Global Trials: Challenges and Opportunities Case Study: The ExTRACT-TIMI 25 Trial

1
Global Trials Challenges and OpportunitiesCase
Study The ExTRACT-TIMI 25 Trial

• Elliott Antman, MD
• Brigham and Womens Hospital
• Harvard Medical School
• Boston, MA

2
Disclosure
The TIMI Study Group has received research /
grant support in the past 2 yrs through the
Brigham Womens Hospital with funding from (in
alphabetical order)
The National Institutes of Health Integrated
Therapeutics Corporation KAI Pharmaceuticals Merck
Co., Inc. Millennium Pharmaceuticals, Inc.
Novartis Pharmaceuticals Nuvelo, Inc.
Ortho-Clinical Diagnostics, Inc. Pfizer, Inc.
Roche Diagnostics Corporation Roche Diagnostics
GmbH Sanofi-Aventis Sanofi-Synthelabo
Recherche Schering-Plough Research Institute St
Jude Medical
Accumetrics, Inc. Amgen, Inc. AstraZeneca
Pharmaceuticals LP Baxter Bayer Healthcare
LLC Beckman Coulter, Inc. Biosite
Incorporated Bristol-Myers Squibb CardioKinetix CV
Therapeutics, Inc.Daiichi-Sankyo Eli Lilly and
Company FoldRx GlaxoSmithKline INO Therapeutics
LLC Inotek Pharmaceuticals Corporation
3
Hamm Lancet 3581533,2001
Ischemic Discomfort
Presentation
Acute Coronary Syndrome
Working Dx
Davies MJ Heart 83361, 2000
No ST Elevation
ST Elevation
ECG
NSTEMI
Biochem. Marker
Myocardial Infarction
Unstable Angina
NQMI
Qw MI
Final Dx
4
Reperfusion Strategies for STEMI
Pharmacologic
PCI
Limited Availability Treatment Delay More
Effective Bleeding Risk Lower
Widely Available Quickly Administered Less
Effective Bleeding Risk
5
Thrombosis of epicardial coronary artery..
Thrombin
Fibrin
Lytic Rx
Antithrombins
Flow
Antiplatelet Rx
.the cause of STEMI
6
Pharmacologic Reperfusion for STEMI Components
of Regimen
FibrinolyticSK? Fibrin- specific? Bolus
AntiplateletASA? GPIIb/IIIaThienopyridine
AnticoagulantUFH ? Alternative Agents
7
X
TF/VIIa
Xa inhibitorsLMWH
Xa
TFPI
UFH
V, Ca2
LMWH
Thrombin
UFH
Prothrombin
DTIs
Clopidogrel
ASA
Platelet
GP IIb/IIIaInhibitor
8
Potential Advantages of Anticoagulation with LMWH
vs UFH
LMWH
UFH
Inhibit thrombus generation
Greater
Less
Route
IV
SC
Monitor A/C
No
Yes
Inhibition by Platelets
No
Yes
9
Enoxaparin for STEMI
No Lytic RxTETAMI
Adjunct to LyticHART IIAMI SKBaird et
alASENOXENTIRE-TIMI 23 ASSENT 3ASSENT 3-PLUS
10
ASSENT 3 Bleeding Stratified by Age
lt 75 yrs(N 5328)
gt75 yrs(N 767, 13)
P lt0.0001
P 0.26
JACC 39 306A, 2002
11
Primary Hypothesis
Am Heart J 200514917-26.

• Compared to UFH, adjunctive antithrombin therapy
  with ENOX reduces the composite end point of
  all-cause mortality or non-fatal re-MI within 30
  days in patients with STEMI who are eligible to
  receive fibrinolytic therapy.

12
Protocol Design
Am Heart J 200514917-26.
STEMI lt 6 hLytic eligible
Lytic choice by MD(TNK, tPA, rPA, SK)
ASA
Double-blind, double-dummy
ENOX lt 75 y 30 mg IV bolus SC 1.0 mg / kg q 12
h (Hosp DC) 75 y No bolusSC 0.75 mg / kg q 12
h (Hosp DC) CrCl lt 30 1.0 mg / kg q 24 h
UFH60 U / kg bolus (4000 U) Inf 12 U / kg / h
(1000 U / h)Duration at least 48 hContd at
MD discretion
Day 301 Efficacy Endpoint Death or Nonfatal
MI1 Safety Endpoint TIMI Major Hemorrhage
13
Trial Features

