Clinical Assessment of Patients with Acute Coronary Syndrome Managed - - PowerPoint PPT Presentation

Clinical Assessment of Patients with Acute Coronary Syndrome Managed with Percutaneous Coronary Intervention and Treated with Prasugrel or Clopidogrel using Academic Center Databases

• The PROMETHEUS Study-

Study Organization

Study Leadership

Roxana Mehran (PI) – Mount Sinai

Steering Committee

Timothy Henry – Cedars Sinai Sunil Rao – Duke University William Weintraub – Christiana

Study Sponsor

Mark B. Effron Stuart Keller Douglas Faries Cliff Molife Brian A. Baker Elizabeth Marrett Eli Lilly and Daiichi Sankyo

Data Coordination/Analysis

Usman Baber Samantha Sartori Melissa Aquino Sonny Sayseng Icahn School of Medicine at Mount Sinai

Statistical Consultation

• Prof. Stuart Pocock

London School of Hygiene

Study Rationale

• Randomized studies have shown that compared to clopidogrel,

prasugrel significantly reduces thrombotic events, albeit with an excess risk of bleeding, in ACS patients undergoing PCI with stent implantation

• Data examining the safety and efficacy profile of prasugrel in

non-randomized patients presenting across the entire clinical spectrum of ACS remains limited

• The overall frequency and determinants of prasugrel use

remains poorly characterized

Study Design

• Cohort study evaluating ACS (UA, NSTEMI or STEMI)

patients undergoing PCI with stent implantation in academic centers in the United States between 1/1/2010 – 6/30/2013

• Each center maintains an institutional database with baseline

clinical, procedural and outcome data on patients undergoing PCI

• Center feasibility to extract and provide relevant data

(baseline and outcomes) was assessed and confirmed with a site selection questionnaire

• The study sponsor had no access to patient level data

Definitions

• Major Adverse Cardiovascular Events (MACE)

¡ Composite occurrence of all-cause death, myocardial

infarction, unplanned revascularization or stroke

• Bleeding

¡ Clinically overt hemorrhage requiring hospitalization

• Primary Efficacy Endpoint

¡ MACE at 90 days from index PCI

• Primary Safety Endpoint

¡ Bleeding requiring hospitalization

Study Schema

Index ¡PCI ¡ ¡

Data Extraction and Analysis

• Data Elements

pre-specified by Steering Committee

Extraction Sheet

• Local Sites

Data Extraction

• Data

Coordinating Center

Validation/ Quality Check Aggregation Analysis

• Sample Size Assumptions

¡ MACE rate ~ 8.0% at 90 days ¡ Relative reduction with prasugrel ~ 20% ¡ Minimum of 4,300 prasugrel patients required to

achieve 80% power

• Propensity score adjustment

¡ Stratification (Primary) ¡ Inverse Probability weighting ¡ Matching ¡ Covariate adjustment

Statistical Methods

Propensity Score Histogram

Predicted Probability to Receive Prasugrel

PROMETHEUS Sites

Cleveland Clinic, Samir Kapadia Duke University, Sunil Rao Intermountain Heart Institute, Brent Muhlestein Minneapolis Heart Institute, Craig Strauss UPMC, Catalin Toma Christiana Care, Sandra Weiss Mount Sinai Medical, Annapoorna Kini Aurora Healthcare Anthony DeFranco

19,914 patients over 8 US Sites from 01 Jan 2010 to 30 June 2013

Prasugrel and Clopidogrel Use in Overall Cohort

20% 80% Prasugrel Clopidogrel n = 4058 n = 15856

CAD Presentation in Overall Cohort

17% 27% 56%

STEMI (n=3285) NSTEMI (n=5412) Unstable angina (n=11,216) 52.3% 57.3% 28.6% 26.8% 19.1% 15.8%

0% 20% 40% 60% 80% 100% Prasugrel Clopidogrel

STEMI (n=3,285) NSTEMI (n=5,412) Unstable Angina (n=11,216)

CAD Presentation by Therapy

Baseline Characteristics*

Prasugrel (n = 4,058) Clopidogrel (n = 15,856) Age, years 58.7 ± 10.3 65.8 ± 12.3 Female sex 989 (24.4%) 5315 (33.5%) African-American 253 (6.2%) 1,872 (11.8%) Weight <60 kg 124 (3.1%) 1145 (7.2%) Diabetes 1382 (34.1%) 6198 (39.1%) Diabetes on insulin 394 (9.7%) 2140 (13.5%) Hypertension 2915 (71.8%) 13466 (84.9%) Dyslipidemia 3220 (79.3%) 13469 (84.9%) Smoking 1175 (29.0%) 3832 (24.2%) Prior MI 833 (20.5%) 5130 (32.4%) Prior PCI 788 (19.4%) 4250 (26.8%) Prior CABG 359 (8.8%) 3074 (19.4%) Prior cerebrovascular disease 188 (4.6%) 2197 (13.9%) CKD 619 (15.3%) 4995 (31.6%) Anemia 339 (8.4%) 2553 (16.1%)

