bicuspid aortic valves. On
average prostheses were deployed 3.4 mm below the
aortic annulus, in the BAVARD registry, with excellent
clinical outcomes. Devices keep consistent diameters
from distal edge to 12 mm above. Prosthesis underex-
pansion is constantly observed in BAV. Annular-based siz-
ing is accurate in type 0 and type 1 bicuspid valves with
minimal oversizing. The ICD, 4 mm above the annulus,
could be integrated in the sizing process for gray zones.
ARTICLE INFORMATION
Received June 23, 2018; accepted November 29, 2018.
The Data Supplement is available at https://www.ahajournals.org/doi/
suppl/10.1161/CIRCINTERVENTIONS.118.007107.
Downloaded from http://ahajournals.org by on December 9, 2019
Authors
Didier Tchetche, MD; Chiara de Biase, MD; Lennart van Gils, MD; Radoslaw
Parma, MD; Andrzej Ochala, MD; Thierry Lefevre, MD; Thomas Hovasse,
MD; Ole De Backer, MD, PhD; Lars Sondergaard, MD, PhD; Sabine Bleiziffer,
MD; Rudiger Lange, MD; Ran Kornowski, MD; Uri Landes, MD; Bjarne Linde
Norgaard, MD, PhD; Luigi Biasco, MD; Raphaël Philippart, MD; Javier Molina-
Martin de Nicolas, MD; Darren Mylotte, MD; Caroline Lemee, PhD; Nicolas
Dumonteil, MD; Nicolas M. Van Mieghem, MD, PhD
Correspondence
Didier Tchetche, MD, Groupe Cardiovasculaire Interventionnel, Clinique Pasteur,
45 avenue de Lombez, 31076 Toulouse cedex 3, France. Email d.tchetche@
clinique-pasteur.com
Affiliations
Groupe CardioVasculaire Interventionnel, Clinique Pasteur, Toulouse, France
(D.T., C.d.B., R.P., J.M.-M.d.N., N.D.). Thoraxcenter, Erasmus Medical Center,
Rotterdam, the Netherlands (L.v.G., N.M.V.M.). 3rd Department of Cardiology,
Medical University of Silesia, Katowice, Poland (R.P., A.O.). Institut cardiovascu-
laire Paris sud, Hôpital prive Jacques Cartier, Massy, France (T.L., T.H.). The Heart
Center-Rigshospitalet, Copenhagen, Denmark (O.D.B., L.S.). German Heart
Center, Munich (S.B., R.L.). Cardiology Department, Rabin Medical Center,
Petach Tikva, Tel-Aviv University, Israel (R.K., U.L.). Department Cardiology,
Aarhus University Hospital, Denmark (B.L.N.). Centro Cardiotoracico TICINO,
Lugano, Switzerland (L.B.). Department of Cardiology, University Hospital
Galway, Ireland (D.M.). Independent biostatistician, Toulouse, France (C.L.).
Disclosures
Dr Molina-Martin de Nicolas has received the Magda Heras Grant from
the Spanish Society of Cardiology (SEC). Dr Lefevre is proctor for Edwards
LifeSciences. Dr Tchetche is proctor for Edwards LifeSciences, Medtronic and
Boston Scientific. Dr Dumonteil is proctor for Edwards LifeSciences, Medtronic
Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107
and Boston Scientific. Dr Van Mieghem has received research grant support
from Abbott vascular, Edwards LifeSciences, Medtronic, Boston Scientific,
Claret Medical. Dr de Biase has been supported by a research grant provided by
the Cardiopath PhD program. The other authors report no conflicts.
