[name of file], 1
Pre- and Post-Marketing Continuum Models for
Cardiovascular Safety with Diabetic Drugs:
Academic View
Mitchell W. Krucoff, MD, FACC Professor, Medicine/Cardiology
Duke University Medical Center
Director, Cardiovascular Devices Unit
Co-Director, CSRC
Duke Clinical Research Institute
Cardiac Safety
Research Consortium
Challenges Underlying The Premise
Shifting from:
Safety as a barrier to innovation to
facilitating innovation and improving
safety evaluation
Nice talk to pragmatic direction & traction
www.cardiac-safety.org
Facilitating Innovation & Ensuring Safety:
Managing (Mis)Perception
No effective therapy is perfectly safe
Upper confidence interval boundary more
important than point estimate
Dichotomous “safe or not” less informative
than continuum of unknown: “residual risk”*
Underpowered/unrepresented:
CV safety endpoints in DM trials
Patient subgroups
Duration of exposure
* GHTF SG 5 final documents: http://www.ghtf.org/sg5/sg5-final.html
ISO 14971: http://www.isosert.ru/isosert_iso_14971.pdf
Cardiac Safety
Research Consortium
Classical NME Clinical R&D:
Separate & Unequal
Pre-market RCT(s):
Well defined population
Emphasis on efficacy
endpoints
Statistically robust for
primary questions
Highly controlled
monitoring & quality
Independent adjudication
Expensive & logistically
challenging
Post-market surveillance:
• Registry/single arm designs
• Large numbers
• Real world mix
• More efficient, less
expensive
• Event reporting, monitoring
& quality ???
• SAP ???
www.cardiac-safety.org
Pre- & Post-market continuum:
gaps & bridges
Consensus definitions
Integrated trial designs
Enriched populations &
“all comer” designs
Efficient quality solutions
From selected patients to
real world cohorts
From RCTs to registries
From controlled settings to
practice of medicine
Cardiac Safety
Research Consortium
[name of file], 6
The Academic Research Consortium
(ARC)*
Consensus Definitions
www.cardiac-safety.org
Consistency is more important than perfection!
*Krucoff MW et al, JACC Intervention & EuroIntervention, in press
…ethical, clinical, and
logistical constraints must
be integrated into the
balance of how best to
encourage innovation and
brisk access to better
therapies with potential
safety concerns and their
evaluation…One of the most
fundamental barriers to
meaningful data aggregation
is the use of different
nomenclature and definitions
of key descriptors and
endpoints from one trial to
another, or from one
manufacturer to another.”
Krucoff MW et al, JACC Intervention & EuroIntervention, in press
Cardiac Safety
Research Consortium
www.cardiac-safety.org
ARC LAST:
>500 reported trials
Cypher
Cypher Select
Taxus
Taxus Liberte
Xience
Endeavor
Resolute
Costar
Nevo
Biomatrix
Biofreedom
Absorb
Valve ARC: VARC
Standardized Bleeding Definitions for Cardiovascular
Clinical Trials: A Consensus Report from the Bleeding
Academic Research Consortium (BARC)
Roxana Mehran, Sunil V. Rao, Deepak L. Bhatt, C. Michael Gibson,
Adriano Caixeta, John Eikelboom, Sanjay Kaul, Stephen D. Wiviott,
Venu Menon, Eugenia Nikolsky, Victor Serebruany, Marco Valgimigli,
Pascal Vranckx, David Taggart, Joseph Sabik, Donald E. Cutlip, Mitchell
W. Krucoff, E. Magnus Ohman, Philippe Gabriel Steg, and Harvey White
Short title: Standardized Bleeding Definitions
Word count: 3889
Abstract word count:221
Bleeding ARC (BARC)
Cardiac Safety
Research Consortium
Mehran R et al, Circulation in press
[name of file], 11
Integrated Clinical Trial
Designs:
Pivotal RCTs &COA Registries
www.cardiac-safety.org
Case Study: The Xience V Stent
New DES (EES) platform
RCTs (vs Taxus) at PMA filing:
SPIRIT II, III, IV: 3,000 pt exposures to EES
Clinical & angio: non-inferior & superior
Residual risk concerns:
LAST: (are bigger lumens bad?)
