J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S
`ª 2 0 1 5 B Y T H E A M E R I C A N C O L L E G E O F C A R D I O L O G Y F O U N D A T I O N
`
`V O L . 8 , N O . 3 , 2 0 1 5
`
`I S S N 1 9 3 6 - 8 7 9 8 / $ 3 6 . 0 0
`
`P U B L I S H E D B Y E L S E V I E R I N C .
`
`h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j c i n . 2 0 1 4 . 1 0 . 0 1 4
`
`Incidence and Severity of Paravalvular
`Aortic Regurgitation With Multidetector
`Computed Tomography Nominal Area
`Oversizing or Undersizing After
`Transcatheter Heart Valve Replacement
`With the Sapien 3
`A Comparison With the Sapien XT
`
`Tae-Hyun Yang, MD,*y John G. Webb, MD,* Philipp Blanke, MD,* Danny Dvir, MD,* Nicolaj C. Hansson, MD,z
`Bjarne L. Nørgaard, MD,z Christopher R. Thompson, MD,* Martyn Thomas, MD,x Olaf Wendler, MD,k
`Alec Vahanian, MD,{ Dominique Himbert, MD,{ Susheel K. Kodali, MD,# Rebecca T. Hahn, MD,#
`Vinod H. Thourani, MD,** Gerhard Schymik, MD,yy Bruce Precious, MD,* Adam Berger, MD,* David A. Wood, MD,*
`Philippe Pibarot, MD,zz Josep Rodés-Cabau, MD,zz Wael A. Jaber, MD,xx Martin B. Leon, MD,#
`Thomas Walther, MD,kk Jonathon Leipsic, MD*
`
`ABSTRACT
`
`OBJECTIVES This study sought to compare the influence of the extent of multidetector computed tomography (MDCT)
`area oversizing on the incidence of paravalvular aortic regurgitation (PAR) between the Sapien 3 and the Sapien XT
`transcatheter heart valve (THV) to define a new MDCT sizing guideline suitable for the Sapien 3 platform.
`
`BACKGROUND The inverse relationship of PAR occurrence and oversizing has been demonstrated for the Sapien XT but
`the incidence of PAR with comparable oversizing with the Sapien 3 is not known.
`
`METHODS Sixty-one prospectively enrolled patients who underwent transcatheter aortic valve replacement with the
`Sapien 3 THV were compared with 92 patients who underwent transcatheter aortic valve replacement with the Sapien XT
`THV. Patients were categorized depending on the degree of MDCT area oversizing percentage: undersizing (below 0%),
`0% to 5%, 5% to 10%, and above 10%. The primary endpoint was mild or greater PAR on transthoracic echocardiography.
`
`RESULTS Mild or greater PAR was present in 19.7% of patients (12 of 61) in the Sapien 3 group and in 54.3% of patients
`(50 of 92) in the Sapien XT group (p < 0.01). The Sapien 3 group, compared with the Sapien XT group, consistently
`demonstrated significantly lower rates of mild or greater PAR except for oversizing >10% (p for interaction ¼ 0.54).
`Moderate or severe PAR rates were also lower in the Sapien 3 group than in the Sapien XT group (3.3% vs. 13.0%, p ¼
`0.04). In the Sapien 3 group, a MDCT area oversizing percentage value of #4.17% was identified as the optimal cutoff
`value to discriminate patients with or without mild or greater PAR.
`
`CONCLUSIONS Our retrospective analysis suggests that the Sapien 3 THV displays significantly lower rates of PAR than
`does the Sapien XT THV. A lesser degree of MDCT area oversizing may be employed for this new balloon-expandable
`THV. (J Am Coll Cardiol Intv 2015;8:462–71) © 2015 by the American College of Cardiology Foundation.
`
`From the *Department of Medical Imaging and Division of Cardiology, St. Paul’s Hospital, University of British Columbia, Van-
`couver, British Columbia, Canada; yInje University Busan Paik Hospital, Busan, South Korea; zDepartment of Cardiology, Aarhus
`University Hospital Skejby, Aarhus, Denmark; xDepartment of Cardiology, St. Thomas’s Hospital, London, United Kingdom;
`kDepartment of Surgery, King’s College Hospital/King’s Health Partners, London, United Kingdom; {Department of Cardiology,
`Bichat Hospital, Paris, France; #Department of Cardiology, Columbia University Medical Center/New York Presbyterian Hospital,
`New York, New York; **Department of Surgery, Emory University School of Medicine, Atlanta, Georgia; yyDepartment of
`
`Downloaded From: http://interventions.onlinejacc.org/ on 08/03/2016
`
`Page 01 of 10
`
`

`

`J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 8 , N O . 3 , 2 0 1 5
`
`M A R C H 2 0 1 5 : 4 6 2 – 7 1
`
`Yang et al.
