TAVR with next-­generation self-­expanding devices: a multicenter propensity-­matched

comparison of Evolut PRO versus Acurate neobioprostheses – NEOPRO Registry

PRINCIPAL INVESTIGATORSAzeem Latib and Matteo Pagnesi

San Raffaele Scientific Institute, Milan, Italy

On behalf of the NEOPRO Registry investigators

• Several next-­generation THVs have been developed to minimizeTAVR complications and improve outcomes

• The self-­expanding Acurate neo (NEO) and Evolut PRO (PRO)devices have been associated with excellent clinical and echocardiographic outcomes1-­3

• No published data comparing Acurate neo and Evolut PRO THVs

1. Möllmann H et al. EuroIntervention 2017;;13:e1040–6;; 2. Möllmann et al. EuroIntervention2018;;13:e1764–70;; 3. Forrest et al. JACC Cv Intv 2018;;11:160–8.

Background

• To compare short-­term clinical and echocardiographic outcomes after transfemoral TAVR with the next-­generation self-­expanding NEO and PRO devices

Objective

Study devices

Device characteristics Acurate neo Evolut PRO

THV designSelf-­expanding

Supra-­annular porcine pericardial leafletsPericardial sealing-­skirt

Self-­expandingSupra-­annular porcine pericardial leaflets

Pericardial sealing-­skirt

Specific THV features -­ RepositionableRetrievable up to 80%

Easy of use High Intermediate

Sizes 23 mm (S) – 25 mm (M) – 27 mm (L) 23 mm – 26 mm – 29 mm

TF delivery system 20-­F sheath 16-­F sheathless

Radial strength Low Intermediate

Outer skirtheight 13 mm4 mm

• Multicenter, observational, retrospective NEOPRO Registry

• All consecutive patients treated with transfemoral TAVR for symptomatic, severe aortic stenosis of the native aortic valve with either NEO or PRO implantation were included

• A total of 1551 patients (1263 NEO, 288 PRO) treated at 24 centers between Jan 2012 and Mar 2018

Study design

Canada (13 patients)

Brazil (40 patients)

Ireland (5 patients)

United Kingdom (70 patients)

Germany (898 patients)

Italy (306 patients)

Switzerland (102 patients)

Austria (54 patients)

The Netherlands (70 patients)

Spain (11 patients)

Participating centers

Kerckhoff Heart and Lung CenterWon-­Keun Kim, Christian W. HammBad Nauheim, Germany

University Heart Center HamburgLenard Conradi, Oliver Bhadra, Ulrich SchäferHamburg, Germany

University of CataniaMarco Barbanti, Giuliano Costa, Corrado TamburinoCatania, Italy

Humanitas Research HospitalGiulio G. Stefanini, Francesco Cannata, Bernhard ReimersRozzano-­Milan, Italy

San Raffaele Scientific InstituteMatteo Pagnesi, Antonio Colombo, Azeem LatibMilan, Italy

University Hospital DüsseldorfTobias Zeus, Verena VeulemansDüsseldorf, Germany

Bern University HospitalThomas Pilgrim, Masahiko Asami, Stephan WindeckerBern, Switzerland

Albertinen Heart CenterJoachim Schofer, Amnon EitanHamburg, Germany

Medical University of GrazDavid Zweiker, Albrecht SchmidtGraz, Austria

IRCCS Policinico San DonatoLuca Testa, Giovanni Bianchi, Francesco BedogniMilan, Italy

University Hospital of ZürichMaurizio Taramasso, Matteo Saccocci, Francesco MaisanoZürich, Switzerland

Royal Sussex County HospitalDavid Hildick-­Smith, Osama AlsanjariBrighton, United Kingdom

Instituto Dante PazzaneseAlexandre Abizaid, Dimytri SiqueiraSão Paulo, Brazil

Contilia Heart and Vascular CentreAlexander Wolf, Christoph J. Jensen, Christoph K. NaberEssen, Germany

Erasmus Medical CenterNicolas M. Van Mieghem, Francesca ZivielloRotterdam, The Netherlands

University Hospital BonnAlexander Sedaghat, Jan-­Malte SinningBonn, Germany

Ulm University Medical CenterJochen Wöhrle, Julia SeegerUlm, Germany

New Cross HospitalSaib KhogaliWolverhampton, United Kingdom

St. Antonius HospitalJan A.S. Van der Heyden, Jorn BrouwerNieuwegein, The Netherlands

