Structural Heart Disease
`
`Transcatheter Aortic Valve Implantation With
`the New Balloon-Expandable Sapien 3 Versus
`Sapien XT Valve System
`A Propensity Score–Matched Single-Center Comparison
`
`Freek Nijhoff, MD; Masieh Abawi, BSc; Pierfrancesco Agostoni, MD, PhD;
`Faiz Z. Ramjankhan, MD; Pieter A. Doevendans, MD, PhD; Pieter R. Stella, MD, PhD
`
`Background—The new balloon-expandable Sapien 3 transcatheter heart valve (S3-THV) incorporates new features to
`reduce aortic regurgitation (AR) and vascular complications in transcatheter aortic valve implantation. Aim of this study
`is to compare the outcomes of the S3-THV with the preceding Sapien XT THV (SXT-THV) in patients who underwent
`transcatheter aortic valve implantation for symptomatic severe native aortic stenosis.
`Methods and Results—Eligible patients were retrospectively identified in our institutional database and periprocedural
`clinical and imaging data were collected. Non-parsimonious one-to-many propensity score matching was performed to
`account for differences in baseline characteristics. Between November 2011 and December 2014, 167 patients underwent
`balloon-expandable transcatheter aortic valve implantation with either the S3-THV (n=49) or SXT-THV (n=118). Forty-
`four (89.8%) S3-THV patients were matched to 66 (55.9%) SXT-THV patients (mean age 80.3±8.4 and 80.5±7.8 years,
`median EuroSCORE 15.8 and 16.5%, respectively). In the S3-THV and SXT-THV groups, transfemoral approach (77.3%
`versus 78.8%) and postdilatation rates (15.9% versus 12.1%) were similar. Predischarge echocardiography demonstrated
`a lower incidence of ≥mild AR (15.9% versus 46.2%, P=0.003) for the S3-THV, despite reduced annulus area to prosthesis
`oversizing (8.2±5.1 versus 18.2±10.7%, P=0.001). Transfemoral access site–related life-threatening or major bleedings
`and vascular complications were absent in the S3-THV group (0% versus 7.7%, P=0.15). No differences were observed
`in pacemaker implantation rate (9.8% versus 8.8%, P=0.94) and 30-day mortality (both 5%).
`Conclusions—In this retrospective, propensity score–matched analysis, the S3-THV performed superiorly to the SXT-
`THV, as demonstrated by improved valve patency and increased transfemoral access safety. (Circ Cardiovasc Interv.
`2015;8:e002408. DOI: 10.1161/CIRCINTERVENTIONS.115.002408.)
`
`Key Words: aortic regurgitation ◼ aortic valve stenosis ◼ transcatheter aortic valve implantation ◼ vascular complications
`
`Transcatheter aortic valve implantation (TAVI) has become a
`
`firmly established treatment option for symptomatic severe
`aortic stenosis (AS) in inoperable patients1 and patients at high
`operative risk.2,3 Although proven noninferior to surgical aortic
`valve replacement in terms of all-cause mortality, TAVI is asso-
`ciated with a higher incidence of postoperative aortic regurgita-
`tion (AR) and vascular complications.2,3 Other non-negligible
`TAVI-related complications are cerebral embolic events and
`advanced conduction disturbances.4 The extension of TAVI to
`lower risk populations requires minimization of these adverse
`events, partly depending on technological developments in
`transcatheter heart valves (THVs) and delivery systems.
`Recently, the new balloon-expandable Sapien 3 THV
`(S3-THV; Edwards Lifesciences, Irvine, CA) has become
`commercially available.5 The S3-THV embodies the next
`
`generation balloon-expandable valve of the Sapien valve fam-
`ily, building on the clinical experiences gained with the pre-
`vious Sapien XT-THV (SXT-THV; Edwards Lifesciences).
`With its new features, including an outer annular sealing
`cuff, improved delivery system, and low crimped profile, the
`S3-THV is thought to achieve better results than the preceding
`SXT-THV.5 Initial data from a multicenter registry prospec-
`tively evaluating the S3-THV looked promising, reporting low
`rates of AR, vascular complications, and stroke.6 Comparative
`studies on the clinical outcomes of TAVI with the S3-THV
`and SXT-THV are currently scarce.
`Aim of this study was to retrospectively compare the
`hemodynamic and clinical outcomes of TAVI with the
`S3-THV versus the SXT-THV in patients with symptomatic
`severe native AS.
`
`Received January 21, 2015; accepted May 13, 2015.
`From the Departments of Cardiology (F.N., M.A., P.A., P.A.D., P.R.S.) and Cardiothoracic Surgery (F.Z.R.), University Medical Center Utrecht, Utrecht,
`Netherlands.