`Incidence of PAR With the SAPIEN 3
`
`463
`
`METHODS
`
`A B B R E V I A T I O N S
`
`A N D A C R O N Y M S
`
`AUC = area under the curve
`
`CI = confidence interval
`
`MDCT = multidetector
`computed tomography
`
`OR = odds ratio
`
`PAR = paravalvular aortic
`regurgitation
`
`TAVR = transcatheter aortic
`valve replacement
`
`THV = transcatheter heart
`valve(s)
`
`STUDY POPULATION. Patients enrolled in
`the prospective safety and performance study
`of the Edwards Sapien 3 THV were considered
`for this analysis. The SAPIEN 3 THV study
`was a nonrandomized, prospective multi-
`center study that consisted of 150 patients
`with severe symptomatic calcific aortic valve
`stenosis (intermediate or high operative risk)
`at 15 participating centers across Canada,
`France, Germany,
`Italy, and the United
`Kingdom. Patients in whom the THV was
`implanted too high or low were excluded
`from the current study (6,7). Patients who did not
`undergo pre-TAVR MDCT or post-TAVR echocardiog-
`raphy, patients with nondiagnostic image quality on
`pre-TAVR MDCT, and patients without systolic image
`data on pre-TAVR MDCT were also excluded. At the
`time of this analysis, 135 of the anticipated 150 pa-
`tients had been enrolled into the SAPIEN 3 THV study
`of whom 61 patients met the inclusion criteria for this
`analysis from January 1, 2013 to September 30, 2013
`(Figure 1). For comparison, an historical cohort of 92
`patients who underwent TAVR with the Sapien XT
`THV between January 1, 2011 and June 30, 2012 was
`investigated. These 92 patients were derived from our
`previous study to analyze MDCT annular dimension
`for the prediction of PAR (6). An experienced inter-
`ventional cardiologist (J.W.) who was unaware of the
`grade of PAR, THV size, and MDCT annular dimension
`reviewed the THV position on pre- and post-implant
`aortic root angiography. Written consent was ob-
`tained from all patients.
`THV SIZE SELECTION. At each participating center,
`the THV size selection was at the discretion of the
`operators who were aware of
`the MDCT size
`
`T ranscatheter
`
`replacement
`valve
`aortic
`(TAVR) has become an effective treatment
`of severe aortic stenosis in patients consid-
`ered to be of high perioperative mortality risk (1)
`and those deemed to be high risk for surgery (2,3).
`Despite continuous procedural refinements, para-
`valvular aortic regurgitation (PAR) and aortic root
`injury remain important limitations in the widening
`of its application to lower-risk patients (4,5). Appro-
`priate sizing of the transcatheter heart valve (THV)
`is critically important to minimize the incidence of
`PAR. Moderate to severe PAR has been shown
`to be an independent predictor of mortality (3,5).
`Multidetector computed tomography (MDCT) has
`recently been shown to be predictive of PAR owing
`to its 3-dimensional capabilities and better apprecia-
`tion of
`the noncircular annular geometry (6,7).
`Importantly,
`integration of a MDCT annulus area
`sizing algorithm reduced PAR in patients receiving
`Sapien XT THV in a prospective multicenter ap-
`proach (8). However, it is well accepted that a single
`sizing algorithm for all valve platforms is not accept-
`able and that sizing needs to be refined depending
`on valve design.
`The Sapien 3 THV has been recently introduced
`with an external sealing cuff and improved delivery
`catheter with the goal of reducing PAR by means of its
`enhanced paravalvular sealing and more accurate
`positioning. In the first case series, an MDCT sizing
`algorithm designed for a previous valve generation
`was used (9); however, it is unclear whether this is
`appropriate. We herein sought to compare the in-
`cidences of PAR with comparable MDCT area over-
`sizing between the Sapien 3 and the Sapien XT
`THV and to serve as an exploratory analysis to
`help define a new MDCT sizing guideline for the
`Sapien 3 platform.
