Title: Cardiovascular Assessment in the PreApproval and PostApproval Settings for Drugs and Biologics Devel
1Cardiovascular Assessment in the Pre-Approval
and Post-Approval Settings for Drugs and
Biologics Developed for the Treatment of Type 2
Diabetes MellitusEndocrinologic and Metabolic
Drugs Advisory Committee July 1-2,
2008Hylton V. Joffe, M.D., M.M.Sc.Lead Medical
Officer, Diabetes Drug GroupU.S. Food and Drug
Administration
2Overview
- Introduction
- Type 2 diabetes mellitus
- FDA approach to diabetes drug development
- Cardiovascular assessment considerations
- Points for discussion and questions to the panel
3Introduction
- Anti-diabetic drugs are indicated to improve
glycemic control and are approved on the basis of
HbA1c - Safety concerns with some anti-diabetic drugs
have led to suggestions for a more extensive
cardiovascular assessment during the approval
process
N Engl J Med. 2007 357844 N Engl J Med.
2007 3571775-7
4Introduction
- This advisory committee meeting will explore this
proposal and associated complex issues - Should a long-term cardiovascular trial be
required? - Show benefit vs. rule out harm?
- All new therapies or only those with a safety
signal? - Challenges related to trial design?
- Timing relative to approval?
- What to do with currently marketed therapies?
5Presentations
- Natural history of type 2 diabetes and
macrovascular complications (Dr. David Nathan) - HbA1c as a surrogate for glycemic control and
microvascular complications (Dr. Robert Ratner) - Evaluating benefit and risk in type 2 diabetes
statistical considerations (Dr. Thomas Fleming) - Clinical macrovascular outcomes with
anti-diabetic drugs What we already know
(Professor Rury Holman)
6Presentations (continued)
- Clinical macrovascular outcomes with
anti-diabetic drugs Ongoing studies (Dr. Hertzel
Gerstein) - Need for cardiovascular assessment during
approval process for antidiabetic drugs (Dr.
Steven Nissen) - Challenges in designing a cardiovascular trial
for type 2 diabetes (Dr. Robert Califf)
7Agenda
- Day 1
- Introductory FDA presentation
- Presentations from experts in the field
- Panel questions to the presenters
- Panel discussion (will be continued on Day 2)
- Day 2
- Open public hearing
- FDA Comments (Dr. Mary Parks)
- Continued panel discussion
- Questions to the panel
8Type 2 Diabetes Mellitus
- gt150 million people worldwide
- gt18 million people in the United States
- 2-4 fold higher risk of cardiovascular death
- Most deaths due to cardiovascular disease/stroke
- Other important long-term complications
- Peripheral vascular disease
- Retinopathy, nephropathy, and neuropathy
N Engl J Med. 1998 339229-34 JAMA. 2002
2872570-81 Lancet. 2005 3651333-46
9Pharmacological Therapies
- Alpha-glucosidase inhibitors
- Amylin analogues (pramlintide acetate)
- Biguanides (metformin)
- Bile acid sequestrants (colesevelam)
- Dipeptidyl peptidase 4 inhibitors (sitagliptin
phosphate) - Glinides
- Glucagon-like peptide 1 analogues (exenatide)
- Insulin
- Sulfonylureas
- Thiazolidinediones
10Macrovascular Complications
- Intensive glycemic control appears to reduce
macrovascular complications in type 1 diabetes - No conclusive evidence of macrovascular risk
reduction with any of the FDA-approved treatments
for type 2 diabetes
11Older Trials with Cardiovascular Findings
- University Group Diabetes Program (UGDP)
- Tolbutamide increased cardiovascular mortality
- UK Prospective Diabetes Study (UKPDS)
- Non-significant reduction (p0.052) in myocardial
infarction/ sudden death with intensive therapy - Reduction in diabetes-related death and all-cause
mortality in a substudy of 342 overweight
patients given metformin - Increase in diabetes-related death with metformin
add-on to sulfonylurea
Diabetes. 1970 19 (Supp 2) 747-830 Lancet.
1998 352837-53 and 854-65
12Trials with Cardiovascular Assessments
- ACCORD
- ADVANCE
- BARI 2D
- HEART 2D
- VADT
- PROactive
- RECORD
- ORIGIN
- NAVIGATOR
- ACE
- STOP-NIDDM
- DREAM
Treatment regimens
Type 2 diabetes
Primary CV or mortality endpoint
Specific drugs
Pre-diabetes
Secondary CV endpoint
13Current FDA Approval Process
- Anti-diabetic drugs are indicated to improve
glycemic control - HbA1c is the primary efficacy endpoint
- Correlates with mean blood glucose over 3 months
- Unlike some other biomarkers relied upon for drug
approval, there is symptomatic benefit with daily
control of hyperglycemia - Lowering of HbA1c reduces the risk of onset,
progression of microvascular complications
Lancet. 1998 352837-53 N Engl J
Med. 2008 3582560-72
N Engl J Med. 1993 329977-86 Diabetes Care.