REFERENCES
1. Leon MB, Smith CR, Mack M, Miller DC, Moses JW, Svensson LG, Tuzcu
EM, Webb JG, Fontana GP, Makkar RR, Brown DL, Block PC, Guyton RA,
Pichard AD, Bavaria JE, Herrmann HC, Douglas PS, Petersen JL, Akin JJ,
Anderson WN, Wang D, Pocock S, Investigators PT. Transcatheter aortic-
valve implantation for aortic stenosis in patients who cannot undergo sur-
gery. N Engl J Med. 2010;363:1597–607. doi: 10.1056/NEJMoa1008232
2. Adams DH, Popma JJ, Reardon MJ. Transcatheter aortic-valve replacement
with a self-expanding prosthesis. N Engl J Med. 2014;371:967–968. doi:
10.1056/NEJMoa1400590
3. Leon MB, Smith CR, Mack MJ, Makkar RR, Svensson LG, Kodali SK,
Thourani VH, Tuzcu EM, Miller DC, Herrmann HC, Doshi D, Cohen DJ,
Pichard AD, Kapadia S, Dewey T, Babaliaros V, Szeto WY, Williams MR,
Kereiakes D, Zajarias A, Greason KL, Whisenant BK, Hodson RW, Moses
JW, Trento A, Brown DL, Fearon WF, Pibarot P, Hahn RT, Jaber WA,
Anderson WN, Alu MC, Webb JG; PARTNER 2 Investigators. Transcatheter
or surgical aortic-valve replacement in intermediate-risk patients. N Engl J
Med. 2016;374:1609–1620. doi: 10.1056/NEJMoa1514616
4. Siu SC, Silversides CK. Bicuspid aortic valve disease. J Am Coll Cardiol.
2010;55:2789–2800. doi: 10.1016/j.jacc.2009.12.068
5. Rodríguez-Caulo EA, Araji OA, Barquero JM. Transapical aortic valve im-
plantation in bicuspid aortic valves: must be an absolute contraindication?
Res Cardiovasc Med. 2012;1:37–39. doi: 10.5812/cardiovascmed.4498
6. Philip F, Faza NN, Schoenhagen P, Desai MY, Tuzcu EM, Svensson LG,
Kapadia SR. Aortic annulus and root characteristics in severe aortic ste-
nosis due to bicuspid aortic valve and tricuspid aortic valves: implica-
tions for transcatheter aortic valve therapies. Catheter Cardiovasc Interv.
2015;86:E88–E98. doi: 10.1002/ccd.25948
7. Sabet HY, Edwards WD, Tazelaar HD, Daly RC. Congenitally bicuspid aortic
valves: a surgical pathology study of 542 cases (1991 through 1996) and a
literature review of 2,715 additional cases. Mayo Clin Proc. 1999;74:14–
26. doi: 10.4065/74.1.14
8. Mylotte D, Lefevre T, Søndergaard L, Watanabe Y, Modine T, Dvir D,
Bosmans J, Tchetche D, Kornowski R, Sinning JM, Thériault-Lauzier P,
O’Sullivan CJ, Barbanti M, Debry N, Buithieu J, Codner P, Dorfmeister M,
Martucci G, Nickenig G, Wenaweser P, Tamburino C, Grube E, Webb JG,
Windecker S, Lange R, Piazza N. Transcatheter aortic valve replacement in
bicuspid aortic valve disease. J Am Coll Cardiol. 2014;64:2330–2339. doi:
10.1016/j.jacc.2014.09.039
9. Kong WK, Delgado V, Poh KK, Regeer MV, Ng AC, McCormack L, Yeo
TC, Shanks M, Parent S, Enache R, Popescu BA, Liang M, Yip JW, Ma
LC, Kamperidis V, van Rosendael PJ, van der Velde ET, Ajmone Marsan N,
Bax JJ. Prognostic implications of raphe in bicuspid aortic valve anatomy.