`
`Cardiology, Städtisches Klinikum Karlsruhe, Karlsruhe, Germany; zzDepartment of Cardiology, Quebec Heart and Lung Institute,
`Laval University, Quebec City, Quebec, Canada; xxCleveland Clinic, Cleveland, Ohio; and the kkDepartment of Surgery, Kerckhoff
`Heart Center, Bad Nauheim, Germany. Dr. Yang received unrestricted research grants from the Inje Research and Scholarship
`Foundation in 2013. Dr. Webb is a consultant to Edwards Lifesciences. Dr. Hansson has received grant support from Edwards
`Lifesciences. Dr. Nørgaard has received unrestricted research grants from Siemens and Edwards Lifesciences; and is a consultant
`to Edwards Lifesciences. Dr. Thompson has received minor honoraria from Edwards Lifesciences. Dr. Thomas is a consultant to
`Edwards Lifesciences. Dr. Wendler is a proctor for the Edwards Lifesciences Transcatheter Heart Valve program; and in the local
`principal investigator for the SAPIEN 3 trial. Dr. Vahanian serves on the advisory board of Medtronic; and has received speaking
`fees from Edwards Lifesciences. Dr. Himbert is a consultant and proctor for Edwards Lifesciences. Dr. Kodali is a consultant to
`Edwards Lifesciences and Meril; serves on the scientific advisory boards of Thubrikar Aortic Valve, Inc., and Meril; is a principal
`investigator for Claret Medical; has received research and grant support from Edwards Lifesciences and Claret Medical; and has
`equity in Thubrikar Aortic Valve Inc. Dr. Hahn is a speaker for and consultant to TC3. Dr. Thourani is a consultant to Edwards
`Lifesciences. Dr. Schymik is a proctor for Edwards Lifesciences. Dr. Wood is a consultant to Edwards Lifesciences. Dr. Pibarot has
`received a research grant from Edwards Lifesciences. Dr. Rodés-Cabau is a consultant to and has received a research grant from
`Edwards Lifesciences. Dr. Leon is a consultant to Edwards Lifesciences. All other authors have reported that they have no re-
`lationships relevant to the contents of this paper to disclose.
`
`Manuscript received July 21, 2014; revised manuscript received September 10, 2014, accepted October 8, 2014.
`
`Downloaded From: http://interventions.onlinejacc.org/ on 08/03/2016
`
`Page 02 of 10
`
`

`

`464
`
`Yang et al.
`Incidence of PAR With the SAPIEN 3
`
`J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 8 , N O . 3 , 2 0 1 5
`
`M A R C H 2 0 1 5 : 4 6 2 – 7 1
`
`F I G U R E 1 Study Flowchart
`
`The flowchart provides information about the included and excluded patients in the Sapien 3 and Sapien XT THV groups. MDCT ¼ multidetector
`computed tomography; TAVR ¼ transcatheter aortic valve replacement; THV ¼ transcatheter heart valve; TTE ¼ transthoracic
`echocardiography.
`
`recommendation provided by the core laboratory
`(J.L.),
`transesophageal echocardiography annular
`measurements, and vendor specifications.
`
`MDCT IMAGE ACQUISITION. MDCT protocols were
`applied according to each center’s standard practice.
`Scans were performed with electrocardiography-
`synchronized data acquisition with a requirement
`for systolic phase imaging.
`
`MDCT IMAGE ANALYSIS. All MDCT examinations
`were reviewed by 2 experienced cardiac CT readers
`(J.L. and T.H.Y.) at the MDCT core lab of the SAPIEN 3
`THV study (St. Paul Hospital, Vancouver, British
`Columbia, Canada). A double-oblique transverse view
`at transecting through the basal hinge points of
`3 aortic cusps was obtained, defining the aortic
`annulus plane (10). Planimetry of the annulus con-
`tours yielded area,
`long diameter, short diameter,
`and circumference. Annulus area was measured by
`manually tracking the luminal contours on double-
`oblique transverse plane. Annular measurements
`were performed in systole at 25% or 35% of the RR
`interval. When systolic imaging data was not avail-
`able, systolic annulus dimensions were modeled
`based on the findings of a previous study (11). Aortic
`
`annular eccentricity was calculated as: 1 – short
`diameter/long diameter. The annulus was separately
`analyzed for calcification. If present, the distribution
`of calcification and extension into the left ventricular
`outflow tract were also assessed in a semiquantitative
`fashion as
`follows: mild,
`1 nodule of calcium
`extending <5 mm in any direction and covering <10%
`of the perimeter of the annulus; moderate, 2 nodules
`of calcification or 1 extending >5 mm in any direction
`or covering >10% of the perimeter of the annulus;
`severe, multiple nodules of calcification of single
`focus extending >1 cm in length or covering >20% of
`the perimeter of the annulus (12). All image data were
`analyzed offline on a 3-dimensional workstation (AW
`4.4, GE Healthcare, Waukesha, Wisconsin, or Intui-
`tion, Terarecon, Foster City, California). By conven-
`tion, all measurements were repeated 2 times by 2
`cardiac CT readers with the final measurements rep-
`resenting the mean of both measurements.
`
`ECHOCARDIOGRAPHIC ASSESSMENT. Pre-TAVR trans-
`esophageal echocardiography and post-TAVR trans-
`thoracic echocardiography were performed by highly
`experienced echocardiographers at each participating
`center. Post-TAVR echocardiography was performed
`at 30 days after TAVR in the Sapien 3 group and at
`
`Downloaded From: http://interventions.onlinejacc.org/ on 08/03/2016
`
`Page 03 of 10
`
`

`

`J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 8 , N O . 3 , 2 0 1 5
`
`M A R C H 2 0 1 5 : 4 6 2 – 7 1
`
`Yang et al.