0.6% vs 1.2% ???
Real world populations
Cardiac Safety
Research Consortium
Registry (n = 2,700), RCT Diabetics 2:1 vs. TAXUS (n = 300) OUS
Registry (n = 1,550) RCT 2:1 vs. CYPHER® (n = 450) OUS
Integrated Pre-Approval and Post-Approval
Clinical Program (N > 16,000)
SPIRIT First RCT 1:1 XIENCE V vs. VISION (n = 60) OUS
SPIRIT II RCT 3:1 XIENCE V vs. TAXUS® (n = 300) OUS
SPIRIT III RCT 2:1 XIENCE V vs. TAXUS (n = 1,002) US
SPIRIT III 4.0 Registry 4.0 mm (n = 80) US
SPIRIT III Japan Registry (n = 88) Japan
SPIRIT IV RCT XIENCE V vs. TAXUS 2:1 Continued Access (n = 3,690) US
SPIRIT V
XIENCE V SPIRIT Women
XIENCE V USA Post-approval Registry – real world (n ~ 5,000) US
Ongoing and Planned Clinical Data
XIENCE India Post-approval Registry – real world (n ~ 1,000) OUS
Pre-approval Clinical Data
RCTs in
progressively
complex lesions
Registries in
progressively
complex patients
& lesions
SE2930040 Rev. A
Gregg Stone, SPIRIT IV 1 Year Results, TCT 2009
© 2009 Abbott Laboratories
SPIRIT IV Stent Thrombosis (ARC Def or Prob)
Number at risk
XIENCE V 2458 2426 2412 2388 2376
TAXUS 1229 1195 1184 1174 1166
1.06%
0.29%
Months
p=0.003
HR [95%CI] =
0.27 [0.11, 0.67]
Δ 0.77%
Ste
nt
thro
mb
osis
(%
) XIENCE V
TAXUS
FOR IMPORTANT SAFETY INFORMATION SEE FINAL SLIDE.
SE2930040 Rev. A
Gregg Stone, SPIRIT IV 1 Year Results, TCT 2009
© 2009 Abbott Laboratories
Group
EES
(%)
PES
(%)
Relative Risk
(95% CI)
Relative Risk
(95% CI)
P
interaction
All randomized (n=3687) 4.2 6.8 0.62 [0.46, 0.82] —
Single lesion treated (n=2710) 3.9 5.7 0.68 [0.48, 0.96] 0.34
Two or more lesions treated (n=901) 5.1 10.0 0.51 [0.32, 0.83]
RVD > 2.75 mm (n=1351) 3.9 4.7 0.83 [0.49, 1.41] 0.29
RVD ≤ 2.75 mm (n=1352) 3.9 6.8 0.57 [0.35, 0.91]
Lesion length > 13.3 mm (n=1346) 4.5 6.9 0.65 [0.41, 1.03] 0.79
Lesion length ≤ 13.3 mm (n=1349) 3.2 4.5 0.71 [0.41, 1.23]
Bailout stent required (n=221) 5.0 14.5 0.35 [0.14, 0.86] 0.19
No bailout stent required (n=3390) 4.1 6.4 0.65 [0.48, 0.87]
Diabetes (n=1140) 6.4 6.9 0.94 [0.59, 1.49] 0.02
No diabetes(n=2467) 3.1 6.7 0.47 [0.32, 0.68]
SPIRIT IV Subgroup Analysis: TLF at 1 Year
TLF = cardiac death, target vessel MI, or ischemia-driven TLR; 1 Year = 365
28 days
Single lesion treated data are used for RVD & lesion length subgroup analyses
0.1 1 10
EES better PES better
FOR IMPORTANT SAFETY INFORMATION SEE FINAL SLIDE.
