Listen to this manuscript’s

audio summary by

JACC Editor-in-Chief

Dr. Valentin Fuster.

J O U R N A L O F T H E AM 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 V O L . 6 9 , N O . 1 8 , 2 0 1 7

ª 2 0 1 7 B Y T H E AM 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

P U B L I S H E D B Y E L S E V I E R

I S S N 0 7 3 5 - 1 0 9 7 / $ 3 6 . 0 0

h t t p : / / d x . d o i . o r g / 1 0 . 1 0 1 6 / j . j a c c . 2 0 1 7 . 0 2 . 0 5 7

ORIGINAL INVESTIGATIONS

Transcatheter Aortic Valve ImplantationWithin Degenerated AorticSurgical BioprosthesesPARTNER 2 Valve-in-Valve Registry

John G. Webb, MD,a Michael J. Mack, MD,b Jonathon M. White, MD,c Danny Dvir, MD,d Philipp Blanke, MD,e

Howard C. Herrmann, MD,f Jonathon Leipsic, MD,e Susheel K. Kodali, MD,g Raj Makkar, MD,h D. Craig Miller, MD,i

Philippe Pibarot, DVM, PHD,j Augusto Pichard, MD,k Lowell F. Satler, MD,k Lars Svensson, MD, PHD,l

Maria C. Alu, MS,g Rakesh M. Suri, MD, DPHIL,m Martin B. Leon, MDg

ABSTRACT

Fro

SudD

Bri

BACKGROUND Early experience with transcatheter aortic valve replacement (TAVR) within failed bioprosthetic

surgical aortic valves has shown that valve-in-valve (VIV) TAVR is a feasible therapeutic option with acceptable acute

procedural results.

OBJECTIVES Theauthors examined30-dayand 1-year outcomes ina large cohortof high-risk patientsundergoingVIVTAVR.

METHODS Patients with symptomatic degeneration of surgical aortic bioprostheses at high risk ($50% major morbidity

or mortality) for reoperative surgery were prospectively enrolled in the multicenter PARTNER (Placement of Aortic

Transcatheter Valves) 2 VIV trial and continued access registries.

RESULTS Valve-in-valve procedures were performed in 365 patients (96 initial registry, 269 continued access patients).

Mean age was 78.9 � 10.2 years, and mean Society of Thoracic Surgeons score was 9.1 � 4.7%. At 30 days, all-cause

mortality was 2.7%, stroke was 2.7%, major vascular complication was 4.1%, conversion to surgery was 0.6%, coronary

occlusion was 0.8%, and new pacemaker insertion was 1.9%. One-year all-cause mortality was 12.4%. Mortality fell

from the initial registry to the subsequent continued access registry, both at 30 days (8.2% vs. 0.7%, respectively;

p ¼ 0.0001) and at 1 year (19.7% vs. 9.8%, respectively; p ¼ 0.006). At 1 year, mean gradient was 17.6 mm Hg, and

effective orifice area was 1.16 cm2, with greater than mild paravalvular regurgitation of 1.9%. Left ventricular ejection

fraction increased (50.6% to 54.2%), and mass index decreased (135.7 to 117.6 g/m2), with reductions in both mitral

(34.9% vs. 12.7%) and tricuspid (31.8% vs. 21.2%) moderate or severe regurgitation (all p < 0.0001). Kansas City

Cardiomyopathy Questionnaire score increased (mean: 43.1 to 77.0) and 6-min walk test distance results increased

(mean: 163.6 to 252.3 m; both p < 0.0001).

CONCLUSIONS In high-risk patients, TAVR for bioprosthetic aortic valve failure is associated with relatively low

mortality and complication rates, improved hemodynamics, and excellent functional and quality-of-life outcomes

at 1 year. (The PARTNER II Trial: Placement of AoRTic TraNscathetER Valves [PARTNER II]; NCT01314313)

(J Am Coll Cardiol 2017;69:2253–62) © 2017 by the American College of Cardiology Foundation.

m the aDivision of Cardiology, St. Paul’s Hospital, Vancouver, British Columbia, Canada; bDepartment of Cardiothoracic

rgery, Baylor Scott and White Health, Plano, Texas; cDepartment of Cardiovascular Medicine, Cleveland Clinic, Cleveland, Ohio;

ivision of Cardiology, University of Washington, Seattle, Washington; eDepartment of Radiology, St. Paul’s Hospital, Vancouver,

tish Columbia, Canada; fPerelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania; gStructural

ABBR EV I A T I ON S

AND ACRONYMS

AS = aortic stenosis

EOA = effective orifice area

KCCQ = Kansas City

Cardiomyopathy Questionnaire

LVEF = left ventricular ejection

fraction

PPM = patient-prosthesis

mismatch

SAVR = surgical aortic valve

replacement

TAVR = transcatheter aortic

valve replacement

THV = transcatheter heart

valve

VIV = valve-in-valve

Heart and V

New York,

of Cardioth

Lung Instit

WashingtomDepartme

Transcathe

Vascular, a

Lifescience

consultant

consultants

Neovasc, G

in Thubrika

Abbott Vas

National H

Medical, an

is a consul

property w

Abbott and

that they h

Manuscript

Webb et al. J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7

THV Implantation in Degenerated Bioprostheses M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2

2254

S urgical aortic valve replacement(SAVR) has long been the gold stan-dard for the management of severe

aortic stenosis (AS). Recent years have seena trend towards the use of bioprostheses asa result of the increased risk of bleedingand thrombotic complications with mechani-cal prostheses (1–3). Combined with an agingpopulation, these factors are likely to resultin increasing numbers of patients presentingwith structural degeneration of aortic bio-prostheses. Reoperation for failed surgicalvalves carries important risks (4–6); how-ever, valve-in-valve (VIV) transcatheteraortic valve replacement (TAVR) has recentlyemerged as a less invasive treatment forpatients with degenerated bioprostheses (7).

