COMPARE-ABSORB 1 year results

Pieter C. Smits, MD&

Robert-Jan van Geuns, MDOn behalf of the COMPARE-ABSORB

investigators

StephanACHENBACH

EmanueleBARBATO

BernardCHEVALIER

JavierESCANED

TommasoGORI

VictorKOCKA

GuiseppeTARANTINI

Robert-JanVAN GEUNS

PeterSMITS

Nick WEST

JanTIJSSEN

YoshiONUMA

Marie-ClaudeMORICE

DariuszDUDEK

Steering Committee

EmanueleBARBATOAalst (66)

BernardCHEVALIERMassy (99)

TommasoGORI

Mainz (72)

Robert-JanVAN GEUNS

Rotterdam (55)

PeterSMITS

Rotterdam (201)

Nick WEST

Cambridge (89)

GiovanniESPOSITONaples (62)

MohanedEGRED

Newcastle (50)

RalphTÖLG

Bad Segeberg (67)

MohamedABDEL-WAHAB

Bad Segeberg (67)

AdrianWŁODARCZAK

Lubin (178)

TOP 10ENROLLERS

B. Vaquerizo Monttilla Hospital del Mar,Barcelona, Spain

47

G. Tarantini Azienda Ospedaliera,Padova, Italie

47

V. Kocka University Hospital Kralovs, Prague, Czech Republic

43

P. O’Kane Royal Bournemouth Hospital,Bournemouth, UK

39

P. Buszman American Heart of Poland,Chrzanow, Poland

36

P. Kala University Hospital,Brno, Czech Republic

34

S. Achenbach Universitatsklinikum,Erlangen, Germany

33

M. Maly Central Military Hospital,Prague, Czech Republic

30

K. Milewski American Heart of Poland,Tychy, Poland

30

S. Ijsselmuiden Albert Schweitzer Hospital,Dordrecht, The Netherlands

29

U. Landmesser Charité Campus Benjamin Berlin, Germany

29

C. Naber ElisabethkrankenhausEssen, Germany

28

P. Tonino Catharina Ziekenhuis,Eindhoven, The Netherlands

26

D. Dudek University Hospital,Krakow, Poland

25

H. Nef UniversitatsklinikumGiessen, Germany

25

P. Motreff CHU Clermont-FerrandClermont Ferrand

25

J. Sainsous Clinique Rhône Durance, Avignon, France

24

S. Brugaletta Hospital Clinic,Barcelona, Spain

21

C. Liebetrau Kerckhoff Klinik,Bad Nauheim, Germany

19

G. Saad CHR de la Citadelle,Liege, Belgium

19

A. Menozzi Universitaria di ParmaParma, Italie

17

J. Fajadet Clinique Pasteur,Toulouse, France

15

C. Cernetti Ospedale San Giacoma,Castelfranco Veneto, Italie

13

O. Valssecchi Ospedale Papa Giovanni XXIII,Bergamo, Italie

12

M. Meeuwissen Amphia Ziekenhuis,Breda, The Netherlands

11

J. Escaned Hospital Clinico San Carlos,Madris, Spain

11

T. Rudolph Universitatsklinikum,Koln, Germany

11

C. Indolfi Universita Degli Studi Magna Graecia, Catanzaro, Italie

6

B. Loi Azienda Ospeldaliera BrotzuCagliari, Italie

6

R. Koning Clinique Saint Hilaire,Rouen, France

4

W. Desmet UZ Leuven,Leuven, Belgium

4

J. Mehilli Klinikum der Universitat München,München, Germany

3

P. Lurz Universitatsklinikum,Leipzig, Germany

3

M. Caruso Arnas Civico,Palermo, Italie

3

COMPARE-ABSORB trial

Investigator initiated,prospective, single blind, multicenter

randomized controlled trial comparing Xience versus Absorb

in a high-risk patient and/or complex lesion population

Cum

ulat

ive in

ciden

ce (%

)

0

5

10

15

20

25

Days since initial procedure

0 180 360 540 720 900 1080 1260 1440 1620 1800

BESEES

Study background

5 year results COMPARE II

13.7 %

11.8 %TLF

XienceBP-BES

Vlachojannis et al. JACC Cardiovasc Int 2017;10:1215-21

Objectives• To show non-inferiority between

Absorb and Xience on TLF at 1 year

• Secondary objective is to show superiority of Absorb on TLF between 1 and 5 years in a landmark analysis