N~2,952
N~1,080
N~800
N~2,000
N~300
~80%
~36%
~40%
~40%
~30%
3 vessel CAD: N=738
Real world (EU): N=1620
Diabetics: N=300
Gender specific N~1,200
Real world (US) N~3,000
Optimal DAP
Real world (India) N~700
~64%
~60%
~60%
~70%
~20%
SPIRIT/XIENCE V Integrated Post-Approval Strategy
Post-Approval Trials N=14,690; (Randomized: 4,440)
Adding Certainty
(On Label)
Advance Knowledge
(Real World)
3 vessel CAD: N=738
Real world (EU): N=1620
Diabetics: N=300
Gender specific N~1,200
Real world (US) N~3,000
Optimal DAP
Real world (India) N~700
~64%
~60%
~60%
~70%
~20% SPIRIT IV
N~3,690
SPIRIT V
N~3,000
SPIRIT Women
N~2,000
XIENCE V USA
N~5,000
XIENCE V INDIA
N~1,000
Integrated Trial Design:
Xience V USA (COA Registry) & SPIRIT IV
Objective: LAST
Identical descriptor and endpoints definitions
Elimination of exclusion criteria (“Off label” pts):
Renal failure
Low EFx
STEMI
Anatomic: Lmain, CTO, bifurcation, SVG
Etc
Prospective analysis plan:
Event rates in matched Xience V & SPIRIT IV pts
Objective: quality reliability of benchmark in off-label pts
www.cardiac-safety.org
SE2932151 Rev. A
1 Year ST Rate (ARC Def/Prob) Overall Population
0.44%
0.39%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
ST
(A
RC
Def/
Pro
b)
(%)
Early (< 30 Days) Late (30 Days - 1 Year)
18
XIENCE V USA
XIENCE V
0.84%
SE2932151 Rev. A
0.11%0.22%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
ST
(A
RC
Def/
Pro
b)
(%)
Early (< 30 Days) Late (30 Days - 1 Year)
19
XIENCE V
0.34%
1-Year ST Rate (ARC Def/Prob) Standard Risk Cohort
*XIENCE V USA patients who have lesion length ≤ 28mm, RVD between 2.5mm and 4.25mm and do not have any one of the following: CTO, graft lesion, bifurcation
with side branch ≥ 2mm, ostial, LM, ISR, > 2 lesions stented in the same vessel, > 2 vessels treated, AMI, renal insufficiency, EF < 30%, staged procedure. †The
number of standard risk patients who have reached 1 year for ST analysis. Total number of standard risk cohort is 1,827, with 54 patients terminated before 1 year
without ST event and therefore excluded in this analysis.
XIENCE V USA
Standard Risk* (N = 1,773 †)
SE2932151 Rev. A
0.2%0.6%0.11% 0.2%
0.5%
0.22%0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
ST
(A
RC
Def/
Pro
b)
(%)
Early (< 30 Days) Late (30 Days - 1 Year)
20
SPIRIT III & SPIRIT IV**
XIENCE V TAXUS XIENCE V
0.34% 0.4%
1.0%
p = 0.03
*XIENCE V USA patients who have lesion length ≤ 28mm, RVD between 2.5mm and 4.25mm and do not have any one of the following: CTO, graft lesion, bifurcation
with side branch ≥ 2mm, ostial, LM, ISR, > 2 lesions stented in the same vessel, > 2 vessels treated, AMI, renal insufficiency, EF < 30%, staged procedure. †The
number of standard risk patients who have reached 1 year for ST analysis. Total number of standard risk cohort is 1,827, with 54 patients terminated before 1 year
without ST event and therefore excluded in this analysis. ** Presented by Dr. James Hermiller at SCAI 2010
XIENCE V USA
Standard Risk* (N = 1,773 †)
1-Year ST Rate (ARC Def/Prob) Standard Risk Cohort
SE2932151 Rev. A
Late ST Rates (30 Days - 1 Year) After DAPT Interruption
Su
bseq
uen
t L
ate
ST
(A
RC
Def/
Pro
b)
(%)
No Interruption Interruption
After 30 Days*
13/3500 0/292
Interruption
After 180 Days*
2/435
Interruption
After 90 Days*
1/378
Overall
21
*Out to 1 year
SE2932151 Rev. A