`Incidence of PAR With the SAPIEN 3
`
`465
`
`F I G U R E 2 Sapien 3 THV
`
`The nominal external valve area, diameter, and perimeter are shown for the Sapien 3 transcatheter heart valve (THV).
`
`discharge in the Sapien XT group. Echocardiogra-
`phies performed within a range of post-TAVR 21 to
`100 days in the Sapien XT group were also analyzed.
`Post-TAVR studies were interpreted by core labs
`blinded to the pre-TAVR annulus dimensions and
`implanted THV size (for
`the Sapien 3 THV, C5
`Cleveland Clinic, Cleveland, Ohio; for the Sapien XT
`group, St. Paul Hospital). PAR was graded in the
`same fashion as none, mild, moderate, and severe
`according to the Valve Academic Research Con-
`sortium 2 criteria (13).
`
`DEGREE OF AREA OVERSIZING OR UNDERSIZING.
`The nominal external valve area of an expanded Sa-
`pien 3 THV is 328 mm2 (20-mm device), 409 mm2
`(23-mm), 519 mm2 (26-mm), 649 mm2 (29-mm) ac-
`cording to the manufacturer (Figure 2) as compared to
`the Sapien XT THV with expanded areas of 314 mm2,
`415 mm2, 531 mm2, and 661 mm2, respectively. A THV
`was considered oversized when the THV nominal area
`was greater than the systolic MDCT annular area was.
`The percentage of oversizing (positive percentage) or
`undersizing (negative percentage) was calculated
`using the following formula: % oversizing ¼ (THV
`nominal area/MDCT annular area – 1)  100.
`
`STUDY ENDPOINTS. The primary endpoint was the
`presence of mild or greater PAR. The secondary
`endpoint was the presence of moderate or severe
`PAR.
`
`STATISTICAL ANALYSIS. The Student t test and the
`Mann-Whitney U test were used for comparison of
`continuous variables with normal or skewed distri-
`bution, when appropriate. The Fisher exact test was
`used to test for significant differences between cate-
`gorical variables. A priori, we categorized patients
`
`depending on the degree of MDCT area oversizing
`percentage in the following cohorts: undersizing
`(below 0%); 0% to 5%; 5% to 10%; and above 10%. The
`interactions among THV type implanted, MDCT area
`oversizing category, and rates of mild or greater PAR
`were analyzed using the multivariable model with
`addition of an interaction term of THV type  MDCT
`area oversizing variables. A receiver-operating char-
`acteristics curve analysis was performed to evaluate
`the discriminative abilities of MDCT annular mea-
`surements in predicting mild or greater PAR and to
`identify the cutoff level for MDCT area oversizing
`percentage in predicting mild or greater PAR. The
`relationship of PAR, annulus calcification, or area
`oversizing percentage was investigated by univariate
`or multivariate logistic regression analysis. A p
`value #0.05 was considered statistically significant.
`All statistical analyses were performed by using SPSS
`(version 18.0, SPSS Inc., Chicago, Illinois) software.
`
`RESULTS
`
`PATIENT POPULATION. Baseline clinical character-
`istics are shown in Table 1. Except for patient age,
`there was no significant difference in regard to clin-
`ical or demographic baseline characteristics.
`
`ANNULAR ASSESSMENT AND PROCEDURE. Annular
`dimensions and procedural characteristics are pre-
`sented in Table 2. Average oversizing was 6.5  11.0%
`in the Sapien 3 group (range –15.2 to 32.7%) and
`4.7  17.2% in the Sapien XT group (range –21.7 to
`59.6%) with no significant different between groups
`(p ¼ 0.42). Undersizing was present in 18 patients
`(29.5%) in the Sapien 3 group and in 41 patients
`(44.6%) in the Sapien XT group (p ¼ 0.07). MDCT area
`
`Downloaded From: http://interventions.onlinejacc.org/ on 08/03/2016
`
`Page 04 of 10
`
`

`

`466
`
`Yang et al.
`Incidence of PAR With the SAPIEN 3
`
`J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 8 , N O . 3 , 2 0 1 5
`
`M A R C H 2 0 1 5 : 4 6 2 – 7 1
`
`TABLE 1 Baseline Clinical Characteristics
`
`Age, yrs
`Male
`Height, cm
`Weight, kg
`Body mass index, kg/m2
`Diabetes
`Hypertension
`COPD
`Atrial fibrillation
`Peripheral vascular disease
`Smoking history
`Previous cerebrovascular
`accident
`Previous myocardial infarction
`Previous open heart surgery
`Previous permanent pacemaker
`GFR, ml/min
`STS PROM, %
`
`Sapien XT
`Sapien
`Group
`3 Group
`(n ¼ 92)
`(n ¼ 61)
`p Value
`84.6  5.2
`81.1  8.5 <0.01
`25 (41.0)
`48 (52.2)
`0.19
`164.4  10.1 164.9  25.0
`0.89
`74.9  17.7
`77.4  21.9
`0.50
`27.7  6.3
`26.9  6.1
`0.46
`19 (31.1)
`25 (27.2)
`0.72
`53 (86.9)
`71 (77.2)
`0.15
`16 (26.2)
`20 (21.7)
`0.56
`27 (44.3)
`34 (37.0)
`0.40
`11 (18.0)
`17 (18.5)
`1.00
`24 (39.3)
`31 (33.7)
`0.50
`12 (19.7)
`12 (13.0)
`0.36
`
`7 (11.5)
`12 (19.7)
`10 (16.4)
`58.1  26.7
`7.2  4.1
`
`15 (16.3)
`26 (28.3)
`8 (8.7)
`57.4  20.6
`6.2  2.9
`
`0.49
`0.26
`0.20
`0.87
`0.10
`
`Values are mean  SD or n (%).