St. Paul’s HospitalJohn G. Webb, Abdullah Alenezi, David A. WoodVancouver, Canada

Hospital of LeónRodrigo Estévez-­LoureiroLeón, Spain

Galway University HospitalsDarren MylotteGalway, Ireland

King's College HospitalPhilip MacCarthy, Vasileios TzalamourasLondon, United Kingdom

University Hospital Clinic, IDIBAPSSalvatore Brugaletta, Ander RegueiroBarcelona, Spain

Participating centers

Primary endpoint• Device success according to VARC-­2 criteria

Secondary outcomes• Procedural outcomes (VARC-­2 criteria)• Pre-­discharge echocardiographic outcomes• 30-­day clinical outcomes (VARC-­2 criteria)

Study endpoints

• Primary and secondary endpoints compared between NEO and PRO groups in the entire population and after PS matching

• In the overall cohort, binary logistic regression was also performed to adjust the treatment effect for the PS quintiles

Statistical analysis

Propensity score (PS) calculation• PS was estimated by means of a non-­parsimonious multivariable logistic regression including the following variables:

-­ age -­ prior cardiac surgery-­ sex -­ prior BAV-­ BMI -­ previous PM or ICD-­ diabetes mellitus -­ NYHA class III or IV-­ COPD -­ LVEF-­ eGFR -­ STS-­M score-­ prior MI -­ moderate-­to-­heavy AV calcification-­ PVD -­ moderate-­to-­severe LVOT calcification

• C-­statistic 0.78;; Hosmer-­Lemeshow goodness-­of-­fit test p-­value 0.35

Statistical analysis

PS matching

• 1-­to-­1 nearest neighbor matching without replacement to identify PS matched pairs

• Pseudo-­R2was 0.08 (p < 0.0001) before matching and very low (0.007;; p=0.995) after matching

Statistical analysis

PS matched cohortN=502 (251 NEO, 251 PRO)

Overall cohortN=1551 (1263 NEO, 288 PRO)

Baseline characteristicsEntire

population

Acurate neo(n=1263)

Evolut PRO(n=288)

p-­value

Age (years) 81.8 ± 5.8 81.7 ± 5.9 0.911

Male sex 35.2% 36.1% 0.785

eGFR (mL/min/1.73/m2) 58.4 ± 21.7 55.7 ± 20.2 0.052

Prior cardiac surgery 12.2% 8.4% 0.080

PVD 12.4% 24.3% <0.001

Prior stroke 10.3% 7.7% 0.224

NYHA class III-­IV 77.8% 66.9% <0.001

STS-­M 5.02 ± 3.23 5.35 ± 3.87 0.137

AV calcification <0.001

None 0.4% 0.4%

Mild 28.9% 15.4%

Moderate 43.8% 44.0%

Heavy 26.9% 40.2%

Baseline characteristicsPS matchedpopulation

Acurate neo(n=251)

Evolut PRO(n=251)

p-­value

Age (years) 81.4 ± 6.5 81.6 ± 6.1 0.742

Male sex 34.3% 34.3% 1.000

eGFR (mL/min/1.73/m2) 57.2 ± 21.4 56.2 ± 20.1 0.609

Prior cardiac surgery 8.4% 8.0% 1.000

PVD 24.7% 24.7% 1.000

Prior stroke 10.0% 7.6% 0.430

NYHA class III-­IV 69.7% 67.3% 0.631

STS-­M 5.08 ± 3.05 5.25 ± 3.72 0.577

AV calcification 0.313

None 0.4% 0.4%

Mild 14.7% 15.5%

Moderate 51.8% 44.2%

Heavy 33.1% 39.8%

Procedural characteristicsEntire

population

Acurate neo(n=1263)

Evolut PRO(n=288)

p-­value

Conscious sedation 86.7% 92.3% 0.007

Valve size -­

S or 23 mm 27.6% 2.4%

M or 26 mm 41.2% 37.5%

L or 29 mm 31.2% 60.1%

Predilatation 83.3% 37.1% <0.001

Postdilatation 41.7% 23.7% <0.001

Valve repositioning -­ 8.0% -­

Second THV implanted 1.1% 2.1% 0.240

Valve embolization 1.0% 1.0% 1.000

Annular rupture 0.3% 0.0% 1.000

Pericardial tamponade 1.6% 0.0% 0.036

Conversion to surgery 1.0% 0.0% 0.144

Mean AV gradient (mmHg) 8.5 ± 4.0 7.2 ± 3.5 <0.001

Procedural characteristicsAcurate neo(n=1263)