• Tertiary objective is to show superiority of Absorb on TLF from start to 5 year

COMPARE-ABSORBInclusion criteria

• Patients with at least one of the following:

i) High-risk characteristics for restenosisKnown diabetes and/or multivessel disease of which more than one de-novo target lesion to be treated with the study scaffold/stent

ii) Complex de-novo target lesion- Lesion length >28 mm- Small vessels: RVD between 2.25-2.75 mm- Lesion with pre-existing total occlusion- Bifurcation with single device strategy

COMPARE-ABSORBExclusion criteria

Patient level• Patients at age >75 yr• Renal insufficiency

(GFR <30 ml/min)• Known LVEF < 30%• life expectancy < 7 years• Known non-adherence to

DAPT• Patients on OAC or NOAC• Cardiogenic Shock, Killip >2

Lesion / vessel level• Target lesion reference

vessel diameter < 2.25 and > 4.0 mm

• STEMI with target lesion RVD of >3.5mm

• Target lesion with in-stent/scaffold thrombosis or re-stenosis

• Graft lesions as target lesions

• Severe tortuosity of target vessel

• Bifurcation target lesion with intended 2 device strategy

Key features of COMPARE-ABSORBSpecific patient population and implantation technique

• To study a patient population which potentially might benefit the most by the vascular restoration therapy concept on the long term

• Selection of specific patients and complex lesions not investigated in previous RCT’s like: STEMI, acute non-STEMI, bifurcations and long lesions and CTO’s

• PSP implantation technique from the start

Key features of COMPARE-ABSORBPSP protocol for ABSORB implantation from the start

• Pre-dilatation mandatory with 1:1 balloon-artery ratio

• Treatment of target vessels < 2.75 mm QCA, IVUS or OCT highly recommended

• High pressure (>16 Atm) post-dilatation is mandatory

• Post-dilatation with NC balloons up to 0.5 mm larger than the scaffold is recommended

REmergent/ Elective PCI

2100 pts

ABSORB

XienceRandomization after successful w iring first target lesion

1Y 5Y1st analysis:

Non-inferiority in TLF at 1Y

3rd analysis: Cumulative

superiority in TLF at 5Y (or 7Y)

2nd analysis: Superiority in TLF between 1 and 5

years45 sites across Europe

Original protocolApril 2015

• TLF 1yr in Xience: 8.5% • Non-inferiority margin: 4.5%• Alpha = 0.05• Power=90% • Sample size: 808 x 2 = 1616 pts

• TLF 1-5yr in Xience: 10.5% • TLF 1-5yr in ABSORB 6.3%• Power=90% • Sample size: 1004 x 2 = 2008 pts• Sample size 5% attr. = 2100 pts

Trial design (original)

Start regulatory

submission

July2015

Sep.2015

Oct.2016

Dec.2015

May2016

Study Start-up period

Enrollment period Follow-up period

First Patient enrolled

28 Sep 2015Maassstad

Hospital

First Initiated site

28 Sep 2015

100 Patients enrolled

600 Patients enrolled

All sites enrolling

1200Patientsenrolled

Nov.2016

ABSORB II3 yr results

March2017

FDA warningletter

ABSORB III2 yr results

Study Flow

Protocol revisions

• Lower limit of small vessels was increased from 2.25 to 2.5 mm

• Secondary long term endpoint was changed to 3 to 7 year landmark analysis

• Prolonged 36 months DAPT regimen in the scaffold arm was advised

REmergent/ Elective PCI

2100 pts

ABSORB

XienceRandomization after successful w iring first target lesion

1Y 7Y1st analysis:

Non-inferiority in TLF at 1Y

3rd analysis: Cumulative

superiority in TLF at 7Y (or 10 yr)

2nd analysis: Superiority in TLF between 3

and 7 years45 sites across Europe

• Changed the secondary analysis• Extended the follow -up to 7 years• Excluded target vessel ref. diam. < 2.5 mm• Excluded high risk bleeding patients• Recommendation on extension DAPT