We present the 1-year follow-up results of a largecohort of high-risk patients who underwent VIVTAVR for failing surgical aortic bioprostheses.

SEE PAGE 2263

METHODS

The PARTNER (Placement of Aortic TranscatheterValves) 2 trial was a prospective, multicenterstudy that enrolled patients with symptomatic AS.This study included a nested registry of patientswith degenerated surgical aortic bioprostheseswho were at high risk of complications duringreoperation. Following the enrollment of amaximum of 100 patients in the nested registry,additional patients were enrolled in a continuedaccess registry.

alve Center, Center for Interventional Vascular Therapy, Division

New York; hHeart Institute, Department of Medicine, Cedars-Sinai

oracic Surgery, Stanford University, Stanford, California; jDepartm

ute, Quebec City, Quebec, Canada; kDivision of Interventional

n, DC; lSydell and Arnold Miller Family Heart and Vascular In

nt of Thoracic and Cardiovascular Surgery, Cleveland Clinic, Cl

ter Valves) 2 trial was sponsored by Edwards Lifesciences. Dr. W

nd St. Jude Medical. Dr. Dvir is a consultant for Edwards Lifescien

s, St. Jude Medical, Medtronic, Boston Scientific, Abbott Vascula

for Edwards Lifesciences and Siemens; and holds equity in Micr

for Edwards Lifesciences and provide uncompensated CT core

DS, and Tendyne Holdings. Dr. Kodali has received grants from Ed

r Aortic Valve, Inc. Dr. Makkar has received grants from Edwards

cular, Cordis, and Medtronic; and holds equity in Entourage Me

eart Lung Blood Institute HL67025; and has received consulting

d Medtronic. Dr. Pibarot has uncompensated Core Lab contracts w

tant for Edwards Lifesciences. Dr. Svensson holds equity in Card

ith Postthorax. Ms. Alu is a consultant for Claret Medical. Dr. Su

St. Jude; is a consultant with Sorin and Abbott; and has a patent ap

ave no relationships relevant to the contents of this paper to dis

received December 19, 2016; revised manuscript received Febru

Patients were required to have a bioprosthesissuitable for VIV treatment with either a 23- or 26-mmSapien XT transcatheter heart valve (THV) (EdwardsLifesciences Corp., Irvine, California). Severe aorticstenosis was defined as an effective orifice area (EOA)of <0.8 cm2 or an indexed EOA of <0.5 cm2/m2 anda mean gradient of >40 mm Hg or peak velocityof>4m/s. Patients with at least moderate stenosis andregurgitation were classified as having mixed bio-prosthetic failure.

Patients were deemed to be at high risk if the heartteam considered the risk of surgical mortality ormajor morbidity to be $50%. Key exclusion criteriawere a bioprosthetic valve with a labeledsize <21 mm, more than mild paravalvular regurgi-tation, left ventricular ejection fraction (LVEF)of <20%, or an estimated life expectancy of <2 years.The complete list of inclusion and exclusion criteriaare shown in Online Table 1. All patients were pre-sented to a Web-based conference call where imagingand clinical data were reviewed by a screening com-mittee and approved. Transcatheter valve sizing wasdetermined by expert consensus, manufacturers’tables, and smartphone application-based sizing data(8,9). The trial was approved by the institutionalreview boards of all participating sites, and written,informed consent was provided by all patients.

The THV used was balloon expandable and con-sisted of bovine pericardial leaflets sutured to a cobaltchromium frame and a polyethylene terephthalatecuff that covered the lower portion of the frame. Itwas delivered through expandable 16-F (23-mm THV)or 18-F (26-mm THV) transfemoral delivery sheaths orby using transapical or transaortic access. Following

of Cardiology, Columbia University Medical Center,

Medical Center, Los Angeles, California; iDepartment

ent of Medicine, Laval University, Quebec Heart and

Cardiology, Medstar Washington Hospital Center,

stitute, Cleveland Clinic, Cleveland, Ohio; and the

eveland, Ohio. The PARTNER (Placement of Aortic

ebb is a consultant for Edwards Lifesciences, Abbott

ces. Dr. Herrmann has received grants from Edwards

r, Gore, Siemens, Cardiokinetix, and Mitraspan; is a

ointerventional Devices. Drs. Blanke and Leipsic are

lab services for Edwards Lifesciences, Medtronic,

wards Lifesciences and Medtronic; and holds equity

Lifesciences and St. Jude Medical; is a consultant for

dical. Dr. Miller is supported by research grant R01

fees and honoraria from Abbott Vascular, St. Jude

ith Edwards Lifesciences and Medtronic. Dr. Pichard

iosolution and Valvexchange; and holds intellectual

ri is a member of the clinical steering committee at

plication with Sorin. All other authors have reported

close.

ary 28, 2017, accepted February 28, 2017.

TABLE 1 Baseline Characteristics

InitialRegistry(n ¼ 96)

ContinuedAccessRegistry(n ¼ 269)

Overall(N ¼ 365) p Value

Age, yrs 80.1 � 9.3 78.5 � 10.5 78.9 � 10.2 0.18

Male 55.2 67.3 64.1 0.03

STS score* 9.9 � 5.1 8.8 � 4.6 9.1 � 4.7 0.058

Logistic EuroSCORE* 15.7 � 12.9 11.1 � 8.2 12.3 � 9.8 0.002

NYHA functionalclass III or IV

95.8 88.1 90.1 0.03

Coronary disease 67.7 70.3 69.6 0.64

Previous MI 10.4 14.1 13.2 0.36

Previous CABG 47.9 53.5 52.1 0.34

Previous PCI 17.7 25.7 23.6 0.12

Previous stroke 15.6 10.8 12.1 0.21

Peripheral vasculardisease

29.2 21.9 23.8 0.15

COPD 292 30.9 30.4

Creatinine >2 mg/dl 14.6 12.3 12.9 0.56

Atrial fibrillation 50.0 45.7 46.8 0.40

Permanentpacemaker

28.1 21.2 23.0 0.17

Severe PAH 14.6 13.9 14.0 0.86

Frailty 37.5 22.8 26.7 0.005

Porcelain aorta 7.3 6.0 6.3 0.65

Prior chest radiation 0.0 0.7 0.6 –

Chest wall deformity 2.1 1.5 1.7 0.66

Liver disease 9.4 6.7 7.4 0.39

LVEF, % 48.5 � 14.0 48.7 � 12.9 48.6 � 13.2 0.92

LVEF <40% 22.5 21.5 21.8 0.85

Values are mean � SD or %. *The STS (Society of Thoracic Surgeons) score andlogistic EuroSCORE (European System for Cardiac Operative Risk Evaluation)estimate the risk of 30-day mortality following cardiac surgery.