`The Data Supplement is available at http://circinterventions.ahajournals.org/lookup/suppl/doi:10.1161/CIRCINTERVENTIONS.115.002408/-/DC1.
`Correspondence to Pieter R. Stella, MD, PhD, Department of Cardiology, University Medical Center Utrecht, Heidelberglaan 100, Room E.04.210, 3584
`CX, Utrecht, the Netherlands. E-mail p.stella@umcutrecht.nl
`© 2015 American Heart Association, Inc.
`Circ Cardiovasc Interv is available at http://circinterventions.ahajournals.org
`
`DOI: 10.1161/CIRCINTERVENTIONS.115.002408
`
`1
`
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`Downloaded from
`
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`http://circinterventions.ahajournals.org/
`
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
` by guest on June 22, 2017
`
`Page 01 of 20
`
`

`

`2
`
` Nijhoff et al
`
` Comparison of Sapien 3 and Sapien XT
`
`WHAT IS KNOWN
`• The new Sapien 3 is designed to reduce aortic regur-
`gitation and vascular complications after transcathe-
`ter aortic valve implantation, with its annular sealing
`cuff and lower delivery profile.
`• Initial series suggest a higher pacemaker rate for the
`Sapien 3 compared with the preceding Sapien XT.
`
`WHAT THE STUDY ADDS
`• In this propensity score–matched analysis, the
`Sapien 3 performs better than the Sapien XT in terms
`of postoperative aortic regurgitation and vascular
`complications.
`• Better valve patency was achieved with the Sapien 3,
`despite reduced oversizing, which may lower the risk
`of mechanical complications.
`• A high implantation of the Sapien 3 is feasible and
`may provide similar pacemaker implantation rates
`compared with Sapien XT, compensating for the
`higher stent frame.
`
`Methods
`This is a retrospective single-center study. All patients who underwent
`TAVI for severe native AS at the University Medical Center Utrecht
`by means of S3-THV or SXT-THV implantation were identified in
`our institutional database and included in the study. Implantation of
`the SXT-THV had to be state-of-the-art, meaning performed with the
`Novaflex+ delivery system through an expandable sheath. Patients
`with bicuspid aortic valve anatomy or a degenerated surgical aortic
`valve were excluded from the study.
`Patients were selected for TAVI based on Heart Team discussion
`involving at least one interventional cardiologist and one cardiac
`surgeon. Reasons to refrain from surgical aortic valve replacement
`included high operative risk (logistic EuroSCORE-I≥15%) and the
`presence of contraindications (eg, porcelain aorta, frailty, patent
`grafts in proximity of the sternum). Workup of TAVI candidates in-
`cluded transthoracic echocardiography (TTE), coronary angiography,
`thoracic and abdominal multislice computed tomography, and con-
`sultation of the geriatrics department.
`Relevant periprocedural clinical and imaging data were collected
`and registered in a database. Follow-up was obtained using documen-
`tation of standard-of-care outpatient visits, and survival status was
`attained by interrogation of the Dutch municipal personal records
`database. Outcomes were registered in compliance with the Valve
`Academic Research Consortium 2 (VARC-2) criteria. Device success
`was defined accordingly as the proper implantation of the first valve
`prosthesis used, with intended performance of the valve (peak aortic
`flow velocity <3 m/s and no moderate or severe AR) and no proce-
`dural mortality. Vascular complications were documented for all pro-
`cedural access sites, defined as any location traversed by a guidewire,
`a catheter, or a sheath during the procedure, including arteries, veins,
`left ventricular apex, and the aorta.
`All patients gave informed consent for the TAVI procedure, and
`the study was performed under a waiver obtained from the institu-
`tional medical ethics committee (14–661/C).
`
`Echocardiographic, Angiographic, and Multislice
`Computed Tomography Evaluation
`All patients underwent TTE examination one day before TAVI and
`before discharge. Native valve function was assessed according to the
`guidelines.7,8 Left ventricular ejection fraction was measured by the
`
`modified Simpson’s biplane method or, in case of insufficient image
`quality, visually estimated and quantified in incremental steps of 5%.