• Central Randomization Toll free
  phone Stratified by center
• Double Blind
• Double Dummy
• Ratio 11 (Enox UFH)

14
Statistical Considerations

• Sample Size UFH 10.5 RRD 13 Enox 9.13
  ARD 1.37 2-sided ? 5 Power gt 90 2080
  events (approx 21,000 pts)
• Interim Stopping Rules 3 Interim looks
  (25,50,75 of events) If Mortality lower with
  UFH (Plt0.01) at first 2 looks--consider
  stopping If Mortality lower with Enox (Plt0.01)
  AND D/MI lower (Plt0.02) at 3rd
  look-consider stopping Final P value 0.043

15
Study Medication Kits

Drug A32 single-use ampules (1 ml)
Drug B4 multi-use vials (10 ml)
A
UFH (5000 U/ml) / Placebo
Enox (100mg/ml) / Placebo
IV bolus A
SC Injections
IV bolus B
IV Infusion
16
CBF ProceduresaPTT/ACT via Hemochron Jr.
17
Enrollment Oct 2002 - Oct 2005N 20,479 (ITT)
Argentina Finland Latvia Singapore
Australia France Lebanon Slovakia
Austria Germany Lithuania South Africa
Belarus Greece Malaysia Spain
Belgium Hong Kong Mexico Sweden
Brazil Hungary Netherlands Switzerland
Bulgaria India New Zealand Thailand
Canada Ireland Norway Turkey
Chile Israel Poland Ukraine
China Italy Portugal United Kingdom
Croatia Jordan Romania United States
Estonia Republic of Korea Russian Federation Uruguay
48 Countries 674 Sites
18
Top 10 Enrolling Countries-1
Country
Lead Inv
Subjects

• Russia
• Poland
• Spain
• Turkey
• 5. Israel

Ruda Sadowski/ Budaj Lopez-Sendon Guneri
Hod
4,887 1,792 1,281 953 870
19
Top 10 Enrolling Countries-2
Country
Lead Inv
Subjects
790 753 645 639 626
6. Ukraine 7. India 8. Netherlands 9. Great
Britain 10. Italy
Parkhomenko SomaRaju Molhoek Jacob Ardissino
20
Other Countries-through 48
Country
Lead Inv
Subjects
57 49 46 43 39
36. Belarus 37. United States 38.
Lithuania 39. Norway 40. Latvia
Polonetsky Antman Sanofi-Aventis Dickstein San
ofi-Aventis
21
Baseline Characteristics ITT N 20,479
N Engl J Med 20063541477-88.
59
Age (yrs)-median
82
CrCl (ml/min)-median
77
Male ()
16
UFH within 3 h ()
44
Hypertension ()
0.5
LMWH within 7 d ()
18
Hyperlipidemia ()
89
Killip Class I ()
47
Current smoker ()
TIMI Risk Score (STEMI)
15
Diabetes ()
64
lt 3 ()
13
Prior MI ()
36
gt 3 ()
44
Anterior MI ()
ALL P NS
22
Medications ITT N 20,479
N Engl J Med 20063541477-88.
Fibrinolytic
20
SK ()
Fibrin-specific ()
80
95
ASA ()
86
Beta Blocker ()
80
ACEI / ARB ()
70
Statin ()
ALL P NS
23
Main Results
NEJM 3541477, 2006
Bleeding by 30 days
Primary Endpoint Death or non-fatal re-MI by 30
days
UFH
12.0
ARD 0.7RR 1.53Plt0.0001
ARD 0.1RR 1.27P 0.14
ARD 0.4RR 1.84P 0.001
9.9
ENOX
Events