CABG ¼ coronary artery bypass grafting; COPD ¼ chronic obstructive pulmonarydisease; LVEF ¼ left ventricular ejection fraction; MI ¼ myocardial infarction;NYHA ¼ New York Heart Association; PAH ¼ pulmonary hypertension; PCI ¼percutaneous coronary intervention.

TABLE 2 Procedural Characteristics

Bioprosthesis age

<5 yrs 14/205 (6.8)

5–10 yrs 55/205 (26.8)

>10 yrs 136/205 (66.3)

Mode of bioprosthesis failure

Stenosis 197/357 (55.2)

Regurgitation 84/357 (23.5)

Mixed 76/357 (21.3)

Type of bioprosthesis

Stented 337/365 (92.3)

Stentless or homograft 22/365 (6.0)

Unknown 6/365 (1.6)

Bioprosthesis labeled size

#21 mm 96/354 (26.8)

23-25 mm 218/361 (60.4)

>25 mm 44/361 (12.2)

THV size

23 mm 252/365 (69.0)

26 mm 113/365 (31.0)

Access

Transfemoral 273/362 (75.4)

Transapical 87/362 (24.0)

Transaortic 2/362 (0.6)

Conscious sedation 44/365 (12.1)

General anesthesia 321/365 (87.9)

Anesthesia duration, min 201.1 � 60.0

Total procedure time, min 92.7 � 39.1

Fluoroscopy time, min 19.7 � 13.5

Post-dilation 37/365 (10.2)

Second valve required 7/365 (1.9)

Coronary obstruction 3/365 (0.8)

Urgent cardiac surgery 2/365 (0.6)

IABP inserted 4/365 (1.1)

Cardiopulmonary bypass 4/365 (1.1)

Median hospital stay, days 5 (3-8)

Values are n/N (%), mean � SD, or median (interquartile range).

IABP ¼ intra-aortic balloon pump; THV ¼ transcatheter heart valve.

J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7 Webb et al.M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2 THV Implantation in Degenerated Bioprostheses

2255

the procedure, treatment with aspirin and clopidogrelfor 6 months was recommended.

The trial was designed by members of the execu-tive steering committee and the sponsor. The firstauthor and coprincipal investigators had unrestrictedaccess to the data. The sponsor had no role in dataanalysis or drafting of the manuscript.

Clinical assessments were performed at baselineand at all subsequent follow-up time points andincluded formal examination by a neurologist. Serialechocardiograms (intraprocedural, within 24 h ofdischarge and at 30 days) were analyzed indepen-dently by a core laboratory. A clinical events com-mittee adjudicated all clinical events, and a data andsafety monitoring board reviewed all adverse events.

The primary endpoint was all-cause mortality at1 year. Nonpowered secondary endpoints includedmajor vascular complications, stroke, acute kidneyinjury (according to Valve Academic Research

Consortium-2 criteria), new permanent pacemakerinsertion, myocardial infarction, and clinicalimprovements in symptoms, quality of life (QOL),Kansas City Cardiomyopathy Questionnaire (KCCQ)responses, and functional status (6-min walk test).Endpoints are defined in Online Table 2.

STATISTICAL ANALYSIS. Data analysis of the valveimplant population was conducted. Categoricalvariables were reported as percentages and comparedusing the chi-square or Fisher exact test, whereappropriate. Continuous variables were reported asmean � SD and compared using Student t test.Longitudinal data were modeled using a linear mixedmodel over the following time points: baseline,30 days, and 1 year. The model included time as a fixedeffect and assumed a compound symmetry covariancestructure for observations within a patient. Results are

TABLE 3 30-Day and 1-Year Clinical Outcomes

30 Days 1 Year

Death

Any cause 10 (2.7) 43 (12.4)

Cardiovascular 9 (2.5) 31 (9.0)

Repeat hospitalization 21 (5.9) 53 (15.9)

Stroke (all) 10 (2.7) 16 (4.5)

MI

All 5 (1.4) 5 (1.4)

Periprocedural 5 (1.4) 5 (1.4)

Vascular complications

All 27 (7.4) 28 (7.7)

Major 15 (4.1) 16 (4.4)

Acute kidney injury 27 (7.5) 31 (8.7)

Stage 1 2 (0.5) 2 (0.5)

Stage 2 23 (6.4) 26 (7.3)

Stage 3 2 (0.5) 3 (0.8)

Life-threatening or major bleeding 76 (20.8) 84 (23.2)

New pacemaker 7 (1.9) 9 (2.6)

Values are n (%); all percentages are Kaplan-Meier estimates.

MI ¼ myocardial infarction.