`Prosthetic valve function was evaluated as recommended. Left ven-
`tricular outflow tract (LVOT) diameter measurements and Doppler
`velocity recordings for calculation of the effective orifice area were
`performed just beneath the ventricular margin of the prosthetic stent,
`as previously validated.9 AR severity (both trans- and paraprosthetic)
`was assessed and classified by means of the integrative approach en-
`dorsed by the VARC-2 recommendations.10
`Angiographic assessment of AR severity was performed accord-
`ing to the Sellers classification,11 directly after valve implantation and
`after any countermeasures to address significant AR. Measurements
`of implantation depth and skewness of valve position were performed
`on angiographic images as previously described (see Figure in Data
`Supplement).12 Implantation depth is presented both in percentages of
`stent frame height extending below the annulus plane and in millime-
`ters. All measurements were performed by operators experienced in
`angiographic evaluations and independent from the procedure itself.
`Preprocedural multislice computed tomography evaluation in-
`cluded measurement of the aortic annulus and aortic root dimensions
`and eligibility assessment of the TAVI access sites. Aortic annulus
`dimensions (minimum diameter, maximum diameter, perimeter,
`and area) were measured according to standard procedures using
`dedicated software (3Mensio; Pie Medical Imaging, Maastricht, The
`Netherlands). Valve prosthesis size was selected in accordance with
`the manufacturer’s recommendations (Figure 1).
`
`Valve Devices
`The SXT-THV and its new iteration, S3-THV, are both balloon-ex-
`pandable valves that consist of a trileaflet bovine pericardial valve
`sewn into a cobalt-chromium frame. The lower two-thirds of the
`frame are covered with an internal polyethylene terephthalate skirt.
`New feature to the S3 is the external polyethylene terephthalate seal-
`ing cuff designed to improve apposition with the aortic annulus and
`minimize paravalvular leakage. Other improvements to the S3-THV
`are the enhanced frame geometry that allows lower delivery profiles
`and the higher radial strength for better maintenance of circularity
`after deployment.13 Both the SXT-THV and S3-THV valves are avail-
`able in the sizes 23-mm, 26-mm, and 29-mm, whereas currently only
`the SXT-THV has a 20-mm version to accommodate small annuli.
`In the transfemoral approach, the SXT-THV and S3-THV valves
`are implanted with the NovaFlex+ and Commander delivery sys-
`tems, respectively, introduced through expandable sheaths (eSheath;
`Edwards Lifesciences). In transapical or direct aortic procedures, the
`SXT-THV and S3-THV are deployed with the Ascendra and Certitude
`delivery systems, respectively. New refinement to the transfemoral
`S3-THV delivery system is a fine alignment wheel that allows small
`changes to prosthesis position without having to push or pull the
`whole delivery system, increasing positioning precision. Device
`characteristics and sizing charts for the SXT-THV and S3-THV are
`provided in Figure 1.
`
`Implantation Procedure and Technique
`Valve implantation was performed per transfemoral, transaortic, or
`transapical approach, in order of our institutional preference, depend-
`ing on the presence of suitable access sites. Common access techniques
`were used. All transfemoral procedures involved a full percutane-
`ous technique. Suture-mediated closure devices (Perclose ProGlide;
`Abbott Laboratories, Abbott Park, IL) were inserted into the femoral
`arteries to facilitate vascular closure. Conscious sedation was the de-
`fault anesthetic method in transfemoral procedures; in surgical TAVI,
`general anesthesia was instituted. Fluoroscopic guidance was used to
`guide prosthesis positioning and deployment, whereas intraprocedural
`imaging support was accounted for by intracardiac echocardiography
`in the transfemoral and transesophageal echocardiography in surgi-
`cal TAVI procedures. After routine predilatation to prepare the device
`landing zone, valve implantation was performed under rapid ventricu-
`lar pacing (180–200 bpm) by means of a 2-step inflation technique,
`as previously described for SXT-THV.12 The S3-THV was deployed
`
`Downloaded from
`
`http://circinterventions.ahajournals.org/
`
` by guest on June 22, 2017
`
`Page 02 of 20
`
`

`

`3
`
` Nijhoff et al
`
` Comparison of Sapien 3 and Sapien XT
`
`Figure 1. Device characteristics and sizing recommendations for the Sapien 3 (A) and Sapien XT (B).
`
`during one slow inflation (5–10 s). Prosthesis position and function
`and patency of the coronary ostia were evaluated with angiography
`and intracardiac echocardiography or transesophageal echocardiogra-
`phy. Significant aortic regurgitation was addressed by postdilatation or
`second valve implantation, at discretion of the operator.
`
`Statistical Analysis
`Data were analyzed using IBM SPSS Statistics software version
`20 (IBM Corp., Armonk, NY) and R version 2.12.0 (http://www.r-
`project.org). Continuous variables are presented as means±standard
`deviation or medians [interquartile range], as considered appropri-
`ate, and categorical variables as counts and percentages. Comparison
`of continuous variables between groups was performed with the
`Student’s t test or Mann–Whitney test, depending on data distri-
`bution. Categorical variables were compared using Chi-square or
`Fisher’s Exact test, as considered appropriate. Two-tailed P values
`<0.05 were considered statistically significant.