`COPD ¼ chronic obstructive pulmonary disease; GFR ¼ glomerular filtration
`rate; STS PROM ¼ Society of Thoracic Surgeons predicted risk of mortality.
`
`TABLE 2 Annular Assessment and Procedural Characteristics
`
`Sapien 3 Group
`(n ¼ 61)
`
`Sapien XT Group
`(n ¼ 92)
`
`Pre-TAVR aortic regurgitation
`None/trivial
`35 (57.4)
`Mild
`25 (41.0)
`Moderate
`1 (1.6)
`Severe
`0 (0.0)
`Left ventricular ejection fraction, % 52.3  10.2
`44.7  14.3
`Mean gradient, mm Hg
`0.64  0.19
`Aortic valve area, cm2
`22.5  2.4
`MDCT short-axis diameter, mm
`26.6  2.2
`MDCT long-axis diameter, mm
`24.5  2.1
`MDCT mean-diameter, mm
`485.8  77.5
`MDCT annular area, cm2
`80.7  6.3
`MDCT annular circumference, mm
`0.15  0.07
`MDCT annular eccentricity
`Labeled prosthesis size
`20-mm
`23-mm
`26-mm
`29-mm
`MDCT nominal area oversizing, %
`Post-dilation
`Access type
`Transfemoral
`Transapical
`Transthoracic
`
`0 (0.0)
`16 (26.2)
`33 (54.1)
`12 (19.7)
`6.5  11.0
`2 (3.3)
`
`43 (70.5)
`15 (24.6)
`3 (4.9)
`
`79 (85.2)
`12 (13.0)
`1 (1.1)
`0 (0.0)
`55.0  12.9
`43.9  17.3
`0.69  0.17
`21.0  2.4
`26.8  2.9
`23.9  2.4
`467.9  90.6
`78.5  8.9
`0.21  0.08
`
`2 (2.2)
`41 (44.6)
`45 (48.9)
`4 (4.3)
`4.7  17.2
`9 (9.8)
`
`65 (70.7)
`27 (29.3)
`0 (0.0)
`
`p Value
`
`<0.01
`
`0.33
`0.77
`0.18
`<0.01
`0.71
`0.10
`0.21
`0.08
`<0.01
`<0.01
`
`0.42
`0.20
`0.10
`
`Values are n (%) or mean  SD.
`MDCT ¼ multidetector computed tomography; TAVR ¼ transcatheter aortic valve replacement.
`
`oversizing percentage of 0% to 5%, 5% to 10%, and
`above 10% was present in 10 patients (16.4%), 15
`(24.6%), and 18 (29.5%) in the Sapien 3 cohort and
`in 10 (10.9%), 11 (12.0%), and 30 (32.6%) in the Sapien
`XT cohort.
`
`PAR. Post-TAVR transthoracic echocardiography data
`is summarized in Table 3. Overall, mild, or greater
`PAR was present in 19.7% of patients (12 of 61) in the
`Sapien 3 group and in 54.3% of patients (50 of 92) in
`the Sapien XT group (p < 0.01). Figure 3 illustrates the
`rates of mild or greater PAR stratified by the extent of
`MDCT area oversizing percentage. The Sapien 3 group
`demonstrated significantly lower rates of mild or
`greater PAR compared with the Sapien XT group
`except for the subgroup with area oversizing above
`10% (p for interaction ¼ 0.54). In the Sapien 3 group,
`patients with 0% to 5% oversizing exhibited numeri-
`cally higher rates of mild or greater PAR without
`statistical significance compared with patients with
`5% to 10% oversizing (30.3% vs. 13.3%, p ¼ 0.36).
`Similarly, patients with undersizing exhibited slightly
`higher rates of mild or greater PAR compared with
`patients with oversizing (27.8% vs. 16.3%, p ¼ 0.31).