FIGURE 1 All-Caus

All-

Caus

e M

orta

lity

(%)

50

40

10

20

30

0

2Initial RegistryContinued Access

Number at risk:

Comparing the initia

registry, there was a

the later continued a

Webb et al. J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7

THV Implantation in Degenerated Bioprostheses M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2

2256

presented as least squares means with 95% confidenceintervals (CI). Cumulative incidence graphs andKaplan-Meier estimates reflect time-to-event out-comes. Comparisons are based on the log-rank testresults. A multivariate Cox proportional hazardregression model was used to assess the adjustedassociation betweenmortality and risk factors (Society

e Mortality

Initial Registry Continued Access Registry

Time in Months123 6

HR: 2.29 [95% CI: 1.25, 4.18]Log-Rank P-Value =0.0055

8.3%

0.7%

90

77168

9.8%

19.8%

85256

80230

77214

9669

l nested registry population with the extended continued access

significant reduction in all-cause mortality at 30 days and 1 year in

ccess experience. CI ¼ confidence interval; HR ¼ hazard ratio.

of Thoracic Surgeons [STS] score, labeled valve size,THV size, mean gradient $20 mm Hg, and severepatient-prosthesis mismatch [PPM]). All statisticalanalyses were performed using SAS version 9.4software (SAS Institute Inc., Cary, North Carolina).

RESULTS

Between June 2012 and December 2014, 367 patientswere enrolled in both the initial registry and thecontinued access registry at 34 sites. Two patients didnot undergo the procedure (1 withdrew consent, andthe other was deemed inappropriate for the studyvalve), leaving 365 patients who underwent VIVprocedures and are included in this analysis (96 in theinitial registry, 269 in the continued access patients).At 1 year, no patients were lost to follow-up (OnlineFigure 1).

PATIENT AND PROCEDURAL CHARACTERISTICS.

Mean age was 78.9 � 10.2 years, and mean STS scorewas 9.1 � 4.7%. The ages of the surgical bioprostheseswere <5 years in 6.8%, 5 to 10 years in 26.8%, and >10years in 66.3% of patients. Initial registry patients,compared to those in continued access, had similarSTS scores but higher mean logistic EuroSCOREs(15.7 vs. 11.1, respectively; p ¼ 0.002) and weremore often frail (37.5% vs. 22.8%, respectively;p ¼ 0.005), but no other significant differences wereobserved. Other baseline characteristics are shown inTable 1.

The labeled surgical valve sizes were <23 mm in27.1%, 23 to 25 mm in 60.5%, and >25 mm in 12.4%.Pre-procedural computed tomography was performedin 90 patients (24.7%), and transesophageal echocar-diography (either pre-procedural or intraprocedural)was performed in 351 patients (96.2%). ImplantedTHV sizes were 23 mm in 69% and 26 mm in 31%. Theexisting bioprostheses were stented in 92.3% ofcases, stentless or a homograft in 6.0% of cases, andof unknown type in 1.6% of cases. The predominantfailure modes were stenosis in 55.2%, regurgitation in23.5%, and mixed in 21.3% of cases.

Transfemoral access was used in 75.4% (60.6%initial vs. 80.6% continued access; p ¼ 0.0001). In thecombined population of 365 patients, general anes-thesia was used in 321 patients (87.9%). Post-dilationwas performed in 37 of patients (10.1%). Valveembolization occurred in no patients, more than 1THV was required in 7 patients (1.9%), coronary arteryocclusion was required in 3 patients (0.8%), andconversion to open surgery was required in 2 patients(0.6%). Intra-aortic balloon pump counterpulsationand cardiopulmonary bypass were each required in4 patients (1.1%). No patients died during the

FIGURE 2 Stratified All-Cause Mortality Curves

50

40

Valve Size 21mm Valve Size > 21mm Mean Gradient ≥ 20mm Hg Mean Gradient < 20mm Hg

All-

Caus

e M

orta

lity

(%)

Time in Months

10

20

30

0123 6

HR: 1.39 [95% CI: 0.73, 2.63]Log-Rank P-Value =0.3156

90

59178

11.4%

15.5%

87243

80220

74207

95259

21mm> 21mm

Number at risk:

50

40

All-

Caus

e M

orta

lity

(%)

Time in Months

10

20

30

0123 6

HR: 2.27 [95% CI: 1.16, 4.46]Log-Rank P-Value =0.0140

90

72155

7.7%

16.7%

107210

94194

85185

114218

≥ 20mm Hg< 20mm Hg

Number at risk:

Severe PPM No PPMModerate PPM

30

Deat

h (%

)

Time in Months

10

20

0123 6

Log-Rank P-Value = 0.8617

90

1106125

10.3%9.7%6.9%

1558229

1437426

1336926

1628530

Sev PPMMod PPMNo PPM

Number at risk:

A B

C

Although there were no significant differences in mortality comparing (A) previous surgical valve sizes, (B) mortality was significantly greater at 1 year with a larger

post-procedural mean gradient. (C) Trends toward increased mortality in patients with PPM were not significant. PPM ¼ patient-prosthesis mismatch; other abbre-

viations as in Figure 1.

J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7 Webb et al.M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2 THV Implantation in Degenerated Bioprostheses

2257

procedure. The median length of intensive care unitstay was 1 day (interquartile range [IQR]: 1 to 2 days),and the median length of hospital stay was 5 days(IQR: 3 to 8 days). Other procedural characteristics areshown in Table 2.

30-DAY OUTCOMES. The rate of 30-day all-causemortality was 2.7% (initial registry: 8.3%; continuedaccess patients: 0.7%; p < 0.0001) and cardiovasculardeath was 2.5%. The rate of all stroke at 30 days was2.7%, and disabling stroke was 2.2% (modified Rankinscore: $2). Rehospitalization occurred in 5.9% of

patients, major bleeding in 14.6%, major vascularcomplications in 4.1%, and a new permanentpacemaker was inserted in 1.9% of patients. Otherclinical events are shown in Table 3. Significantly moremajor bleeding occurred in transapical than in trans-femoral access patients (24.8% vs. 11.4%, respectively;p ¼ 0.001), but no other significant differences in earlyoutcomes were observed after comparing thesecohorts.