Evolut PRO(n=288)

p-­value

Conscious sedation 96.4% 92.8% 0.112

Valve size -­

S or 23 mm 23.9% 2.0%

M or 26 mm 41.8% 38.3%

L or 29 mm 34.3% 59.8%

Predilatation 86.5% 37.9% <0.001

Postdilatation 41.4% 25.0% <0.001

Valve repositioning -­ 8.6% -­

Second THV implanted 0.8% 1.2% 1.000

Valve embolization 0.4% 0.8% 1.000

Annular rupture 0.0% 0.0% -­

Pericardial tamponade 2.0% 0.0% 0.061

Conversion to surgery 0.8% 0.0% 0.499

Mean AV gradient (mmHg) 8.3 ± 4.0 7.3 ± 3.6 0.003

PS matchedpopulation

VARC-­2 device success

PS matched populationEntire population

Moderate-­to-­severe PAR

PS matched populationEntire population

Overall PAR

PS matched populationEntire population

30-­day clinical outcomesEntire

population

Acurate neo(n=1263)

Evolut PRO(n=288)

p-­value

All-­cause mortality 3.0% 1.8% 0.319

Cardiovascular mortality 2.3% 1.1% 0.249

Any stroke 2.0% 2.5% 0.645

Any bleeding 14.9% 8.5% 0.004

Life-­threatening 2.1% 1.1% 0.335

Major 4.7% 2.8% 0.195

Minor 8.1% 4.6% 0.044

Any vascular complication 17.1% 11.6% 0.025

Major 6.0% 3.5% 0.114

Minor 11.1% 8.1% 0.163

Coronary obstruction 0.2% 0.4% 0.562

New PPI 8.8% 13.2% 0.045

VARC-­2 safety composite endpoint 16.4% 10.9% 0.025

30-­day clinical outcomesAcurate neo(n=251)

Evolut PRO(n=251)

p-­value

All-­cause mortality 3.2% 1.2% 0.221

Cardiovascular mortality 2.4% 0.4% 0.122

Any stroke 2.4% 2.8% 1.000

Any bleeding 13.0% 8.4% 0.112

Life-­threatening 2.4% 0.8% 0.175

Major 4.1% 2.4% 0.323

Minor 6.5% 5.2% 0.572

Any vascular complication 14.2% 11.7% 0.424

Major 4.9% 3.2% 0.372

Minor 9.3% 8.4% 0.754

Coronary obstruction 0.0% 0.0% -­

New PPI 11.0% 12.8% 0.565

VARC-­2 safety composite endpoint 10.6% 10.4% 1.000

PS matchedpopulation

Adjustment for PS quintiles

Outcome OR (95% CI) p-­value

VARC-­2 device success 1.15 (0.68-­1.96) 0.598

Moderate-­to-­severe PAR 0.85 (0.46-­1.60) 0.625

30-­day clinical outcomes

All-­cause mortality 0.42 (0.12-­1.43) 0.166

Cardiovascularmortality 0.19 (0.02-­1.42) 0.104

Any stroke 1.15 (0.47-­2.82) 0.758

Any bleeding 0.63 (0.38-­1.03) 0.067

Any vascular complication 0.75 (0.48-­1.16) 0.196

New PPI 1.49 (0.93-­2.39) 0.100

VARC-­2 safety composite endpoint 0.76 (0.48-­1.20) 0.234

Binary logistic regression to adjust the treatment effect (PRO vs. NEO) for the PS quintiles (entire population)

• Retrospective design• No core-­laboratory analysis• No independent adjudication of clinical events• Potential impact of unknown/unmeasured confounding factors on outcomes

• Potential selection bias (higher cost of Evolut PRO THV)• Different sample size between groups;; small sample size in the PS matched population

Study limitations

• TF TAVR with the next-­generation Acurate neo and Evolut PRO THVs was associated with high device success, acceptable rates of moderate-­to-­severe PAR, and good 30-­day clinical outcomes

• Subtle differences in procedural characteristics and clinical outcomes highlight the different design features of the 2 valves

• After adjustment for potential confounders, short-­term outcomes were similar between both devices

Conclusions