Long Term Clinical Outcomes Following DrugEluting and Bare Metal Stenting in Massachusetts

1
Long Term Clinical Outcomes Following
Drug-Eluting and Bare Metal Stenting in
Massachusetts

• Laura Mauri, MD, MSc Treacy Silverstein, B.Sc.
• Ann Lovett, R.N., M.A. Frederic S. Resnic, M.D.,
  M.Sc.
• and Sharon-Lise T. Normand, Ph.D.
• Brigham and Womens Hospital, Harvard Medical
  School,
• Harvard School of Public Health
• Boston, Massachusetts
• Study funding from Massachusetts Department of
  Public Health
• November 4, 2007
• American Heart Association Scientific Sessions

2
Drug-Eluting and Bare Metal Stenting in
MassachusettsBackground

• The data reported for available DES in the US
  indicate that these devices are important
  advancements by reducing the need for
  revascularization
• Long term clinical outcomes remain uncertain
  following DES use compared to BMS use in the
  general population

3
Drug-Eluting and Bare Metal Stenting in
Massachusetts Background

• Randomized trials of DES have been limited in
  power to detect rare events and in
  generalizability to a broad range of practice
  patterns and patient populations
• Observational studies have been limited by
  possible residual selection bias and incomplete
  follow-up
• Studies comparing DES and BMS use have been
  performed in populations with restricted DES use

4
Drug-Eluting and Bare Metal Stenting in
Massachusetts Objectives

• To evaluate whether the use of DES is associated
  with increased rates of death or MI compared with
  BMS in general use representative of current US
  practice
• To confirm whether the use of DES is associated
  with reduction in revascularization compared with
  BMS in general use representative of current US
  practice

5
Drug-Eluting and Bare Metal Stenting in
Massachusetts Methods Data source Mass-DAC

• All PCI in Massachusetts (gt25000 stent
  procedures per year)
• Massachusetts Dept. of Public Health mandatory
  reporting
• Clinical and procedural factors collected
  prospectively using ACC NCDR instrument
• Mortality from hospital record, Mass. vital
  statistics and Social Security website
• Myocardial infarction and revascularization from
  Mass-DAC PCI and CABG data merged with hospital
  discharge data
• Index procedures performed April 2003 September
  2004
• Non-Massachusetts residents excluded

6
Drug-Eluting and Bare Metal Stenting in
Massachusetts Methods Propensity Score Matched
Analysis

• Patients assigned to DES or BMS groups based on
  treatment at index procedure
• Propensity score matching
• Logistic regression to predict DES treatment by
  patient, procedural, hospital variables
• Caliper matching of DES to BMS patients
• Primary outcomes Matched risk differences for
  mortality, myocardial infarction and
  revascularization rates at 2 years
• Paired t-test, 2-sided alpha 0.05