Thirty-day echocardiographic evaluation was per-formed in 327 patients (90%; data not shown). Frombaseline to 30 days, mean EOA increased from0.93 cm2

FIGURE 3 Change

40

35

30

mm

Hg

10

5

20

15

25

0Baseline

Change from baselin

which then remained

effective orifice area

TABLE 4 Echocardiographic Outcomes

Baseline(n ¼ 353)

1 Year(n ¼ 232)

Difference(1 Year–Baseline) p Value*

EOA, cm2 0.93 (0.89–0.98) 1.16 (1.11–1.21) 0.23 <0.0001

EOA index, cm/m2 0.49 (0.47–0.51) 0.60 (0.57–0.63) 0.11 <0.0001

Mean gradient, mm Hg 35.0 (33.7–36.2) 17.6 (16.2–19.1) �17.4 <0.0001

LVEF, % 50.6 (49.0–52.1) 54.2 (52.3–56.1) 3.6 <0.0001

LV stroke volumeindex, ml/m2

41.1 � 12.0 34.8 � 9.9 �6.3 <0.0001

LV mass index, g/m2 135.7 (131.9–139.5) 117.6 (113.3–121.8) �18.1 <0.0001

Total AR

None 29/247 (11.7) 67/106 (63.2)

Trace 46/247 (18.6) 32/106 (30.2)

Mild 64/247 (25.9) 5/106 (4.7) NA NA

Moderate 67/247 (27.1) 2/106 (1.9)

Severe 41/247 (16.6) 0 (0.0)

Paravalvular AR

None NA 72/105 (68.6)

Trace NA 27/105 (25.7)

Mild NA 5/105 (4.8) NA NA

Moderate NA 1/105 (1.0)

Severe NA 0 (0.0)

Moderate or severe MR 84/241 (34.9) 13/102 (12.7) NA NA

Moderate or severe TR 75/236 (31.8) 22/104 (21.2) NA NA

Values are median (interquartile range), mean � SD or n/N (%). *Change over time.

AR ¼ aortic regurgitation; CI ¼ confidence interval; EOA ¼ effective orifice area; LV ¼ left ventricular;LVEF ¼ left ventricular ejection fraction; MR ¼ mitral regurgitation; NA ¼ not applicable; TR ¼ tricuspidregurgitation.

Webb et al. J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7

THV Implantation in Degenerated Bioprostheses M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2

2258

(95% CI: 0.89 to 0.98) to 1.13 cm2 (95% CI: 1.09 to1.18), indexed EOA increased from 0.49 cm2/m2 (95%CI: 0.47 to 0.51) to 0.60 cm2/m2 (95% CI: 0.57 to 0.62),and mean gradient decreased from 35.0 mm Hg

in EOA and Mean Gradient

1.4

1.2

1

cm2

0.2

0.6

0.4

0.8

01 Year30 Days

p=0.90

p=0.30

p<0.0001

p<0.0001

AV Mean Gradient AV Area (EOA)

e to 30 days was significant for both EOA and mean AV gradient,

relatively unchanged at 1-year follow-up. AV ¼ aortic valve; EOA ¼.

(95% CI: 33.7 to 36.2) to 17.7 mm Hg (16.5 to 19.0; allp < 0.0001).

Improvements in EOA (þ0.33 cm2 vs. þ0.06 cm2,respectively; p < 0.0001) and mean gradient(�21.7 mm Hg vs. �11.9 mm Hg; p < 0.0001) weremore marked in patients with AS than in patients withaortic regurgitation (mixed or predominant). At30 days, 96.8% had mild or less aortic regurgitation,and 3.2% had moderate or severe regurgitation (allparavalvular). No change in LVEF was seen betweenbaseline and 30 days (50.6% vs. 50.1%, respectively;p ¼ 0.52).

An elevated residual gradient ($20 mm Hg) wasobserved on the first post-procedural echocardiogramin 34.3%, with no significant difference observed incomparison between the smallest 21-mm surgicalvalves and the larger valves (41.2% vs. 32.4%, respec-tively; p¼0.14). Severe PPMwas defined as an indexedEOA of<0.65 cm2/m2 (or<0.60 for patients with a bodymass index $30) (10). On the first post-proceduralechocardiogram, severe PPM was present in 58.4% ofpatients. A significant difference was observed incomparison between 21-mm surgical valves and largervalves (69.4% vs. 55.0%, respectively; p ¼ 0.0327).When stratified by labeled surgical valve size, howev-er, all valve sizes showed an increase in EOA and adecrease in mean gradient from baseline to 30 days.

The true internal diameter (ID) of the surgical valve,as determined from published tables, was available in258 of 365 patients (71%), with a true ID of #20 mm in46.9%. This and the THV size were correlated withpost-TAVR gradient on 30-day echocardiography andwith PPM. Surgical valves with a true ID of <20 mmwere strongly associated with post-TAVR meangradient of >20 mm Hg and with severe PPM(p ¼ 0.0042 and p < 0.0001, respectively), as was aTHV size of <23 mm (p ¼ 0.0003 and < 0.0001,respectively). Neither surgical valve true ID nor THVsize, however, were predictive of 1-year mortality.

1-YEAR OUTCOMES. The 1-year overall Kaplan-Meiermortality estimate was 12.4%, and cardiac mortalitywas 9.0%. At 1 year, the rate of stroke was 4.5%,rehospitalization was 15.9%, and pacemaker insertionwas 2.6%. Substantially lower mortality was observedin continued access patients than in those inthe initial registry (9.8% vs. 19.8%, respectively;p ¼ 0.006) (Figure 1). Increased mortality was seen inpatients with an elevated ($20 mm Hg) post-TAVRgradient (16.7% vs. 7.7%, respectively; p ¼ 0.01). Noincreased mortality was observed in patients strati-fied according to mode of valve failure, access route,21-mm surgical valves, or severe PPM (Figure 2), andmultivariate analyses adjusted for these variables and

CENTRAL ILLUSTRATION Transcatheter VIV Implantation

Mea

n Gr

adie

nt (m

m H

g)

Total Aortic Regurgitation40

20

15

30

25

35

10Baseline 30 Days

17.1

35.0

Mean Gradient

17.0

1 Year

100

KCCQ

Ove

rall

Scor

e

KCCQ Overall Summary Score Six Minute Walk Test Distance (m)

20

30

10

50

40

60

70

80

90

0

500

Six Minute W

alk Test Distance (m)

100

150

50

250

200

300

350

400

450

0Baseline 30 Days

p<0.0001

p=0.1135

p=0.0002

76.16

248.01

p<0.0001

43.03

163.74

70.56

229.31

1 Year

100

40

20

60

80

0Baseline

Severe Trace NoneModerate Mild

30 Days 1 Year

A. Changes in hemodynamics

B. Changes in function and quality of life

Webb, J.G. et al. J Am Coll Cardiol. 2017;69(18):2253–62.