`Propensity score matching was performed to adjust for differ-
`ences in baseline characteristics between the S3-THV and the SXT-
`THV groups. The propensity score, reflecting the propensity of being
`treated with the S3-THV, was calculated by means of nonparsimo-
`nious binary logistic regression, including the following baseline
`features: age, sex, body mass index, hypertension, diabetes mellitus,
`glomerular filtration rate, coronary artery disease, peripheral artery
`disease, cerebrovascular disease, myocardial infarction, percutane-
`ous coronary intervention, coronary artery bypass grafting, atrial
`fibrillation, chronic obstructive pulmonary disease, pulmonary hy-
`pertension, porcelain aorta, left ventricular ejection fraction, mean
`transaortic pressure gradient, baseline moderate to severe AR, and
`procedural access. The 2 groups were matched on the logit of the
`propensity score (caliper set to 0.20) using a greedy, nearest neighbor,
`one (S3-THV) to many (SXT-THV) matching algorithm. Balance di-
`agnostics were performed by inspection of the weighted standardized
`mean differences (d) (<0.20 indicating adequate balance) as previ-
`ously devised.14 In the absence of expert consensus on the estima-
`tion of treatment effects in propensity score matching, the 2 matched
`groups were handled as unpaired independent groups.15 Although all
`reported values for the matched control group represent unweighted
`data for transparency purposes, all statistical analyses were weighted
`for the number of matches.
`
`Results
`Between November 2011 and December 2014, out of 253
`patients treated with TAVI because of severe symptomatic
`AS at the University Medical Center Utrecht, 168 (66.9%)
`
`received a balloon-expandable valve. One SXT-THV patient
`was excluded because of degenerated surgical valve disease,
`leaving 167 patients for further analysis. Forty-nine (29.3%)
`patients received a S3-THV (which was introduced in Febru-
`ary 2014) and 118 (70.7%) a SXT-THV. Baseline characteris-
`tics, procedural features, and outcomes of the total population
`are provided in Tables I–III in the Data Supplement. The over-
`all S3 and SXT-THV groups were comparable with respect to
`logistic EuroSCORE (median 15.8% versus 16.0%, P=0.899)
`and comorbidities. In the S3-THV group, there was a trend
`toward a lower incidence of renal impairment (49% ver-
`sus 63%, P=0.088) and atrial fibrillation (25% versus 39%,
`P=0.084), with significantly lower left ventricular ejection
`fraction (48.5±12.4 versus 52.7±12.2%, P=0.043).
`By means of one-to-many propensity score matching, 44
`(89.8%) S3-THV patients were matched to 66 (55.9%) SXT-
`THV patients (1.5 matches per S3-THV patient on average).
`Baseline characteristics were well balanced, whereas imag-
`ing data showed differences between the matched groups
`(see Table 1). Aortic annuli were larger in the S3-THV group,
`demonstrated by the significantly larger diameter and area mea-
`surements on multislice computed tomography (all P<0.05). For
`each prosthesis size (23-mm, 26-mm, and 29-mm), larger annuli
`were treated with the S3-THV, with significantly lower percent-
`age area and mean diameter oversizing (Figure 2A and 2B).
`The majority of S3- and SXT-THV patients were treated
`with transfemoral TAVI (77.3% and 78.8%) under local
`anesthesia. The S3-THV was implanted higher (ie, more
`aortic), indicated by the lesser percentage of stent frame
`height extending below the annulus plane (20.0±11.9 versus
`31.0±11.7%, P<0.001) and lower implantation depth (4.0±2.2
`versus 5.3±2.1 mm, P<0.001). Skewness of final valve posi-
`tion was smaller for the S3-THV (3.5±3.2 versus 5.3±3.8%,
`P=0.034). The incidence of postdilatation was comparable
`among the groups: 15.9% for the S3-THV and 12.1% for the
`SXT-THV (P=0.55). An overview of procedural features is
`provided in Table 2.