Late ST Rates (30 Days - 1 Year) After DAPT Interruption
Su
bseq
uen
t L
ate
ST
(A
RC
Def/
Pro
b)
(%)
No Interruption Interruption
After 30 Days*
13/3500 0/292
Interruption
After 180 Days*
2/435
Interruption
After 90 Days*
1/378 2/1272 0/157 0/147 0/120
Overall
Standard Risk
22
*Out to 1 year
[name of file], 23
Integrated Clinical Trial
Designs: Sustainable
Registries & RCTs:
A “River of Information”
www.cardiac-safety.org
Registries in the Regulatory Landscape:
A “river” of information:
Large numbers
Real world pts
Logistically simple
Prospective SAP, quality controlled
Informative: Rare event safety
Real world safety
Indication extensions
Database: modelling platforms
Informed prior
OPC
Infrastructure for nested RCTs
Leverage RCT findings
www.cardiac-safety.org
Prospective Hypothesis Registry Designs &
Device Evaluation for Regulatory Applications
FDA Guidance: Bayesian Models
(http://www.fda.gov/MedicalDevices/DeviceRegulationandG
uidance/GuidanceDocuments/ucm071072.htm)
FDA Guidance: Adaptive Trial Designs
Post-market COA studies:
Endeavor (Medtronic)
Xience V (Abbott)
Collaborative PPP safety trials:
INTERMACS*
DAPT**
SAFARI***
TREAT*** * NIH/FDA/Advamed
** CSRC/FDA/Advamed/Pharma
*** CSRC/FDA/Advamed/Pharma/ACC/HRS
[name of file], 26
Study of Access site For Enhancing
Percutaneous Coronary Intervention
for Women
(SAFE-PCI for Women) A Randomized Trial of Radial vs. Femoral Approaches in
Women Undergoing Percutaneous Coronary Intervention
Sunil Rao MD, PI
Mitchell Krucoff MD, Study Chair
Duke University Medical Center
Cardiac Safety
Research Consortium
http://www.ncdr.com/webncdr/common/
[name of file], 28
National Cardiovascular Research
Infrastructure
(NCRI)
NIH/NHLBI grant (#: 1RC2HL101512-01)
PI: Robert Harrington MD
David F. Kong MD
Duke Clinical Research Institute
www.cardiac-safety.org
EDC using NCRI* extraction from ACC-NCDR
Figure 1: NCRI Data Interface Strategy
NCRINCRI-only
Site
NCDR Site
Doing NCRI trial standard ACC process
Standard trial web data entry process
NCDR
NCDR d
ata
passed to
NCRI
trial d
atabase
* Tria
l spec
ific
data is
reporte
d
direct
ly
NCRINCRI-only
Site
NCDR Site
Doing NCRI trial standard ACC process
Standard trial web data entry process
NCDR
NCDR d
ata
passed to
NCRI
trial d
atabase
* Tria
l spec
ific
data is
reporte
d
direct
ly
* Trial specific data is data that is not included in NCDR data.
www.cardiac-safety.org
*NIH/NHLBI grant (#: 1RC2HL101512-01)
75% site based workload reduction
Part 11 compliant data structure
Pre- & Post-Market Continuum: Conclusions
Systematic approach to the pre- & post-
market as a continuum can both foster
innovation and improve safety information
“Residual risk” provides a useful,
continuous, quantitative construct to
support this continuum
Consistent use of consensus definitions,
integrated RCT & registry study designs and
integrated quality registries and RCT
infrastructure provide pragmatic tools with
which to build traction toward this end
[name of file], 31
Pre- and Post-Marketing Continuum Models for
Cardiovascular Safety with Diabetic Drugs:
Academic View
Mitchell W. Krucoff, MD, FACC Professor, Medicine/Cardiology
Duke University Medical Center
Director, Cardiovascular Devices Unit
Co-Director, CSRC
Duke Clinical Research Institute
Cardiac Safety
Research Consortium