Long-Term Safety of DES in Off-Label Use: Results of the MATRIX Registry

1
Long-Term Safety of DES in Off-Label
UseResults of the MATRIX Registry
George D. Dangas, MD, PhD, FACC, FSCAI On Behalf
of the Matrix Investigators
2
Disclosures
Personal May be construed as possible COI
Cordis Endovascular, JJ Completed Term Consultancy (lt10K)

MATRIX Funding Support
Cordis Cardiology, JJ Research Grant to the Cardiovascular Research Foundation
3
MATRIX Goals and Design

• Prospective single arm study initiated in 2004
  under an investigator-initiated IDE (1st
  submitted in October 2003)
• Designed to evaluate the outcomes of SES in
  consecutive real world population undergoing
  PCI with SES
• Both on- and off-label SES use
• Clinical follow-up at 1 month, 6 months, 1 year
  and 2 years thus far

MATRIX Registry Dangas et al, SCAI-ACCi2 2008
4
Study Organization

• Principal Investigator George D. Dangas, MD
• Clinical sites Lenox Hill Hospital, Columbia
  University Medical Center
• Data management Data Center of Cardiovascular
  Research Foundation
• Independent CEC (Chair J. Coromilas, MD)
• 100 monitoring of all data fields of the first
  1,000 pts 10 thereafter
• Independent QCA lab for the first 800 lesions
  treated

5
Medication Regimen

• Pre-procedure
• Aspirin 325 mg
• Clopidogrel loading dose of 300-600 mg within 24
  hours followed by 75 mg once daily or Ticlopidine
  loading dose of 500 mg within 24 hours, followed
  by 250 mg twice a day.
• During procedure
• Bivalirudin or Heparin GP IIb/IIIa inhibitors
• Post-procedure and after discharge
• Aspirin 325 mg for 1 month, thereafter Ec-ASA 81
  mg indefinitely
• Clopidogrel 75 mg once daily for at least three
  months but recommend for 1 year to all patients
  physician discretion thereafter

6
Follow-Up
N1,510 patients N1,510 patients N1,510 patients N1,510 patients
Eligible for F/U
30 days 87.9 (1327/1510)
6 months 87.7 (1324/1510)
1 year 88.6 (1338/1510)
2 years 70.3 (877/1248)
7
Baseline Clinical Characteristics
n 1,510 patients
Age, meanSD (years) 64.811.1
Male gender 74.6
Prior myocardial infarction 33.3
History of PCI 44.4
History of CABG 21.0
Diabetes mellitus 33.7
Unstable angina 27.7
ST-Elevation MI within 48 hrs 3.3
Chronic renal insufficiency 10.1
History of Stroke or TIA 8.0
History of peripheral arterial disease 7.3
8
Baseline Angiographic Characteristics
N 2,876 lesions
Target vessel
Unprotected LM 1.6
LAD 37.1
LCX 29.5
RCA 27.4
SVG 4.5
Arterial conduit 0.6
Target lesion location
Ostial 7.6
Proximal 30.6
Chronic total occlusion 3.5
Bifurcation lesion 19.5
Restenotic lesion 7.5
Sirius-2.25 mm
MATRIX Registry
9
Procedural Characteristics
N 1,510 patients
No. of stents per procedure 2.01.2
No. of stents per lesion 1.10.5
Unfractionated heparin 16.0
Bivalirudin used 85.0
IIb/IIIa inhibitors administered 8.2
Procedure success 95.6
Device success 98.6
10
On-Label Use of Cypher Stent

• The CYPHER Sirolimus-eluting Coronary Stent is
  indicated in patients with symptomatic ischemic
  disease due to discrete de novo lesions of length
  lt 30 mm in native coronary arteries with a
  reference vessel diameter of gt 2.5 to lt 3.5 mm
  (http//www.fda.gov/cdrh/PDF2/p020026c.pdf).
• On-label definition in MATRIX De novo lesion 1
  lesion 1 vessel Lesion length lt 30mm RVD
  2.5-3.5mm Also excluding
• Diffuse disease
• Multivessel PCI PCI with 3 of more SES
• Use of rotablator, atherectomy or laser
• Use of thrombectomy or intracoronary thrombus
• Acute ST elevation MI within 72 hours before the
  procedure
• ACS with positive CKMB prePCI
• Ostial lesions
• Bifurcation lesions
• Chronic occlusions, baseline TIMI flow 0 or 1
• Vein grafts, LIMA/RIMA, radial or GEA grafts
• Angioplasty restenosis or in-stent restenosis
• Severe calcification Severe tortuosity