`We found a declining trend in the rates of mild or
`greater PAR in patients in the Sapien 3 group with
`increasing area oversizing, whereas the Sapien XT
`group provided little change (Figure 4). With regard to
`moderate or severe PAR, Sapien 3 patients demon-
`strated lower PAR rates than did Sapien XT patients
`throughout all subgroups: undersizing (0.0% vs.
`19.5%, p ¼ 0.04); area oversizing between 0% and 5%
`(10.0% vs. 20.0%, p ¼ 0.53); and area oversizing be-
`tween 5% and 10% (6.7% vs. 18.2%, p ¼ 0.36). No
`patients with either Sapien 3 or Sapien XT exhibited
`moderate or severe PAR when MDCT area oversizing
`percentage exceeded 10%. A total of 96 Sapien 3
`subjects, encompassing 61 patients with direct sys-
`tolic measurements as well as 35 patients with
`modeled data, also manifested similar findings in the
`rates of mild or greater and moderate or greater PAR
`when compared with all Sapien XT subjects (20.8%
`vs. 54.3%, p < 0.01 and 3.1% vs. 13.0%, p ¼ 0.02)
`(Figure 5). Likewise, the incidences of mild or greater
`and moderate or greater PAR were lower in the Sapien
`3 patients (n ¼ 61), which was consistent when
`compared with the Sapien XT patients with follow-up
`echocardiography conducted from 21 days to 100 days
`(n ¼ 61) (19.7% vs. 47.7%, p < 0.01 and 3.3% vs. 12.3%,
`p ¼ 0.06). Representative case examples are pre-
`sented in Figure 6.
`
`AREA UNDER THE CURVE. Sapien 3 THV nominal
`area and MDCT annular area provided weak predic-
`tion of mild or greater PAR (area under the curve
`
`Downloaded From: http://interventions.onlinejacc.org/ on 08/03/2016
`
`Page 05 of 10
`
`

`

`J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 8 , N O . 3 , 2 0 1 5
`
`M A R C H 2 0 1 5 : 4 6 2 – 7 1
`
`Yang et al.
`Incidence of PAR With the SAPIEN 3
`
`467
`
`[AUC]: 0.67, 95% confidence interval [CI]: 0.54 to
`0.78, p ¼ 0.04). The difference between the Sapien 3
`THV nominal diameter and MDCT annular mean
`diameter (AUC: 0.54, 95% CI: 0.40 to 0.67, p ¼ 0.68),
`the Sapien 3 THV nominal circumference and MDCT
`annular circumference (AUC: 0.58, 98% CI: 0.44 to
`0.71, p ¼ 0.44) did not show significant predictive
`value. In contrast, the relationship between the Sa-
`pien XT THV size and MDCT annular size ([area: AUC:
`0.73, 95% CI: 0.63 to 0.82, p < 0.01], [mean diameter:
`AUC: 0.74, 95% CI: 0.64 to 0.83, p < 0.01], circum-
`ference: AUC: 0.73, 95% CI: 0.63 to 0.81, p < 0.01])
`had moderate predictive value.
`
`THE RELATIONSHIP BETWEEN PAR AND ANNULUS
`CALCIFICATION. Mild or greater PAR occurred more
`frequently in patients with moderate or severe
`annulus calcification than in patients without mod-
`erate or severe annulus calcification in the Sapien 3
`group (odds ratio [OR]: 7.80, 95% CI: 1.95 to 31.22,
`p < 0.01), whereas this difference could not be found
`in the Sapien XT group (Table 4). Similarly, moderate
`or severe PAR was more common in patients having
`moderate or severe annulus calcification (11.1% vs.
`0.0%, p ¼ 0.08) in the Sapien 3 group. In the multi-
`variate logistic regression analysis, moderate or se-
`vere annulus calcification (OR: 10.0, 95% CI: 1.22 to
`81.48, p ¼ 0.03) was found to be significant and in-
`dependent predictors of mild or greater PAR after
`adjusting for age, sex, annulus eccentricity, and
`MDCT area oversizing percentage in the Sapien 3
`patients.