7
Drug-Eluting and Bare Metal Stenting in
Massachusetts
N21,019 PCI Patients in Massachusetts April 1,
2003- September 30, 2004 Complete 2 year follow-up
1,538 non-residents excluded
564 adminstrative files not linkable
N18,917 PCI Patients
1,191 patients with both stent types excluded
N11,516 DES Only Patients
N6,210 BMS Only Patients
7
8
35 BMS
65 DES
72 sirolimus, 28 paclitaxel eluting stents
9
Drug-Eluting and Bare Metal Stenting in
MassachusettsPatient Characteristics before Match
10
Drug-Eluting and Bare Metal Stenting in
MassachusettsPatient Characteristics before Match
11
Drug-Eluting and Bare Metal Stenting in
MassachusettsProcedure Indications before Match
12
Drug-Eluting and Bare Metal Stenting in
MassachusettsProcedural Characteristics before
Match
13
Drug-Eluting and Bare Metal Stenting in
Massachusetts, Unadjusted 2-Year Outcomes
14
n5441 DES n5441 BMS Propensity matched
pairs 63 variables
15
Drug-Eluting and Bare Metal Stenting in
MassachusettsPatient Characteristics after Match
SD Percent Standardized Difference Values
lt10 reflect well-matched characteristics
16
Drug-Eluting and Bare Metal Stenting in
MassachusettsPatient Characteristics after Match
SD Percent Standardized Difference Values
lt10 reflect well-matched characteristics
17
Drug-Eluting and Bare Metal Stenting in
MassachusettsProcedure Characteristics after
Match
SD Percent Standardized Difference Values
lt10 reflect well-matched characteristics
18
Drug-Eluting and Bare Metal Stenting in
MassachusettsProcedural Characteristics after
Match
SD Percent Standardized Difference Values
lt10 reflect well-matched characteristics
19
Drug-Eluting and Bare Metal Stenting in
Massachusetts, Primary ResultsPropensity
Matched 2-Year Outcomes
D -3.8 -5.4,-2.3
D -1.0 -2.2,0.2
D -2.4 -3.6,-1.3
20
Drug-Eluting and Bare Metal Stenting in
Massachusetts 2-Year Outcome in Matched Patients
Mortality
Time after Initial Procedure (days)
21
Drug-Eluting and Bare Metal Stenting in
Massachusetts 2-Year Outcome in Matched Patients
MI
12.3
11.2
Time after Initial Procedure (days)
22
Drug-Eluting and Bare Metal Stenting in
Massachusetts 2-Year Outcome in Matched Patients
Revascularization
Time after Initial Procedure (days)
23
Drug-Eluting and Bare Metal Stenting in
Massachusetts Matched Risk Differences at 2
years
Mortality

• -2.4 -3.6,-1.3
• plt0.0001

DES 9.4 BMS 11.9
Myocardial Infarction

• -1.0 -2.2,0.2
• p0.11

DES 10.8 BMS 11.3
Revascularization

• -3.8 -5.4,-2.3
• plt0.0001

DES 20.1 BMS 23.9
24
Drug-Eluting and Bare Metal Stenting in
Massachusetts Sensitivity analyses

• Effect of differential rates of use of DES vs BMS
  over time
• 2 day mortality difference

25
n3752 DES n3752 BMS Propensity matched
pairs adjusted for time on market
26
Drug-Eluting and Bare Metal Stenting in
Massachusetts Sensitivity Analysis 2-Year
Outcomes adjusted for time on market
D -1.9 -3.9,-0.1
D -1.4 -2.8,0.0
D -3.2 -4.6,-1.8
27
Drug-Eluting and Bare Metal Stenting in
Massachusetts Sensitivity analyses

• Adjustment for time on market is consistent with
  results of the primary analysis
• larger observed difference favoring DES for
  mortality
• 2 day mortality difference after propensity match
  is small
• 2 day mortality 0.4 vs 0.8 (DES vs BMS)
• D -0.4 at 2d vs D -2.4 at 2 y

28
Drug-Eluting and Bare Metal Stenting in
Massachusetts Comparison with PCI Practice
Patterns and Other Studies
Distinguishing features high rate of
PCI/population, high rate of DES/procedure,
study sample chosen to start with introduction
of DES and to end when all patients had complete
2 year follow up.
1. Lagerqvist B et al. N Engl J Med
20073561009-19. 2. Jensen, LO et al. JACC
200750(5)463-70. 3. Tu J et al. NEJM
20073571393-42.
29
Drug-Eluting and Bare Metal Stenting in
MassachusettsConclusions

• From the Massachusetts state dataset of 21,024
  PCI unique patient procedures in Massachusetts
  from April 2003- September 2004 with complete 2y
  follow-up, propensity matched analysis of 10,882
  patients demonstrated
• No increase in rates of death, or myocardial
  infarction associated with DES as compared to BMS
  use at 2 years.
• A significantly lower mortality rate was observed
  for DES at 2 years, and will be monitored as
  follow up proceeds to 3-5 years
• A lower rate of revascularization in patients
  treated with DES compared with BMS.