14 Of Patients in MATRIX w/o any of above
MATRIX Registry
11
Antiplatelet Adherence
Patients ()
N1501 N1503 N1327 N1324
N1338 N877
12
Patients off clopidogrel at 30 days, 6 moths or 1
year, and back at 6 months, 1 year or 2 years
Reason 2 years
Patients off clopidogrel at 30 days and back at 6 months, 1 year or 2 years 46 (23/50)
Patients off clopidogrel at 30 days and back at 6 months 34
Patients off clopidogrel at 30 days and back at 1 year 36
Patients off clopidogrel at 30 days and back at 2 years 24
Patients off clopidogrel at 6 months and back at 1 year or 2 years 49 (65/133)
Patients off clopidogrel at 6 months and back at 1 year 44
Patients off clopidogrel at 6 months and back at 2 years 44
Patients off clopidogrel at 1 year and back at 2 years 10 (23/233)
13
Impact of Clopidogrel Adherence 1-Year Use and
Outcomes gt 365 Days
On-Plavix (N1101) Off-Plavix (N236) p
Death 1.4 4.8 0.005
Cardiac death 0.2 1.1 0.054
Non-cardiac death 1.0 2.7 0.08
Unknown death 0.3 1.0 0.16
Myocardial infarction 0.9 1.1 0.80
Q wave 0 0 N/A
Non-Q wave 0.9 1.1 0.80
TLR 5.5 0.0 0.001
TVR 6.1 0.5 0.002
Death/ MI 2.2 5.9 0.008
Death/ MI/ CD-TVR 7.2 6.3 0.71
Stent thrombosis 0.3 0 0.46
MATRIX Registry
14
Time-dependent (time-updated) Cox regression
model for outcomes up to 2 years. Clopidogrel
adherence is treated as a time- dependent
variable.
2-Year Events Hazard Ratio 95 CI p
Off Clopidogrel (vs. On Clopidogrel)
Death (48 events) 3.10 1.53 6.32 0.0018
Cardiac death 3.82 0.92 15.75 0.0642
Non cardiac death 1.94 0.69 5.46 0.2099
MI (55 events) 0.93 0.20 4.25 0.9266
Death/MI (98 events) 2.19 1.14 4.20 0.0182
Thrombosis (12 events) 0.00 0.00 gt999 0.9928
MATRIX Registry
15
Reasons of Clopidogrel Discontinuation
Reason 6 months 1 year 2 years
Clopidogrel discontinuation 133/1324 pts (10.0) 233/1338 pts (17.4) 291/877 pts (33.2)
Doctors choice 4.5 9.9 11.7
Bleeding 3.8 5.2 1.7
Surgery 2.3 2.1 2.7
Rash or allergy 9.0 1.7 0.3
Cost 0.8 0.4 0.7
Post 1 year N/A 41.2 30.9
Other/unknown 79.7 39.5 52.2
Doctors choice/Bleeding/Surgery /Rash or allergy/Cost 20.3 19.3 16.8
Post 1 year/other/unknown 79.7 80.7 83.2
16
Impact of Reasons of Clopidogrel Discontinuation
at 1 year Outcomes gt 365 Days
Doctors choice /Bleeding/Rash/Allergy/Cost (N45) Post 1 year /Other/Unknown (N188) p
Death 9.0 (3) 3.3 (5) 0.10
Cardiac death 3.0 (1) 0.7 (1) 0.19
Non-cardiac death 6.1 (2) 1.4 (2) 0.07
Unknown death 0.0 (0) 1.3 (2) 0.53
Myocardial infarction 0.0 (0) 1.3 (2) 0.55
Q wave 0.0 (0) 0.0 (0) N/A
Non-Q wave 0.0 (0) 1.3 (2) 0.55
TLR 0.0 (0) 0.0 (0) N/A
TVR 0.0 (0) 0.6 (1) 0.63
Death/ MI 9.0 (3) 4.6 (7) 0.23
Death/ MI/ CD-TVR 9.0 (3) 5.2 (8) 0.33
Stent thrombosis 0.0 (0) 0.0 (0) N/A
17
Prediction of 2-Year Adverse Outcomes
Multivariate Predictors Using Cox Model
2-Year Events 2-Year Events Hazard Ratio 95 CI 95 CI p
Death (48 events) Death (48 events)
Age Age 1.08 1.05 1.12 1.05 1.12 lt.0001
Clopidogrel Clopidogrel 2.03 1.11 3.73 1.11 3.73 0.0218
Diabetes Mellitus Diabetes Mellitus 2.03 1.11 3.73 1.11 3.73 0.0029
Dialysis Dialysis 5.67 1.72 18.76 1.72 18.76 0.0045
Death or MI (95 events) Death or MI (95 events) Death or MI (95 events) Death or MI (95 events) Death or MI (95 events)
Visual lesion length Visual lesion length 1.02 1.00 1.04 1.00 1.04 0.0313
Age Age 1.05 1.03 1.07 1.03 1.07 lt.0001
Diabetes Mellitus Diabetes Mellitus 1.58 1.05 2.38 1.05 2.38 0.0280
Renal insufficiency Renal insufficiency 2.24 1.38 3.64 1.38 3.64 0.0011
Definite or probable stent thrombosis (12 events) Definite or probable stent thrombosis (12 events) Definite or probable stent thrombosis (12 events) Definite or probable stent thrombosis (12 events) Definite or probable stent thrombosis (12 events)
Multivessel stenting 3.34 3.34 3.34 1.08 10.36 0.0368
Renal insufficiency 4.55 4.55 4.55 1.37 15.11 0.0134
Candidate predictors included on-label use, age,
male, DM, renal insufficiency, dialysis, AMI,
ACS, visual length, visual RVD, vessel stented,
2 vessel stented, clopidogrel.
18
MATRIX - Conclusions
In 1,510 patients with complex CAD treated with
SES, we found

• A low but measurable rate of clopidogrel
  discontinuation overtime. This was associated
  with higher all-cause mortality by Cox regression
  time-dependent analysis.
• Patients off-clopidogrel at 12 months had higher
  mortality and fewer repeat coronary procedures at
  follow-up.
• Based on the reasons for clopidogrel
  discontinuation patients who stopped clopidogrel
  for acute events or side-effects appeared to have
  a trend towards higher mortality (particularly
  non-cardiac) at follow-up.
• A significant proportion of patients who
  discontinued clopidogrel was able to be treated
  again with this agent at a later time point
• Continued surveillance on long-term adherence to
  dual antiplatelet therapy after DES is warranted.