`
`THE IMPACT OF MINIMAL AREA OVERSIZING ON
`PAR. In the Sapien 3 group, an MDCT area oversizing
`percentage value of #4.17% was identified as the
`optimal cutoff value to discriminate patients with or
`without mild or greater PAR that provided a sensi-
`tivity of 66.7%, a specificity of 65.3%, a positive pre-
`dictive value of 32.0%, and a negative predictive
`value of 88.9%. We then dichotomized the area
`oversizing value at a practical value of 5.0%. For pa-
`tients with area oversizing below 5.0%, the rate of
`mild or greater PAR was 25.0% (7 of 28) versus 9.1%
`(3 of 33) (OR: 2.90, 95% CI: 0.77 to 10.95, p ¼ 0.11)
`when the area oversizing was above 5.0%. In the
`Sapien XT group, an optimal MDCT area oversizing
`cutoff value to predict mild or greater PAR was 7.27%
`(sensitivity: 84.0%, specificity: 61.9%, positive pre-
`dictive value: 72.4%, negative predictive value:
`76.5%). In the Sapien 3 subgroup (n ¼ 33) whose area
`oversizing was above 5%, if 2 patients with moderate
`or severe annulus calcification and mild or greater
`PAR were excluded from total 4 patients who had
`
`TABLE 3 Post-TAVR Outcomes
`
`Sapien 3 Group
`(n ¼ 61)
`
`Sapien XT Group
`(n ¼ 92)
`
`p Value
`
`Post-TAVR PAR
`None/trivial
`Mild
`Moderate
`Severe
`Mild or greater
`Mean gradient, mm Hg
`Aortic valve area, cm2
`Procedural mortality
`30-day mortality
`Annular rupture
`Device embolization
`Procedural myocardial infarction
`
`49 (80.3)
`10 (16.4)
`2 (3.3)
`0 (0.0)
`12 (19.7)
`10.9  5.5
`1.6  0.5
`0 (0.0)
`0 (0.0)
`0 (0.0)
`0 (0.0)
`0 (0.0)
`
`42 (45.7)
`38 (41.3)
`12 (13.0)
`0 (0.0)
`50 (54.3)
`11.0  4.4
`1.5 0.3
`0 (0.0)
`2 (2.2)
`0 (0.0)
`0 (0.0)
`1 (1.1)
`
`<0.01
`<0.01
`0.04
`1.00
`<0.01
`0.88
`0.31
`1.00
`0.24
`1.00
`1.00
`0.49
`
`Values are n (%) or mean  SD.
`PAR ¼ paravalvular aortic regurgitation; TAVR ¼ transcatheter aortic valve replacement.
`
`mild or greater PAR, only 2 patients (of 31 patients,
`6.5%) would have had mild or greater PAR.
`
`DISCUSSION
`
`In our comparison of the incidence and severity of
`PAR with MDCT nominal area oversizing or under-
`sizing after TAVR with the Sapien 3 or Sapien XT THV,
`the Sapien 3 THV appears to tolerate a lesser degree of
`area oversizing. Our current results demonstrate the
`following. 1) The Sapien 3 THV, compared with the
`
`F I G U R E 3 Rates of Mild or Greater PAR According to MDCT Nominal Area Oversizing
`Group in All Study Populations
`
`The numbers in the columns denote the absolute numbers and rates of mild or greater
`paravalvular aortic regurgitation (PAR). CI ¼ confidence interval; MDCT ¼ multidetector
`computed tomography; OR ¼ odds ratio.
`
`Downloaded From: http://interventions.onlinejacc.org/ on 08/03/2016
`
`Page 06 of 10
`
`

`

`468
`
`Yang et al.
`Incidence of PAR With the SAPIEN 3
`
`J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 8 , N O . 3 , 2 0 1 5
`
`M A R C H 2 0 1 5 : 4 6 2 – 7 1
`
`F I G U RE 4 Rates of Mild or Greater PAR Stratified by Extent of MDCT Nominal Area
`Oversizing in All Study Populations
`
`Each subgroup on the x-axis indicates all patients with area oversizing below each criterion
`and inclusive of undersizing. *All p values between the Sapien 3 group and Sapien XT group
`were <0.02 at all area oversizing criteria. Abbreviations as in Figures 1 and 3.
`
`Sapien XT, appears to exhibit significantly lower rates
`of mild or greater PAR when all minimal area over-
`sizing criteria are applied. 2) MDCT annular size
`measurements provide weaker prediction of mild or
`greater PAR after the Sapien 3 THV implantation than
`the Sapien XT THV implantation. 3) Annulus calcium
`plays a role in the occurrence of mild or greater PAR
`after the Sapien 3 THV implantation. 4) A lesser de-
`gree of MDCT area oversizing might be employed for
`
`F I G U RE 5 Rates of Mild or Greater PAR According to MDCT Nominal Area Oversizing
`Group in Patients With Systolic MDCT Measurements
`
`The numbers in the columns denote the absolute numbers and rates of mild or greater
`PAR. Abbreviations as in Figures 1 and 3.
`
`this new balloon-expandable THV than the currently
`recommended value for the Sapien XT THV.
`PAR continues to be an important complication
`after TAVR because the prognostic implications are
`not negligible (3,5). Even mild PAR has been associ-
`ated with a worse prognosis (3). A meta-analysis of 45
`studies also demonstrated that patients with mild
`PAR had a 1.8 higher risk of 1-year mortality than did
`those without this complication (14).
`
`DECREASED PAR WITH THE SAPIEN 3 THV.
`Two main causes of PAR with balloon-expandable
`prostheses are felt to be implantation of a THV that
`is relatively smaller than aortic annulus (undersizing)
`or positioning the device too high or too low (14,15).