`Although not statistically significant, initial angiogra-
`phy showed a clear trend toward lower Grade ≥2+ AR (4.5%
`
`Downloaded from
`
`http://circinterventions.ahajournals.org/
`
` by guest on June 22, 2017
`
`Page 03 of 20
`
`

`

`4
`
` Nijhoff et al
`
` Comparison of Sapien 3 and Sapien XT
`
`Table 1. Baseline Clinical Characteristics and Imaging Data
`
`Sapien 3 (n=44)
`
`Sapien XT (n=66)
`
`Weighted d
`
`Weighted p
`
`Age, y
`Male sex
`BMI, kg/m2
`BSA, m2
`Logistic EuroSCORE, %
`NYHA class III/IV
`Recent cardiac decompensation (within 3 mo)
`Diabetes mellitus
`Hypertension
`COPD
`Glomerular filtration rate, mL/min
`Renal impairment (GFR <60 mL/min)
`Dialysis
`Cerebrovascular disease
`Peripheral artery disease
`Coronary artery disease
`Prior myocardial infarction
`Prior PCI
`Prior CABG
`Prior BAV
`Atrial fibrillation
`First degree atrioventricular block*
`Right bundle branch block*
`Left bundle branch block*
`Prior pacemaker implantation
`Pulmonary hypertension
`Porcelain aorta
`Echocardiography data
`
` LVEF, %
`
` Moderate to severe left ventricular dysfunction†
`
` Peak aortic gradient, mm Hg
`
` Mean aortic gradient, mm Hg
`
` Indexed aortic valve area, cm2/m2
`
` Systolic pressure of pulmonary artery, mm Hg‡
`
` Moderate or severe aortic regurgitation
`
` Moderate or severe mitral regurgitation
`Multislice computed tomography data
`
` Aortic annulus minor diameter, mm
`
` Aortic annulus major diameter, mm
`
` Aortic annulus mean diameter, mm
`
` Eccentricity index§
`
` Aortic annulus area, mm2
`
` Aortic annulus area derived diameter, mm
`
`80.3±8.4
`19 (43.2)
`25.5±4.0
`1.82±0.20
`15.8 [9.9–23.4]
`22 (50.0)
`10 (22.7)
`14 (31.8)
`26 (59.1)
`10 (22.7)
`63.6±22.3
`23 (52.3)
`0
`10 (22.7)
`14 (31.8)
`23 (52.3)
`5 (11.4)
`16 (36.4)
`6 (13.6)
`0
`11 (25.0)
`7/41 (17.1)
`2/41 (4.9)
`3/41 (7.3)
`3 (6.8)
`4 (9.1)
`4 (9.1)
`
`50.0±11.4
`10 (22.7)
`65.4±21.3
`38.9±15.0
`0.41±0.11
`38.8±14.3
`5 (11.4)
`11 (25.0)
`
`80.5±7.8
`26 (39.4)
`25.6±3.8
`1.80±0.18
`16.5 [11.4–22.2]
`42 (63.6)
`15 (22.7)
`23 (34.8)
`39 (59.1)
`14 (21.2)
`64.5±22.5
`30 (45.5)
`0
`15(22.7)
`18 (27.2)
`33 (50.0)
`13 (19.7)
`23 (34.8)
`8 (12.1)
`3 (4.5)
`19 (28.8)
`6/57 (10.5)
`4/57 (7.0)
`5/57 (8.8)
`9 (13.6)
`4 (6.1)
`10 (15.2)
`
`49.6±13.0
`16 (24.2)
`68.6±21.4
`40.7±15.0
`0.38±0.09
`41.6±11.2
`9 (13.6)
`14 (21.2)
`
`22.0±2.6
`27.7±2.8
`24.8±2.1
`0.21 [0.16–0.23]
`486±76
`24.8±2.0
`
`21.0±2.0
`26.6±2.2
`23.8±2.0
`0.21 [0.19–0.24]
`445±71
`23.7±1.9
`
`0.002
`0.043
`−0.096
`0.109
`−0.070
`−0.163
`0.000
`0.000
`0.000
`0.038
`−0.039
`0.163
`
`0.000
`0.045
`0.048
`−0.168
`0.034
`0.000
`−0.197
`0.038
`0.136
`−0.122
`−0.039
`−0.160
`−0.074
`−0.094
`
`0.023
`−0.066
`−0.174
`−0.191
`0.434
`−0.149
`−0.041
`0.066
`
`0.480
`0.560
`0.459
`−0.212
`0.519
`0.520
`
`0.99
`0.83
`0.99
`0.61
`0.90
`0.28
`1.00
`1.00
`1.00
`0.80
`0.86
`0.29
`
`1.00
`0.76
`0.75
`0.26
`0.82
`1.00
`0.49
`0.80
`0.38
`1.00
`0.80
`0.48
`0.68
`0.74
`
`0.63
`0.67
`0.42
`0.50
`0.081
`0.57
`0.78
`0.80
`
`0.038
`0.011
`0.045
`0.36
`0.023
`0.023
`
`BAV indicates balloon aortic valvuloplasty; BMI, body mass index; BSA, body surface area; CABG, coronary artery bypass grafting;
`COPD, chronic obstructive pulmonary disease; GFR, glomerular filtration rate; LVEF, left ventricular ejection fraction; NYHA, New York Heart
`Association; and PCI, percutaneous coronary intervention.