We evaluated 30-day and 1-year outcomes of high-risk patients undergoing VIV transcatheter aortic valve replacement in failed bioprosthetic surgical aortic valves. At

both time points, significant improvements were seen in (A) hemodynamic measurements of mean gradient and aortic regurgitation, as well as (B) quality of life and

function as seen in KCCQ scores and 6-min walk test distances. KCCQ ¼ Kansas City Cardiomyopathy Questionnaire; VIV ¼ valve-in-valve.

J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7 Webb et al.M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2 THV Implantation in Degenerated Bioprostheses

2259

baseline STS risk score revealed no significantassociations with 1-year mortality.

One-year echocardiographic follow-up is shown inTable 4. At 1 year, the mean gradient was 17.6 mm Hg(95% CI: 16.2 to 19.1 mm Hg), EOA was 1.16 cm2 (95%CI: 1.11 to 1.21 cm2), and indexed EOA was 0.60 cm2/m2

(95% CI: 0.57 to 0.63 cm2/m2). When 30-day and 1-yearechocardiographic data were compared, no significantdifferences in mean EOA (1.13 cm2 vs. 1.16 cm2,

respectively; p ¼ 0.30) or mean gradient (17.7 mm Hgvs. 17.6 mm Hg, respectively; p ¼ 0.90) were seen(Figure 3). Patients with stenotic bioprosthetic failurehad higher 1-year mean gradient (18.9 mm Hg vs. 16.0mm Hg; p < 0.0001) and lower indexed EOA (0.57 vs.0.65 cm2/m2; p < 0.0001) than those with regurgitantor mixed failure and had greater proportional changesin both mean gradient and EOA at 1 year. At 1 year,aortic regurgitation was none/trace in 93.2%, mild in

FIGURE 4 NYHA Functional Class

100%

90%

80%

40%

30%

20%

10%

60%

50%

70%

0%1 Year

33.1

9.31.5

30 Days

35.3

9.51.2

Baseline

27.7

62.5

9.90

IV III II I

56.154

At baseline, >90% of patients were in NYHA functional class III or IV. By 30 days and

extending to 1 year, more than one-half were in class I and another one-third were in

class II. NYHA ¼ New York Heart Association.

Webb et al. J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7

THV Implantation in Degenerated Bioprostheses M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2

2260

4.7%, and moderate or severe in 1.9%. No cases ofmore than mild aortic regurgitation were valvular(within the implanted THV) (Central Illustration).

Mean LVEF increased from 50.6% (95% CI: 49.0%to 52.1%) at baseline to 54.2% (95% CI: 52.3% to56.1%) at 1 year (p < 0.0001). Overall, mean LV massindex decreased from 135.7 g/m2 (95% CI: 131.9 to139.5 g/m2) at baseline to 125.4 g/m2 (95% CI: 121.6 to129.8 g/m2) at 30 days and 117.6 g/m2 (95% CI: 113.3 to121.8 g/m2) at 1 year (p < 0.0001 for trend). Frombaseline to 1 year, both the rate of moderate or severemitral (34.9% vs. 12.7%, respectively; p < 0.0001) andtricuspid regurgitation (31.8% vs. 21.2%, respectively;p < 0.0001) decreased.

SYMPTOMS AND FUNCTIONAL AND QOL OUTCOMES.

Patient symptoms improved from baseline to 30 daysand 1 year (Figure 4). Themean overall summary KCCQscore was 43.0 (least squares: 40.7 to 45.3) at baseline,increasing to 70.6 (68.2 to 72.9) at 30 days and 76.2 (73.5to 78.8) at 1 year (p<0.0001); andmean 6-minwalk testdistance increased from 163.7m (least squares: 145.8 to181.7) at baseline to 229.3m (211.2 to 247.5m) at 30 daysand 248.0 m (226.9 to 269.1 m) at 1 year (p < 0.0001)(Central Illustration). No differences in KCCQ scoreswere seen when patients were stratified according tobioprosthesis size or residual gradient.

DISCUSSION

Results of the PARTNER 2 VIV registry showed thatTHV treatment of high-risk patients who had

degeneration of surgical aortic bioprosthetic valveswas associated with the following: low rates of mor-tality, stroke, rehospitalization, and new permanentpacemakers at 30 days and 1 year; significant re-ductions in transvalvular gradient and LV mass; sig-nificant increases in EOA and LVEF; very low rates ofsubstantial aortic regurgitation; significant reductionsin mitral and tricuspid valve regurgitation; and sig-nificant improvements in functional status and QOL(Central Illustration). These results are impressiveconsidering the observed-to-expected mortality ratiowas 0.3 at 30 days (observed mortality of 2.7% vs. STSpredicted risk of surgical mortality of 9.1%).