ARD 0.021 2.1 RR 0.83 (0.77 to 0.90)RRR
0.17 (0.23 to 0.10)NNT 48
FatalMajor Bleed
Major Bleed(Total)
ICH
Days
33 RRR in reMI by 48 h (P0.002) 19 RRR in
Death/MI by 72 h (Plt0.001)
24
Revised Pkg Insert for Enoxaparin
25
Death or Reinfarction Across the ACS Spectrum
Eur Heart J. 2007282077-86.
Odds ratio
(95 CI)
0.78 (0.63,0.98)
ASSENT 3
7.7
9.6
1.00 (0.49,2.06)
HART II
8.0
8.0
0.60 (0.35,1.01)
BAIRD
20.8
30.5
0.24 (0.09,0.64)
ENTIRE-TIMI 23
4.4
15.9
0.89 (0.65,1.22)
ASSENT 3 Plus
10.3
11.4
0.81 (0.74,0.88)
ExTRACT-TIMI 25
9.9
12.0
STEMI (p0.002)
0.78 (0.67,0.91)
9.6
11.7
0.76 (0.58,1.01)
ESSENCE
5.8
7.5
0.88 (0.70,1.11)
TIMI 11B
7.4
8.3
0.98 (0.51,1.86)
ACUTE II
7.9
8.1
0.53 (0.30,0.95)
INTERACT
5.0
9.0
0.94 (0.73,1.20)
A TO Z
7.4
7.8
SYNERGY
0.96 (0.85,1.07)
13.9
14.5
NSTEACS (p0.043)
0.90 (0.81,0.996)
10.0
11.0
0.84 (0.76,0.92)
TOTAL
p lt 0.001
.2
1
5
Odds ratio
Favors Enox
Favors UFH
26
Major Bleeding Across the ACS Spectrum
Eur Heart J. 2007282077-86.
Odds ratio
Enox ()
UFH ()
(95 CI)
1.27 (0.90,1.79)
ASSENT 3
3.8
3.0
1.18 (0.39,3.57)
HART II
3.6
3.0
0.84 (0.25,2.81)
BAIRD
3.4
4.0
0.76 (0.13,4.67)
ENTIRE-TIMI 23
1.9
2.4
1.70 (1.07,2.68)
ASSENT 3 Plus
6.2
3.8
1.54 (1.24,1.91)
ExTRACT-TIMI 25
2.1
1.4
STEMI (plt0.001)
2.6
1.8
1.45 (1.23,1.72)
0.93 (0.70,1.23)
ESSENCE
1.56 (1.13,2.14)
TIMI 11B
0.33 (0.03,3.68)
ACUTE II
0.58 (0.33,1.03)
INTERACT
3.78 (1.25,11.41)
A TO Z
1.21 (1.05,1.40)
SYNERGY
NSTEACS (p0.42)
6.3
5.4
1.13 (0.84,1.54)
1.25 (1.04,1.50)
TOTAL
1
p 0.019
.2
5
Odds ratio
Favors Enox
Favors UFH
27
(No Transcript)
28
STEMI Treatments Outcomes WorldwideResults
from the Enoxaparin and Thrombolysis Reperfusion
for Acute Myocardial Infarction - Thrombolysis In
Myocardial Infarction (ExTRACT-TIMI) 25 Registry
Benjamin A. Steinberg, Elliott M. Antman, Nazanin
Moghbeli, Jacqueline Buros, Sabina A. Murphy,
Carolyn H. McCabe,C. Michael Gibson, David A.
Morrow, Eugene Braunwald
29
Background

• Significant regional differences in mortality
  rates following STEMI
• Lower mortality rates in clinical trials versus
  registries
• Patients in clinical trials more likely to get
  life-saving medications

Giugliano, et al. Eur Heart J. 2001221702-15.Ba
hit, et al. Am Heart J 2003145109-17.
30
Unanswered Questions

• How do contemporary RCT subjects differ from
  STEMI patients in the general population?
• How well can we generalize the results of the
  ExTRACT-TIMI 25 trial?
• How does regional variation contribute to
  mortality from STEMI?

31
Hypotheses

• STEMI patients in contemporary clinical trials
  have lower baseline risk and better outcomes than
  patients not enrolled in RCTs
• When adjusted for baseline risk and treatments
  received, regional variation in outcomes after
  STEMI is minimized