`In the present study, the rate of mild or greater PAR
`was significantly lower in the Sapien 3 group than in
`the Sapien XT group except for patients with area
`oversizing above 10% where both groups had a low
`incidence of PAR. Possible reasons for this lower PAR
`rate are the outer polyethylene terephthalate sealing
`cuff, which enhances paravalvular sealing and more
`accurate positioning. The enhanced paravalvular
`sealing with the Sapien 3 THV was especially useful in
`patients with area oversizing below 10%.
`
`PREDICTION OF PAR. In the Sapien 3 group, AUC of
`MDCT annular measurements were not as high as
`those in the Sapien XT group. Two possible explana-
`tions for the discriminatory value of these MDCT
`measurements are as follow: 1) the differences be-
`tween oversizing percentage in patients without PAR
`and undersizing percentage in patients with PAR in
`the Sapien 3 group have become smaller than those in
`the Sapien XT group, possibly due to increased
`awareness of appropriate annular oversizing and the
`integration of a MDCT-based THV sizing algorithm;
`and 2) the occurrence of mild or greater PAR in the
`Sapien 3 group can be affected by factors other than
`THV undersizing or inappropriate THV positioning
`(i.e., annulus calcification).
`
`ANNULUS CALCIFICATION AS A POTENTIAL CAUSE
`OF PAR. Moderate or severe annulus calcification
`was associated with mild or greater PAR in the Sapien
`3 group. This finding is in line with previous studies
`demonstrating that annulus calcification is a risk
`factor for PAR in addition to THV undersizing and
`malpositioning (16). Therefore, it is important to un-
`derstand
`that
`appropriate
`annular
`oversizing,
`although crucial, is not the only factor that affects the
`development of PAR after TAVR even with the Sapien
`3 THV. The presence of anatomical modifiers such as
`annulus calcification should be considered particu-
`larly when the degree of annulus area oversizing is
`
`Downloaded From: http://interventions.onlinejacc.org/ on 08/03/2016
`
`Page 07 of 10
`
`

`

`J A C C : C A R D I O V A S C U L A R I N T E R V E N T I O N S V O L . 8 , N O . 3 , 2 0 1 5
`
`M A R C H 2 0 1 5 : 4 6 2 – 7 1
`
`Yang et al.
`Incidence of PAR With the SAPIEN 3
`
`469
`
`F I G U R E 6 Pre- and Post-TAVR Imaging in A Patient in the Sapien 3 Group With No Significant PAR and in A Patient in the Sapien XT
`Group With Moderate PAR
`
`(A) Double oblique transverse reformat of the annulus on MDCT with an area of 480 mm2 in an 82-year-old female patient in the Sapien 3
`group who had 8.1% area oversizing with the use of the 26-mm Sapien 3 THV. (B) The parasternal long-axis 2-dimensional echocardiography
`showing no significant PAR. (C) Double oblique transverse reformat of the annulus on MDCT with an area of 490 mm2 in a 69-year-old male
`patient in the Sapien XT group who had 8.4% area oversizing with the use of 26-mm Sapien XT. (D) The parasternal long-axis 2-dimensional
`echocardiography showing moderate degree of PAR. Abbreviations as in Figures 1 and 3.
`
`Interestingly, the impact of
`modest (below 10%).
`annular calcification appears to be greater for the
`Sapien 3 than for the Sapien XT. This most likely re-
`flects that the Sapien XT is more profoundly affected
`by modest oversizing, thereby muting the potential
`impact of annular calcification.
`
`THE OPTIMAL DEGREE OF AREA OVERSIZING IN THE
`SAPIEN 3 THV. It has been established that signifi-
`cant oversizing of a THV can minimize the risk of
`significant PAR; however,
`excessive oversizing
`comes at the expense of increasing the potential risk
`of aortic root rupture (12), coronary obstruction (17),
`periaortic hematoma, atrioventricular block, or ven-
`tricular septal
`rupture, whereas significant THV
`undersizing will increase the risk of significant PAR
`(6) and, less frequently, prosthesis embolization (18).
`Furthermore, treating significant PAR due to THV
`undersizing is challenging (19), so optimal sizing and
`appropriate selection of THV are crucial to reduce
`
`significant PAR and mechanical aortic root injuries. In
`our previous study, patients who underwent TAVR
`with the integration of MDCT-based area sizing algo-
`rithm had a significant reduction in greater than mild
`PAR (5.3%) as compared with those who underwent
`exclusively 2-dimensional echocardiography-based
`sizing (12.8%) (8). However, this positive finding was
`based on the annulus area sizing algorithm of the
`Sapien XT THV. Therefore, a new area sizing algo-
`rithm that is suitable for the Sapien 3 THV would be
`desirable.
`In the current study, the optimal cutoff value of
`MDCT area oversizing for the prediction of mild or
`greater PAR was lower in the Sapien 3 group than in
`the Sapien XT