`*As a percentage of patients without prior pacemaker implantation.
`†Defined as left ventricular ejection fraction ≤40%.
`‡Presented for 100 patients (n=46 for SXT and n=30 for S3).
`§Calculated as 1-(minor diameter/major diameter), where an index of 1.00 indicates a perfect circular shape.
`
`Downloaded from
`
`http://circinterventions.ahajournals.org/
`
` by guest on June 22, 2017
`
`Page 04 of 20
`
`

`

`5
`
` Nijhoff et al
`
` Comparison of Sapien 3 and Sapien XT
`
`Figure 2. Sizing data for Sapien 3 and Sapien XT prostheses. Area-derived aortic annulus diameter of the treated valves presented per
`prosthesis type and size (A). Percentage oversizing with respect to the aortic annulus area and the mean diameter for both prosthesis
`types (B). Data are presented as mean±SD. MSCT indicates multi-slice computed tomography.
`
`versus 15.1%, P=0.080) in the S3-THV group (Figure 3A).
`On final angiography, a significant difference in AR grade
`was observed, with lower Grade 1+ (13.6% versus 34.8%,
`P=0.002) and Grade 2+ AR (0% versus 6.1%, P=0.148) for the
`S3-THV (Figure 3B). No cases of grade 3 AR were recorded
`in any group. One case of coronary obstruction occurred in
`the S3-THV group (not because of the device, but because of
`compression of a diseased ostium of the right coronary artery
`
`by the deployment balloon), successfully treated with imme-
`diate PCI with stenting. Emergency surgery was required
`in one SXT-THV patient because of cardiac tamponade not
`resolved by subxiphoid pericardiocentesis.
`All patients in the S3-THV and 65 (98.5%) in the SXT-
`THV group underwent predischarge TTE. Peak and mean aor-
`tic transvalvular pressure gradients decreased significantly in
`both groups (both P<0.001), with no cases of residual stenosis.
`
`Table 2. Procedural Features
`
`Procedural access*
`
` Transfemoral
`
` Transapical/transaortic
`General anesthesia
`Use of cardiopulmonary bypass
`Predilatation
`Implanted prosthesis size
`
` 23-mm
`
` 26-mm
` 29-mm
`
`Deployment balloon underfilling (by 1 cc)
`Stent frame height extending below annulus plane, %
`Prosthesis implantation depth, mm
`Skewness of prosthesis final position, %†
`Post-dilatation
`Balloon area to prosthesis nominal area ratio
`Second valve implantation
`Valve malpositioning
`Coronary artery obstruction
`Conversion to cardiac surgery
`Intraprocedural death
`Fluoroscopy time, min
`Contrast volume, mL
`Acute device success
`
`Sapien 3 (n=44)
`
`Sapien XT (n=66)
`
`Weighted p
`
`34 (77.3)
`10 (22.7)
`12 (27.3)
`0
`43 (97.3)
`
`8 (18.2)
`24 (54.5)
`12 (27.3)
`0
`20.0±11.9
`4.0±2.2
`3.5±3.2
`7 (15.9)
`0.92 [0.92–0.92]
`0
`0
`1 (2.3)
`0
`0
`12.8 [10.0–14.6]
`130 [120–150]
`44 (100)
`
`52 (78.8)
`14 (21.2)
`15 (22.7)
`0
`66 (100)
`
`18 (27.3)
`32 (48.5)
`16 (24.2)
`3 (4.5)
`31.0±11.7
`5.3±2.1
`5.3±3.8
`8 (12.1)
`0.92 [0.92–1.00]
`0
`0
`0
`1 (1.5)
`0
`15.1 [12.0–18.3]
`150 [125–180]
`62 (93.9)
`
`0.95
`
`0.76
`
`1.00
`0.82
`
`0.49
`<0.001
`<0.001
`0.034
`0.55
`0.43
`
`1.00
`1.00
`
`0.045
`0.024
`0.32
`
`*The weighted d-value for procedural access was 0.008.
`†Difference in stent frame height extending below the annulus plane at both margins of the aortic annulus.