In patients at high risk of complications fromreoperation, the 1-year all-cause mortality of 12.4%was lower than the 16.8% rate reported in the retro-spective nonadjudicated VIVID (Valve-in-ValveInternational Data) registry (7) and compared veryfavorably with earlier native valve PARTNER experi-ences using first-generation devices in similarlyhigh-risk patients (PARTNER B mortality: 30.7%;PARTNER A mortality: 24.2%) (11,12). A learning curvefor TAVR has been documented in the treatment ofnative valve stenosis but not in the context of VIVTAVR (13,14). A striking reduction in mortality fromthe initial registry to the continued access registry at30 days (8.3% vs. 0.7%, respectively; p < 0.0001) andat 1 year (19.8% vs. 9.8%, respectively; p ¼ 0.006)suggests a learning curve did exist. Important dif-ferences between native valve TAVR and VIV TAVRincluded much lower rates of paravalvular regurgi-tation (more than mild in only 1.9%), new pacemakers(1.9%), and annular rupture (0%). Presumably therigid annular ring of the surgical bioprosthesisprotected against these complications.

This is the first large VIV registry with core labo-ratory echocardiographic follow-up. A significantreduction in mean gradient and increase in valve areawas durable to 1 year. Patients with severe ASbenefited from a 0.33-cm2 increase in EOA and21.7-mm Hg decrease in mean gradient, with residualaortic regurgitation less than mild in 98.1% of theentire cohort. At 1 year, however, an average residualmean gradient of 17.6 mm Hg remained, and criteriafor severe PPM were met in 58.2.1% of VIV cases.Although residual gradient of $20 mm Hg was pre-dictive of increased 1-year mortality, severe PPM wasnot. The impact of PPM on clinical outcomesfollowing native valve SAVR and TAVR has beenuncertain, with variable reports of its prognosticimportance and some suggestions that any increasein mortality may only emerge very late after SAVR(15–23). To put this in perspective, in patientsundergoing SAVR for native AS part of the PARTNER

PERSPECTIVES

COMPETENCY IN PATIENT CARE AND PROCEDURAL

SKILLS: Transcatheter aortic valve replacement (TAVR) may be

an alternative to cardiac reoperation in patients with degener-

ated aortic valve bioprostheses. In a large prospective registry of

patients undergoing valve-in-valve TAVR, procedural and 1-year

clinical, hemodynamic, functional, and QOL outcomes were

generally favorable, but patient-prosthesis mismatch was

relatively frequent.

TRANSLATIONAL OUTLOOK: Additional studies with longer

follow-up are needed to determine the durability of transcath-

eter valves and clinical importance of patient-prosthesis

mismatch in patients undergoing valve-in-valve TAVR.

J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7 Webb et al.M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2 THV Implantation in Degenerated Bioprostheses

2261

2A intermediate-risk trial, 33% had severe PPM (24).Randomized studies have shown TAVR to havehemodynamics slightly superior to those in SAVR,with lower gradients and a lower prevalence ofsevere PPM (12,25–27). The higher gradients andhigher rates of PPM after VIV procedures likelyrepresent a combination of pre-existing PPM dueto an undersized surgical valve and THV under-expansion due to the constraints of pre-existing sur-gical valves (7,26).

This is the first large VIV registry to systematicallyreport functional and QOL outcomes after VIV TAVR.Even patients with PPM, small (21-mm) surgicalvalves, or elevated post-TAVR gradients showedmarked improvements in symptoms, functional ca-pacity, and QOL at 30-day and 1-year follow-up. In ahighly comorbid population such as this, QOL mea-sures take on great importance, perhaps even greaterthan mortality. PPM might have a less importantimpact on symptoms in elderly patients, which de-scribes most VIV TAVR candidates (20).

The availability of VIV TAVR will likely influencesurgical practice in the management of aortic valvedisease, possibly lessening concerns about the dura-bility of surgical bioprostheses. It is essential to realize,however, that specific characteristics of the chosensurgical bioprosthesis, (including its internal diam-eter, design characteristics, and proximity to the cor-onary ostia) will have an impact on whether a VIVprocedure will or will not be a realistic future option.Unfortunately, current real-world data show thatmorethan one-third of all patients undergoing SAVR receivea bioprosthesis smaller than 23 mm (1). Over the longerterm, constrained THV expansion and malcoaptationwithin a bioprosthesis might lead to accelerated leafletdegeneration and reduced durability.

STUDY LIMITATIONS. This investigation had a num-ber of limitations, including the lack of a randomizedcomparator arm and follow-up limited to 1 year. Thestudy was conducted in carefully selected sites with

extensive oversight. The available THV sizes (23 and26 mm) did not allow inclusion of patients withthe largest or the smallest surgical bioprostheses.Clinical and hemodynamic outcomes may vary withdifferent THV devices.

CONCLUSIONS

The use of VIV TAVR for the treatment of high-riskpatients with degenerated aortic bioprostheses isassociated with relatively low rates of mortality andmajor complications, improved hemodynamics andexcellent improvement in functional and QOLoutcomes at 1 year. Longer follow-up is required todetermine the clinical importance of residualstenosis, particularly in smaller surgical valves,and the durability of transcatheter valves in thiscontext.

ADDRESS FOR CORRESPONDENCE: Dr. John G.Webb, St. Paul’s Hospital, 1081 Burrard Street,Vancouver, British Columbia V6Z 1Y6, Canada.E-mail: john.webb@vch.ca.

RE F E RENCE S

1. Thourani VH, Suri RM, Gunter RL, et al.Contemporary real-world outcomes of surgicalaortic valve replacement in 141,905 low-risk, in-termediate-risk, and high-risk patients. AnnThorac Surg 2015;99:55–61.

2. Isaacs AJ, Shuhaiber J, Salemi A, Isom OW,Sedrakyan A. National trends in utilization andin-hospital outcomes of mechanical versusbioprosthetic aortic valve replacements. J ThoracCardiovasc Surg 2015;149:1262–9.

3. Brennan JM, Edwards FH, Zhao Y, et al.Long-term safety and effectiveness of me-chanical versus biologic aortic valve prostheses

in older patients: results from the Society ofThoracic Surgeons Adult Cardiac SurgeryNational Database. Circulation 2013;127:1647–55.