32
ExTRACT-TIMI 25 Program
STEMI
Entered inExTRACT-TIMI 25 Registry
Randomized inExTRACT-TIMI 25 Trial
Major In-Hospital Outcomes1 Death At
Discharge or Day 8
33
ExTRACT-TIMI 25 Chronology
ExTRACT-TIMI 25 Trial
October 31, 2005
January 24 ,2006
October, 2002
ExTRACT-TIMI 25 Registry
April 1, 2005
Registry Enrollment n0 n3,098 n3,726
34
Population Samples
Trial
Registry
3,726
Patients
20,479
(Intention-to-Treat)
109
674
25
48
Antman, et al. NEJM 20063541477-88.
35
Baseline Characteristics
Trial All (n 20,479) Trial Registry Sites(n 7,819) Registry (n 3,726) p (registry sites)
Age yr (median) 59 60 63 lt0.001
Female 23 25 28 0.003
Hypertension 44 50 58 lt0.001
Hyperlipidemia 18 16 34 lt0.001
Current Smoker 47 47 40 lt0.001
Diabetes 15 13 21 lt0.001
Prior MI 13 15 17 0.04
Prior angina pectoris 28 34 36 0.07
Prior PCI 3.2 1.8 3.8 lt0.001
Prior CABG 1.3 0.7 1.4 lt0.001
Antman, et al. NEJM 20063541477-88.
36
Baseline Characteristics Cont.
Trial All (n 20,479) Trial Registry Sites(n 7,819) Registry (n 3,726) p (registry sites)
Anterior MI 44 47 45 0.02
Chronic ASA 13 12 18 lt0.001
Creat. clearance ml/min (median) 82 82 73 lt0.001
Killip II-IV 11 13 26 lt0.001
Rx In-Hosp
Aspirin 94 94 95 0.005
Beta-blockers 81 82 78 lt0.001
ACEIs or ARBs 74 75 76 0.70
Statin 64 51 76 lt0.001
Antman, et al. NEJM 20063541477-88.
37
STEMI Management
Trial All (n 20,479) Trial Registry Sites(n 7,819) Registry (n 3,726) p (registry sites)
Reperfusion Therapy
Fibrinolysis
Fibrin-specific 80 86 61 lt0.001
Streptokinase 20 14 36 lt0.001

Fibrinolysis Only 86 94 80 lt0.001
Lytic?PCI 14 6 20 lt0.001
Primary PCI - - 26
No Reperfusion 0.2 0.2 29 lt0.001
Antman, et al. NEJM 20063541477-88.
38
In-Hospital Mortality
Plt0.001
HR 1.35
UnadjustedMortality
39
TIMI Risk Index (TRI)
HR ? (Age/10)2
SBP
P(trend)lt0.001
In-Hospital Mortality
TRI
Morrow, et al. Lancet. 20013581571-5.Wiviott,
et al. JACC. 200444783-9.
40
TIMI Risk Index (TRI) Profile
Plt0.001
Patient Population
HR ? (Age/10)2
SBP
41
In-Hospital Mortality by TIMI Risk Index (TRI)
p0.92
Mortality
P(trend)lt0.001
P(trend)lt0.001
42
Gross National Income

• Gross national income (GNI) per capita, as a
  surrogate for regionality
• Data from World Bank Development indicators
  database, 2004 statistics

World Bank. World Development Indicators
Database. Washington DC 2004.Orlandini, et al.
Eur Heart J. 2006 27527-33.
43
Registry TIMI Risk Index Profile By Gross
National Income
Patient Population
TRI
44
Registry Gross National Income (GNI) Corrected
for TRI
N3268 HR for In-Hospital Death P
TRI (per 10 unit increase) 1.18 lt0.001
GNI (log) 1.04 0.50
When adjusted for baseline risk, GNI does not
predict in-hospital mortality.
45
Registry Multivariate Analysis for In-Hospital
Mortality
n3501 HR for In-Hospital Death p
TRI (per 10 unit increase) 1.11 lt0.001
Any Reperfusion 0.74 0.02
ASA 0.69 0.04
Beta-blocker 0.24 lt0.001
ACE-I/ARB 0.24 lt0.001
Thienopyridine 0.44 lt0.001
Anticoagulation(antithrombin or warfarin) 0.56 lt0.001
GNI (log) 1.08 0.2
46
Risk of In-Hospital Mortality Statistical
Assessment
TRI Therapy c0.85
GNIc0.85
TRI
c0.77
Sensitivity
Coin Flipc0.5
C statistic(AUC for ROC)
1 - Specificity
47
Registry Predictors of In-Hospital Mortality
GNI
Medical ReperfusionTherapy
Overall c0.85
Baseline Risk (TRI)
48
Limitations

• Registry population not perfect representation of
  general STEMI population
• Reporting and survival biases

49
Conclusions

• TIMI Risk Index
• A simple, robust risk-stratification tool
• ExTRACT-TIMI 25 RCT
• Patients at lower risk and have better outcomes
  than the general STEMI population
• Mortality differences explained by baseline risk
• Regional Variation
• After adjusting for baseline risk (TRI) and
  in-hospital treatments, GNI does not predict
  in-hospital mortality after STEMI.