`
`Downloaded from
`
`http://circinterventions.ahajournals.org/
`
` by guest on June 22, 2017
`
`Page 05 of 20
`
`

`

`6
`
` Nijhoff et al
`
` Comparison of Sapien 3 and Sapien XT
`
`Figure 3. Aortic regurgitation (AR) evaluated by angiography and transthoracic echocardiography. Initial angiographic AR grade immedi-
`ately after valve implantation (A). Final angiographic AR grade following any AR reducing measures (B). Baseline echocardiographic AR
`severity of the native valve (C). Predischarge prosthetic AR severity on echocardiography (D).
`
`There was a significant difference in predischarge AR severity
`on TTE, with lower rates of mild AR (15.9% versus 38.5%,
`P=0.003) and moderate AR (0% versus 7.7%, P=0.075) in
`the S3-THV group (Figure 3D). No cases of severe AR were
`reported in any group. Figure 4 shows the comparative risk
`for ≥mild (or Grade I+) AR on angiography and predischarge
`TTE after TAVI. Exploratory univariable analysis for each
`valve type revealed percentage area and mean diameter over-
`sizing as the only factors associated with ≥mild AR (Table IV
`in the Data Supplement).
`In-hospital clinical outcomes are presented in Table 3.
`Numeric lower rates of life-threatening bleeding (2.3% ver-
`sus 12.1%, P=0.083), major bleeding (2.3% versus 6.1%,
`P=0.70), and major vascular complications (4.5% versus
`16.7%, P=0.044) were observed for the S3-THV. No life-
`threatening or major bleedings neither major vascular com-
`plications were related to the transfemoral access site in the
`S3-THV group compared with 4 (7.7%) in the SXT-THV
`group (P=0.15). Pacemaker implantation rates were similar
`for both valves (9.8% versus 8.8%, P=0.94), and pacemaker
`indications exclusively comprised complete heart block. No
`differences in in-hospital and 30-day mortality were noted
`among the valves. In the S3-THV group, one in-hospital death
`was caused by hematothorax, leading to respiratory failure,
`whereas the second fatality was caused by postoperative mes-
`enteric ischemia related to preexisting severe atherosclerosis
`(both transapical cases). In the SXT-THV group, the in-hos-
`pital deaths (all transfemoral cases) were caused by delayed
`valve embolization toward the left ventricle with acute heart
`
`failure and recurrent apnea of unknown origin (stroke was
`ruled out). One additional death occurred within 30-days in
`the SXT-THV group because of the late sequelae of an intra-
`procedural disabling stroke (transfemoral case). Weighted
`outcome data for the matched Sapien XT cohort are provided
`in Table III in the Data Supplement.
`
`Discussion
`The purpose of the present study was to compare the out-
`comes of the new balloon-expandable S3-THV with the pre-
`ceding SXT-THV in TAVI for severe AS. All of the SXT-THV
`
`Figure 4. Comparative risk for ≥mild aortic regurgitation (AR)
`after transcatheter aortic valve implantation (TAVI) with the
`Sapien 3 versus Sapien XT.
`
`Downloaded from
`
`http://circinterventions.ahajournals.org/
`
` by guest on June 22, 2017
`
`Page 06 of 20
`
`

`

`7
`
` Nijhoff et al
`
` Comparison of Sapien 3 and Sapien XT
`
`Table 3. Predischarge Echocardiography and In-Hospital Clinical Outcome
`
`Sapien 3 (n=44)
`
`Sapien XT (n=66)
`
`Weighted p
`
`Echocardiography data
` Peak aortic gradient, mm Hg
`
` Mean aortic gradient, mm Hg
`
`Effective orifice area, cm2
`Aortic regurgitation
` ≥Mild
`
`
` Moderate
`Number of regurgitant jets
`
` 1
`
` 2
`
` 3
`Type of aortic regurgitation
`
` Paravalvular
`
` Central
`
` Mixed
`Moderate or severe mitral regurgitation
`In-hospital clinical outcome
`All-cause mortality
`Stroke or TIA
`
` Disabling stroke
`
` Non-disabling stroke
`
` TIA
`Myocardial infarction
`Bleeding
`
` Life-threatening
`
` Major
`
` Transfemoral TAVI access site related
`Vascular complication
`
` Major
`
` Transfemoral TAVI access site related
`Blood transfusions ≥1
`Acute kidney injury
`
` Any
` Stage 2 or 3
`
`Permanent pacemaker implantation*
`Any device failure requiring reintervention
`30-day all-cause mortality
`
`(n=44)
`18.7±6.0
`9.6±3.3