3YAmended protocolApril 2017

Trial design (revised)

Start regulatory

submission

July2015

Sep.2015

Oct.2016

Dec.2015

May2016

Study Start-up period

Enrollment period Follow-up period

First Patient enrolled

28 Sep 2015Maassstad

Hospital

First Initiated site

28 Sep 2015

100 Patients enrolled

600 Patients enrolled

All sites enrolling

1200Patientsenrolled

Nov.2016

ABSORB II3 yr results

March2017

31 AugustDSMB letter

Trial put on hold

Aug. & Sept. 2017

8 SeptemberCommercial

stopon ABSORB

1670Patientsenrolled

Sept.2018

Study flowAIDA

NEJM 2017

ABSORB III2 yr results

FDA warningletter

Randomized 1:1N=1670 (ITT)

XienceN=822

1-Year Follow-up

Study Flow and Follow-up

N = 11 no contactN = 11 withdrew consent

96.9% Complete

XienceN=800

97.3% Complete

AbsorbN=848

AbsorbN=822

N = 13 no contactN = 13 withdrew consent

Base-line characteristicsRisk factors ABSORB

848 patientsXIENCE

822 patients P value

Age [yr] ± SD 61.9 ± 9.4 62.2 ± 9.0 0.61

Male 79.5% (674) 76.3% (627) 0.13

Diabetes mellitus 34.6% (293) 36.1% (296) 0.57

Current smoker 28.8% (241) 26.9% (217) 0.41

Previous smoker 51.9% (289) 50.1% (280) 0.55

Hypercholesterolemia 66.3% (546) 65.8% (531) 0.88

Hypertension 71.6% (601) 69.2% (567) 0.31

Family history of CAD 36.2% (278) 31.7% (241) 0.07

Previous PCI 27.0% (229) 20.2% (238) 0.38

Previous CABG 1.9% (16) 2.6% (21) 0.41

Previous MI 18.2% (154) 20.2% (166) 0.29

Previous stroke 3.4% (29) 4.8% (39) 0.18

Renal insufficiency 3.9% (33) 6.0% (49) 0.054

LV ejection fraction [%] ± SD 56.4 ± 10.5 56.3 ±10.2 0.83

Base-line characteristicsIndication and treatment

ABSORB848 patients

1242 target lesions

XIENCE822 patients

1213 target lesionsP value

Acute coronary syndrome (ACS) 52.1% (442) 48.7% (400) 0.17

STEMI 13.0% (110) 12.5% (103) 0.88

Non-STEMI treatment < 72 hours 13.3% (113) 12.4% (102) 0.57

Multi-vessel treatment 35.7% (303) 37.7% (301) 0.56

Mean target lesions treated ± SD 1.5 ± 0.7 1.5 ± 0.7 0.67

Mean Syntax score ± SD 12.2 ± 7.1 12.2 ± 7.3 0.88

Bifurcation lesions 20.5% (254) 22.2 (269) 0.30

Pre-existing total occlusions 14.6% (181) 13.1% (159) 0.32

Long lesions (>28mm) 25.2% (313) 31.5% (370) <0.001

Small vessel lesions (>2.25 ≤ 2.75 mm) 22.5% (279) 30.5% (370) <0.001

Procedural characteristicsVessel and lesion treatment

ABSORB1242 target lesions

962 procedures

XIENCE1213 target lesions

904 proceduresP value

Pre-dilatation 96.5% (1198) 78.6% (1213) <0.001

Largest balloon (mm ± SD) 3.0 ± 1.0 3.0 ± 0.7 0.96

Non-compliant balloon used 67.9% (814) 52.9% (504) <0.001

Max. pressure used (Atm.) 15.3 ± 3.5 14.8 ± 3.4 0.002

Cutting / scoring balloon 5.8% (72) 2.3% (28) <0.001

Mean study devices used 1.3 ± 0.7 1.3 ± 0.6 0.07

Post-dilatation 90.7% (1497) 58.3% (906) <0.001

Non-compliant balloon used 93.0% (1392) 85.5% (775) <0.001

Largest balloon diameter (mm ± SD) 3.3 ± 0.4 3.3 ± 0.5 0.97

Max. pressure largest balloon (Atm) 17.6 ± 3.7 17.5 ± 3.7 0.76

Max. pressure ≧ 16 Atm 79.7% (1193) 79.5% (720) 0.92