*p <0.0001 for all

Procedural Characteristics*

Prasugrel (n = 4,058) Clopidogrel (n = 15,856) Multivessel disease 1672 (41.2%) 6723 (42.4%) PCI vessel Left Main 84 (2.1%) 583 (3.9%) LAD 1972 (48.6%) 6926 (43.7%) Circumflex 1100 (27.1%) 4795 (30.2%) RCA 1430 (35.2%) 5368 (33.9%) At least one B2/C type lesion 2848 (71.7%) 10763 (75.3%) At least one lesion with moderate/severe calcification 468 (12.0%) 2694 (19.3%) Total stent length 31.4 ± 20.2 30.50 ± 20.9 Minimum stent diameter 3.01 ± 0.49 2.96 ± 0.50 At least one 1st gen DES 297 (7.3%) 2496 (15.7%) At least one 2nd gen DES 3297 (81.2%) 10525 (66.4%) At least one BMS 569 (14.0%) 3927 (24.8%) Procedural anticoagulation Bivalirudin 2743 (67.6%) 11730 (74.0%) GP2b3a inhibitor 1178 (29.0%) 3391 (21.4%) *p <0.05 for all except multivessel disease (p=0.17) and RCA (0=0.096)

6.2 [5.8-6.6] 9.6 [9.1-10.1] 14.3 [13.7-14.9] 20.7 [20.0-21.3] 40%

MACE %

4058 3605 3376 3169 2636 Pras 15856 13612 12413 11322 9438 Clop

Number at risk 30 90 180 365

Days After PCI

*MACE is defined as all cause death, myocardial infarction, stroke, or unplanned revascularization.

Overall MACE

0% 10% 20% 30% Clopidogrel Prasugrel p-value<0.001 p-value<0.001 p-value<0.001 p-value<0.001 3.5 [2.9-4.1] 5.7 [5.0-6.5] 8.3 [7.4-9.2] 12.1 [11.0-13.2]

HR [95% CI] p-value 30 Days

Unadjusted

<0.001

Adjusted (PS)

0.09 90 Days

Unadjusted

<0.001

Adjusted (PS)

0.17 180 Days

Unadjusted

<0.001

Adjusted (PS)

0.12 365 Days

Unadjusted

<0.001

Adjusted (PS)

0.01

0.56 0.84 0.58 0.89 0.56 0.90 0.56 0.86

0.1 1 10

Hazard Ratios for Overall MACE

Prasugrel Better Clopidogrel Better

PS – Propensity stratified

HR [95% CI] p-value 30 Days

Unadjusted

<0.001

Adjusted (PS)

0.15 90 Days

Unadjusted

<0.001

Adjusted (PS)

0.18 180 Days

Unadjusted

<0.001

Adjusted (PS)

0.15 365 Days

Unadjusted

<0.001

Adjusted (PS)

0.30

0.53 0.81 0.51 0.84 0.49 0.85 0.54 0.90

0.1 1 10

Hazard Ratios for Myocardial Infarction

Prasugrel Better Clopidogrel Better

PS – Propensity stratified

0.15 0.47 0.21 0.62 0.25 0.69 0.26 0.69

0.01 0.1 1 10 100 HR [95% CI] p-value 30 Days

Unadjusted

<0.001

Adjusted (PS)

0.04 90 Days

Unadjusted

<0.001

Adjusted (PS)

0.04 180 Days

Unadjusted

<0.001

Adjusted (PS)

0.04 365 Days

Unadjusted

<0.001

Adjusted (PS)

0.01

Hazard Ratios for All-Cause Death

Prasugrel Better Clopidogrel Better

PS – Propensity stratified

0.61 0.98 0.65 1.03 0.61 0.97 0.64 0.97

0.1 1 10

HR [95% CI] p-value

30 Days

Unadjusted

<0.001

Adjusted (PS)

0.89 90 Days

Unadjusted

<0.001

Adjusted (PS)

0.79 180 Days

Unadjusted

<0.001

Adjusted (PS)

0.82 365 Days

Unadjusted

<0.001

Adjusted (PS)

0.83

Prasugrel Better Clopidogrel Better

Hazard Ratios for Bleeding

PS – Propensity stratified

Limitations

• Observational design precludes causal inferences
• Outpatient medication adherence not available
• Potential for residual/unmeasured confounding on point

estimates despite extensive statistical adjustment

• Lack of standardized definition for bleeding endpoint
• Potential for therapeutic cross-over may result in biasing

point estimates to the null

Conclusions

• The clinical profile of ACS patients treated with clopidogrel as compared to

prasugrel differs substantially across US academic medical centers

¡ The overall burden of risk factors for both ischemic and hemorrhagic events are

much lower in patients treated with prasugrel

• Although in this study prasugrel use was associated with lower rates of

ischemic adverse events, the magnitude of benefit attenuated and was no longer statistically significant after adjusting for baseline differences

• Analogously, lack of adjusted bleeding difference with prasugrel may reflect

selection of patients at very low risk for hemorrhagic complications

• Recalibrating ‘real-world’ use of prasugrel to better approximate a patient’s

ischemic risk may yield a more appreciable therapeutic benefit, a hypothesis that warrants prospective evaluation

Acknowledgements

Minneapolis Heart Institute Ross Garberich Aurora Medical Center Sandra Hubatch Cleveland Clinic Kanhaiya Poddar Intermountain Heart Institute Stacey Knight Duke University Medical Center Amanda McBroom Brooks Christiana Care Health Services Angela Herman University of Pittsburgh Medical Center Floyd Thoma Icahn School of Medicine at Mount Sinai Jaya Chandrasekhar Ioannis Mastoris Swathi Roy Amreen Rahman