JAMA Cardiol. 2017;2:285–292. doi: 10.1001/jamacardio.2016.5228
10. Tchetche D, Van Mieghem NM. New-generation TAVI devices: description
and specifications. EuroIntervention. 2014;10 Suppl U:U90–U100. doi:
10.4244/EIJV10SUA13
11. Perlman GY, Blanke P, Dvir D, Pache G, Modine T, Barbanti M, Holy EW,
Treede H, Ruile P, Neumann FJ, Gandolfo C, Saia F, Tamburino C, Mak G,
Thompson C, Wood D, Leipsic J, Webb JG. Bicuspid aortic valve stenosis:
favorable early outcomes with a next-generation transcatheter heart valve
in a multicenter study. JACC Cardiovasc Interv. 2016;9:817–824. doi:
10.1016/j.jcin.2016.01.002
12. Yousef A, Simard T, Webb J, Rodés-Cabau J, Costopoulos C, Kochman
J, Hernández-Garcia JM, Chiam PT, Welsh RC, Wijeysundera HC, García
E, Ribeiro HB, Latib A, Huczek Z, Shanks M, Testa L, Farkouh ME, Dvir
D, Velianou JL, Lam BK, Pourdjabbar A, Glover C, Hibbert B, Labinaz M.
Transcatheter aortic valve implantation in patients with bicuspid aortic
valve: a patient level multi-center analysis. Int J Cardiol. 2015;189:282–
288. doi: 10.1016/j.ijcard.2015.04.066
13. Yoon SH, Lefèvre T, Ahn JM, Perlman GY, Dvir D, Latib A, Barbanti M,
Deuschl F, De Backer O, Blanke P, Modine T, Pache G, Neumann FJ, Ruile
P, Arai T, Ohno Y, Kaneko H, Tay E, Schofer N, Holy EW, Luk NHV, Yong G,
Lu Q, Kong WKF, Hon J, Kao HL, Lee M, Yin WH, Park DW, Kang SJ, Lee
SW, Kim YH, Lee CW, Park SW, Kim HS, Butter C, Khalique OK, Schaefer
U, Nietlispach F, Kodali SK, Leon MB, Ye J, Chevalier B, Leipsic J, Delgado
V, Bax JJ, Tamburino C, Colombo A, Søndergaard L, Webb JG, Park SJ.
January 2019 9
 Tchetche et al; MDCT for Transcatheter Heart Valves in Bicuspid Patients
Transcatheter aortic valve replacement with early- and new-generation
devices in bicuspid aortic valve stenosis. J Am Coll Cardiol. 2016;68:1195–
1205. doi: 10.1016/j.jacc.2016.06.041
14. Zhao ZG, Jilaihawi H, Feng Y, Chen M. Transcatheter aortic valve im-
plantation in bicuspid anatomy. Nat Rev Cardiol. 2015;12:123–128. doi:
10.1038/nrcardio.2014.161
15. Tanaka R, Yoshioka K, Niinuma H, Ohsawa S, Okabayashi H, Ehara S.
Diagnostic value of cardiac CT in the evaluation of bicuspid aortic ste-
nosis: comparison with echocardiography and operative findings. AJR Am
J Roentgenol. 2010;195:895–899. doi: 10.2214/AJR.09.3164
16. Mylotte D, Dorfmeister M, Elhmidi Y, Mazzitelli D, Bleiziffer S, Wagner A,
Noterdaeme T, Lange R, Piazza N. Erroneous measurement of the aortic
annular diameter using 2-dimensional echocardiography resulting in inap-
propriate CoreValve size selection: a retrospective comparison with mul-
tislice computed tomography. JACC Cardiovasc Interv. 2014;7:652–661.