4. Balsam LB, Grossi EA, Greenhouse DG, et al.Reoperative valve surgery in the elderly:predictors of risk and long-term survival. AnnThorac Surg 2010;90:1195–200; discussion 1201.

5. Kaneko T, Loberman D, Gosev I, et al. Reoper-ative aortic valve replacement in theoctogenarians-minimally invasive technique in theera of transcatheter valve replacement. J ThoracCardiovasc Surg 2014;147:155–62.

6. Kaneko T, Vassileva CM, Englum B, et al.Contemporary outcomes of repeat aortic valvereplacement: a benchmark for transcathetervalve-in-valve procedures. Ann Thorac Surg 2015;100:1298–304.

7. Dvir D, Webb JG, Bleiziffer S, et al. Trans-catheter aortic valve implantation in failed bio-prosthetic surgical valves. JAMA 2014;312:162–70.

8. Blanke P, Soon J, Dvir D, et al. Computed to-mography assessment for transcatheter aorticvalve in valve implantation: the Vancouverapproach to predict anatomical risk for coronary

Webb et al. J A C C V O L . 6 9 , N O . 1 8 , 2 0 1 7

THV Implantation in Degenerated Bioprostheses M A Y 9 , 2 0 1 7 : 2 2 5 3 – 6 2

2262

obstruction and other considerations. J CardiovascComput Tomogr 2016;10:491–9.

9. Bapat V. Valve-in-valve apps: why and howthey were developed and how to use them.EuroIntervention 2014;10:44–51.

10. Kappetein AP, Head SJ, Genereux P, et al.Updated standardized endpoint definitions fortranscatheter aortic valve implantation: the ValveAcademic Research Consortium-2 consensusdocument (VARC-2). Eur J Cardiothorac Surg2012;42:45–60.

11. Leon MB, Smith CR, Mack M, et al. Trans-catheter aortic-valve implantation for aorticstenosis in patients who cannot undergo surgery.N Engl J Med 2010;363:1597–607.

12. Smith CR, Leon MB, Mack MJ, et al. Trans-catheter versus surgical aortic-valve replacementin high-risk patients. N Engl J Med 2011;364:2187–98.

13. Alli O,Rihal CS, Suri RM,et al. Learning curves fortransfemoral transcatheter aortic valve replacementin the PARTNER-I trial: Technical performance.Catheter Cardiovasc Interv 2016;87:154–62.

14. Minha S, Waksman R, Satler LP, et al. Learningcurves for transfemoral transcatheter aortic valvereplacement in the PARTNER-I trial: Success andsafety. Catheter Cardiovasc Interv 2016;87:165–75.

15. Chen J, Lin Y, Kang B, Wang Z. Indexedeffective orifice area is a significant predictor ofhigher mid- and long-term mortality ratesfollowing aortic valve replacement in patientswith prosthesis-patient mismatch. Eur JCardiothorac Surg 2014;45:234–40.

16. Head SJ, Mokhles MM, Osnabrugge RL, et al.The impact of prosthesis-patient mismatch onlong-term survival after aortic valve replacement: asystematic review and meta-analysis of 34 obser-vational studies comprising 27 186 patients with133,141 patient-years. EurHeart J 2012;33:1518–29.

17. Dayan V, Soca G, Stanham R, Lorenzo A,Ferreiro A. Is patient-prosthesis mismatch apredictor of survival or a surrogate marker ofco-morbidities in cardiac surgery? Int J Cardiol2015;190:389–92.

18. Hong S, Yi G, Youn YN, Lee S, Yoo KJ,Chang BC. Effect of the prosthesis-patientmismatch on long-term clinical outcomes afterisolated aortic valve replacement for aortic ste-nosis: a prospective observational study. J ThoracCardiovasc Surg 2013;146:1098–104.

19. Kandler K, Moller CH, Hassager C, Olsen PS,Lilleor N, Steinbruchel DA. Patient-prosthesismismatch and reduction in left ventricular massafter aortic valve replacement. Ann Thorac Surg2013;96:66–71.

20. Price J, Toeg H, Lam BK, Lapierre H,Mesana TG, Ruel M. The impact of prosthesis-patient mismatch after aortic valve replacementvaries according to age at operation. Heart 2014;100:1099–106.

21. Tully PJ, Aty W, Rice GD, Bennetts JS,Knight JL, Baker RA. Aortic valve prosthesis-patient mismatch and long-term outcomes:19-year single-center experience. Ann Thorac Surg2013;96:844–50.

22. Koene BM, Soliman Hamad MA, Bouma W,et al. Impact of prosthesis-patient mismatch on

early and late mortality after aortic valvereplacement. J Cardiothorac Surg 2013;8:96.

23. Takagi H, Umemoto T. Prosthesis-patientmismatch after transcatheter aortic valve implan-tation. Ann Thorac Surg 2016;101:872–80.

24. Leon MB, Smith CR, Mack MJ, et al. Trans-catheter or surgical aortic-valve replacement inintermediate-risk patients. N Engl J Med 2016;374:1609–20.

25. Adams DH, Popma JJ, Reardon MJ, et al.Transcatheter aortic-valve replacement with aself-expanding prosthesis. N Engl J Med 2014;370:1790–8.

26. Pibarot P, Weissman NJ, Stewart WJ, et al.Incidence and sequelae of prosthesis-patientmismatch in transcatheter versus surgicalvalve replacement in high-risk patients withsevere aortic stenosis: a PARTNER trial cohort—an analysis. J Am Coll Cardiol 2014;64:1323–34.

27. Ewe SH, Muratori M, Delgado V, et al.Hemodynamic and clinical impact of prosthesis-patient mismatch after transcatheter aorticvalve implantation. J Am Coll Cardiol 2011;58:1910–8.

KEY WORDS 6-min walk test, aorticstenosis, mortality, registry, regurgitation

APPENDIX For supplemental tables and afigure, please see the online version of thisarticle.