IVUS performed post 14.3% (138) 14.3% (129) 1.0

OCT performed post 9.4% (90) 2.7% (24) <0.001

Procedural characteristics

Vessel and lesion treatmentABSORB

962 procedures1242 target lesions

XIENCE904 procedures

1213 target lesionsP value

Procedural success 98.4% (947) 98.9% (894) 0.43

Device success 92.4% (1148) 96.8% (1174) <0.001

Successful delivery of device 95.0% (1180) 99.2% (1203) <0.001

Residual stenosis < 30% 96.9% (1203) 97.4% (1182) 0.40

DAPT usage

97.2 97.6 96.7

80.7

97.8 97.4 94.5

72.4

0

20

40

60

80

100

120

Discharge 1 month 6 months 12 months

Absorb Xience

% P=NS P=0.04P=NS P<0.001

Primary endpoint1 year TLF non-inferiority analysis

• Assumed difference between Xience and Absorb : 0 %• Non inferiority margin : 4.5 %• One sided 2.5% significance level• TLF rate Xience 4.2%• TLF rate Absorb 5.1%

0-1-2-3- 4

Δ Prim EP %

1

Absorb not inferior

Δ Prim. EP: Absorb - Xience = 0.9 % (95% CI: -1.2 – 3.0 %)

Absorb is non-inferior compared to Xience

P < 0.001

- 5 2 3 4 5

TLF @ 1 yearCardiac death, target vessel myocardial infarction, clinically-indicated target lesion revascularization

5.1%

4.2%

HR 1.24 (0.79-1.94)Plogrank = 0.35

Components of TLFCardiac death

Target vessel myocardial infarction

Clinically indicated target vessel revascularization

2.4%2.7%

4.0%

2.1%

0.6%0.1%

MI definition: • SCAI (peri-procedural)• TUD (spontaneous)

HR 4.87 (0.57-41.70)Plogrank = 0.11

HR 0.89 (0.48-1.62)Plogrank = 0.69

HR 1.96 (1.10-3.51)Plogrank = 0.02

Stent/Scaffold Thrombosis @ 1 yearDefinite Stent/Scaffold Thrombosis (ARC definition)

1.9%

0.6%

HR 3.12 (1.14-8.51)Plogrank=0.02

Stent/Scaffold Thrombosis @ 1 yearDefinite and Probable Stent/Scaffold Thrombosis

(ARC definition)

2.0%

0.6%

HR 3.32 (1.22-8.99)Plogrank=0.01

Clinical events

0.7 0.6

4.0 4.0

7.1

3.6

2.4

6.3

5.14.4

0.60.1

2.52.1

7.4

3.7

2.7

4.84.2

2.6

0.0

1.0

2.0

3.0

4.0

5.0

6.0

7.0

8.0

9.0

10.0

All causedeath

Cardiacdeath

All MI TV MI AnyRevasc

CI-TVR CI-TLR TVF TLF CD or MI

Absorb Xience

p = 0.04

p = 0.82

p = 0.02

% e

vent p = 0.35

p = 0.17

p = 0.69

p = 0.92p = 0.07

p = 0.11p = 0.79

Device thrombosis

1.9

0.1 0.2

2.0

0.6

0.0 0.1

0.6

0.0

1.0

2.0

3.0

4.0

5.0

Def. device thrombosis Prob. device thrombosis Poss. device thrombosis Def. or prob. devicethrombosis

Absorb Xience

p = 0.01

p = 0.32 p = 0.58

p = 0.02

Subgroup analyses on TLF

Subgroup analyses on TLF

Conclusions

• In a high risk population for restenosis the primary endpoint of non-inferiority for target lesion failure at 1 year was met between Absorb and Xience

• The target lesion failure rate in a high risk population was low, potentially due to the applied implantation protocol

Conclusions• Although non-inferiority was met, scaffold

thrombosis and myocardial infarction rate for Absorb was significantly higher compared to Xience, despite applied PSP protocol

• COMPARE-ABSORB will have a 7 year follow-up to investigate the potential benefit of Absorb in a high risk population