doi: 10.1016/j.jcin.2014.02.010
17. Binder RK, Webb JG, Willson AB, Urena M, Hansson NC, Norgaard BL,
Pibarot P, Barbanti M, Larose E, Freeman M, Dumont E, Thompson C,
Wheeler M, Moss RR, Yang TH, Pasian S, Hague CJ, Nguyen G, Raju R,
Toggweiler S, Min JK, Wood DA, Rodés-Cabau J, Leipsic J. The impact of
integration of a multidetector computed tomography annulus area siz-
ing algorithm on outcomes of transcatheter aortic valve replacement: a
prospective, multicenter, controlled trial. J Am Coll Cardiol. 2013;62:431–
438. doi: 10.1016/j.jacc.2013.04.036
18. Hayashida K, Bouvier E, Lefèvre T, Chevalier B, Hovasse T, Romano M,
Garot P, Watanabe Y, Farge A, Donzeau-Gouge P, Cormier B, Morice
MC. Transcatheter aortic valve implantation for patients with severe bi-
cuspid aortic valve stenosis. Circ Cardiovasc Interv. 2013;6:284–291. doi:
10.1161/CIRCINTERVENTIONS.112.000084
19. Kappetein AP, Head SJ, Généreux P, Piazza N, van Mieghem NM,
Blackstone EH, Brott TG, Cohen DJ, Cutlip DE, van Es GA, Hahn RT,
Kirtane AJ, Krucoff MW, Kodali S, Mack MJ, Mehran R, Rodés-Cabau J,
Vranckx P, Webb JG, Windecker S, Serruys PW, Leon MB. Updated stan-
dardized endpoint definitions for transcatheter aortic valve implantation:
Downloaded from http://ahajournals.org by on December 9, 2019
Circ Cardiovasc Interv. 2019;12:e007107. DOI: 10.1161/CIRCINTERVENTIONS.118.007107
the Valve Academic Research Consortium-2 consensus document. Eur
Heart J. 2012;33:2403–2418. doi: 10.1093/eurheartj/ehs255
20. Sievers HH, Schmidtke C. A classification system for the bicuspid
aortic valve from 304 surgical specimens. J Thorac Cardiovasc Surg.
2007;133:1226–1233. doi: 10.1016/j.jtcvs.2007.01.039
21. Yoon SH, Bleiziffer S, De Backer O, Delgado V, Arai T, Ziegelmueller J,
Barbanti M, Sharma R, Perlman GY, Khalique OK, Holy EW, Saraf S,
Deuschl F, Fujita B, Ruile P, Neumann FJ, Pache G, Takahashi M, Kaneko H,
Schmidt T, Ohno Y, Schofer N, Kong WKF, Tay E, Sugiyama D, Kawamori H,
Maeno Y, Abramowitz Y, Chakravarty T, Nakamura M, Kuwata S, Yong G,
Kao HL, Lee M, Kim HS, Modine T, Wong SC, Bedgoni F, Testa L, Teiger E,
Butter C, Ensminger SM, Schaefer U, Dvir D, Blanke P, Leipsic J, Nietlispach
F, Abdel-Wahab M, Chevalier B, Tamburino C, Hildick-Smith D, Whisenant
BK, Park SJ, Colombo A, Latib A, Kodali SK, Bax JJ, Søndergaard L, Webb
JG, Lefèvre T, Leon MB, Makkar R. Outcomes in transcatheter aortic valve
replacement for bicuspid versus tricuspid aortic valve stenosis. J Am Coll
Cardiol. 2017;69:2579–2589. doi: 10.1016/j.jacc.2017.03.017
22. Son JY, Ko SM, Choi JW, Song MG, Hwang HK, Lee SJ, Kang JW.
Measurement of the ascending aorta diameter in patients with severe bi-
cuspid and tricuspid aortic valve stenosis using dual-source computed to-
mography coronary angiography. Int J Cardiovasc Imaging. 2011;27(suppl
1):61–71. doi: 10.1007/s10554-011-9956-5
23. Watanabe Y, Chevalier B, Hayashida K, Leong T, Bouvier E, Arai T, Farge
A, Hovasse T, Garot P, Cormier B, Morice MC, Lefèvre T. Comparison of
multislice computed tomography findings between bicuspid and tricuspid
aortic valves before and after transcatheter aortic valve implantation.
Catheter Cardiovasc Interv. 2015;86:323–330. doi: 10.1002/ccd.25830
24. Kawamori H, Yoon SH, Chakravarty T, Maeno Y, Kashif M, Israr S,
Abramowitz Y, Mangat G, Miyasaka M, Rami T, Kazuno Y, Takahashi N,
Jilaihawi H, Nakamura M, Cheng W, Friedman J, Berman D, Sharma R,
Makkar RR. Computed tomography characteristics of the aortic valve
and the geometry of SAPIEN 3 transcatheter heart valve in patients
with bicuspid aortic valve disease. Eur Heart J Cardiovasc Imaging.
2018;19:1408–1418. doi: 10.1093/ehjci/jex333
January 2019 10