`1.66±0.30
`
`7 (15.9)
`0
`
`3/7 (42.9)
`3/7 (42.9)
`1/7 (14.3)
`
`7/7 (100)
`0
`0
`4 (9.1)
`
`2 (4.5)
`1 (2.3)
`0
`0
`1 (2.3)
`1 (2.3)
`
`1 (2.3)
`1 (2.3)
`0/34 (0)
`
`2 (4.5)
`0/34 (0)
`3 (6.8)
`
`3 (6.8)
`2 (4.5)
`4/41 (9.8)
`0
`2 (4.5)
`
`(n=65)
`17.1±6.6
`9.3±3.3
`1.58±0.26
`
`30 (46.2)
`5 (7.7)
`
`17/30 (56.7)
`11/30 (36.7)
`2/30 (6.7)
`
`29/30 (96.7)
`0
`1/30 (3.3)
`9 (13.6)
`
`2 (3.0)
`2 (3.0)
`1 (1.5)
`1 (1.5)
`0
`0
`
`8 (12.1)
`4 (6.1)
`4/52 (7.7)
`
`11 (16.7)
`4/52 (7.7)
`10 (15.2)
`
`4 (6.1)
`0
`5/57 (8.8)
`1 (1.5)
`3 (4.5)
`
`0.19
`0.58
`0.28
`
`0.003
`0.12
`0.53
`
`1.00
`
`0.50
`
`1.00
`1.00
`1.00
`1.00
`0.39
`1.00
`
`0.083
`0.70
`0.15
`
`0.044
`0.15
`0.18
`
`1.00
`0.49
`0.94
`1.00
`1.00
`
`TAVI indicates transcatheter aortic valve implantation; and TIA, transient ischemic attack.
`*As a percentage of patients without prior pacemaker implantation.
`
`procedures were state-of-the-art: using the Novaflex+ deploy-
`ment catheter and expandable sheaths. Main findings of this
`study are (1) a lower incidence of paravalvular AR for the
`S3-THV, in the presence of similar postdilatation rates and less
`extensive prosthesis to aortic annulus oversizing, (2) absence
`of major transfemoral access site bleedings and vascular com-
`plications for the S3-THV, and (3) similar pacemaker implan-
`tation rates for both valves.
`Although the outcomes of TAVI have improved over the
`past years, the technique remained sensitive to complications
`related to the arterial introduction of large bore sheaths and the
`
`implantation of valve prostheses without suturing under direct
`sight. Unsurprisingly, the pivotal PARTNER IA trial evaluat-
`ing TAVI with the first Sapien prosthesis against surgical aortic
`valve replacement showed higher rates of vascular complica-
`tions and paravalvular AR.2 Important progress in reducing
`vascular adverse events was made with the introduction of the
`subsequent SXT-THV, as demonstrated in the PARTNER IIB
`trial.16 Unfortunately, the amount of downsizing established
`with the SXT-THV system did not reduce the rate of stroke,
`nor did prosthesis design positively impact paravalvular leak-
`age.16 Therefore, the S3-THV has been developed: a new THV
`
`Downloaded from
`
`http://circinterventions.ahajournals.org/
`
` by guest on June 22, 2017
`
`Page 07 of 20
`
`

`

`8
`
` Nijhoff et al
`
` Comparison of Sapien 3 and Sapien XT
`
`that not only incorporates further miniaturizing of the delivery
`system, but also includes an outer sealing cuff at the bottom
`of the stent frame to reduce paravalvular AR. In addition, it
`comes with a sophisticated delivery system that offers better
`flexing and fine-tuning abilities to improve coaxial engage-
`ment and precise valve positioning.5
`The first reports on S3-THV short-term performance indi-
`cate that the refinements to valve and delivery system design
`paid off.6,13,17 The prospective Sapien 3 safety and performance
`study demonstrated one of the lowest rates of stroke (2.7%),
`major vascular complications (5.3%), and mild (23.6%) and
`moderate AR (3.8%) to date in a combined population of trans-
`femoral and transapical TAVI.6 Our present results show similar
`trends for the S3-THV. More importantly, the S3-THV yielded
`results superior to the preceding SXT-THV in a matched popu-
`lation of TAVI patients, especially with respect to the incidence
`and severity of postprocedure AR.
`Ever since the introduction of TAVI, postprocedural AR
`has been a major limitation of this technique and source of
`criticism from conservative cardiac surgeons. Local anatomic
`challenges (heavy cusp and LVOT calcifications),18,19 but also
`inappropriate valve sizing, cause incomplete apposition of the
`prosthesis stent to the aortic annulus, resulting in paravalvular
`leaks. Initial data reported ≥moderate AR ranging from 2% to
`71% for balloon-expandable valves, achieved in the setting of
`echocardiographic valve sizing using unidirectional annular
`measurements.20 The introduction of more precise multislice
`computed tomography–guided sizing le