2004 27 1761-73
14Package Inserts
- Drug X is indicated as an adjunct to diet and
exercise to improve glycemic control in adults
with type 2 diabetes mellitus - There have been no clinical studies establishing
conclusive evidence of macrovascular risk
reduction with Drug X or any other oral
antidiabetic drug - Package inserts do not mention improvement in
long-term sequelae of diabetes
15Phase 2 Program
- Dose-finding, randomized, double-blind,
controlled, trials in patients treated only with
diet/exercise or on a stable dose of metformin
Dose A
Dose B
Dose C
Run-in
Placebo
12 weeks
FDA Guidance to Industry on Diabetes Mellitus
(Draft)
16Phase 3 Program
- 6-month, randomized, double-blind, controlled
trials with 6-18 month extensions - Placebo-controlled or active-controlled
(non-inferiority) - Monotherapy
- Add-on to other commonly used anti-diabetic drugs
- Placebo-controlled trials usually 6 months in
duration to limit long-term exposure to
hyperglycemia -
FDA Guidance to Industry on Diabetes Mellitus
(Draft)
17Phase 3 Program - Monotherapy
- Enrollment of patients treated with diet and
exercise only or after washout of a single
anti-diabetic drug - Generally low cardiovascular risk
Dose A
Dose B
Run-in
Placebo (or active comparator)
24 weeks
FDA Guidance to Industry on Diabetes Mellitus
(Draft)
18Phase 3 Program Add-on Trials
Enrollment of patients with HbA1c 7-10 despite
stable maximal/near-maximal doses of a background
anti-diabetic drug
Stable near-maximal/maximal doses of background
anti-diabetic drug
Dose A background anti-diabetic drug
Dose B background anti-diabetic drug
Run-in
Placebo background anti-diabetic drug
24 weeks
FDA Guidance to Industry on Diabetes Mellitus
(Draft)
19Phase 3 Development Program
- A typical program has the following core phase 3
trials - Placebo-controlled monotherapy trial
- Add-on to metformin
- Add-on to sulfonylurea
- Add-on to thiazolidinedione
- Other phase 3 trials
- Active-controlled, monotherapy trial
- Add-on to dipeptidyl-peptidase 4 inhibitor
- Add-on to insulin
- Add-on to dual agents (e.g., metformin
sulfonylurea)
20Phase 3 Program Extension Trials
- In active-controlled trials, the randomized
treatments can be continued into the extension
trial - In placebo-controlled trials, placebo-treated
patients are given active therapy during the
extension trial to limit prolonged hyperglycemia
(e.g., another anti-diabetic drug or the
investigational agent) - Uncontrolled extensions limit interpretability of
efficacy and safety
21Phase 3 Program - Efficacy
- Primary efficacy endpoint HbA1c
- Sensitivity analyses
- Subgroup analyses
- Secondary efficacy endpoints
- Fasting plasma glucose
- HbA1c responder analyses
- e.g., proportion of patients achieving HbA1c lt7
- Body weight
- Mechanistic (e.g., insulin sensitivity,
postprandial glucose)
22Phase 3 Program - Safety
- Deaths
- Non-fatal serious adverse events
- Discontinuations
- Other adverse events
- Events of special interest (some are
drug-specific) - Laboratory data
- Vital signs
- Electrocardiograms
23Phase 3 Program - Safety
- Unpooled analyses
- Individual trial data
- Pooled analyses
- Grouping of data from two or more similar trials
- Useful for analyzing infrequent events (e.g.,
deaths) - Meta-analyses
- Not routinely performed but multiple clinical
trials required for the marketing application
may allow for this to be incorporated in the
safety analysis
24Phase 3 Program - Safety
- Some Caveats
- Multiplicity if 100 associations are tested, 5
may be significant by chance (alpha0.05) - Studies rarely powered for safety
- Assessments of infrequent events (e.g., deaths,
myocardial ischemia) are inconclusive - Do not usually have adjudication committees
25Phase 3 Program Sample Sizes
- ICH E1A recommended pre-approval exposures for
drugs developed for chronic, non-life-threatening
conditions - 1,500 subjects total
- 300-600 subjects for 6 months
- 100 subjects for 1 year
- Minimum pre-approval sample sizes for type 2
diabetes - 2,500 phase 2/3 total
- 1,300-1,500 exposed 1 year
- 300-500 exposed 18 months
- Specific safety concern(s) may require larger
exposures
26Rule of Three
- If we expose 2,500 patients to the study drug and
see no cases of event A then we have ruled out
incidence rates for event A of 0.12 or higher
with 95 certainty
27Clinical Trial Challenges
- Worsening glycemia over time if therapy is not
altered - Protecting participants from prolonged
hyperglycemia - HbA1c entry criteria
- Duration of placebo-controlled phase
- Glycemic rescue therapy
- Progressive nature of type 2 diabetes
necessitates a multidrug anti-diabetic regimen in
long-term trials, limiting the ability to tease
apart the effects of a single drug
28Points for Consideration and Discussion
- What changes are recommended to phase 2/3 trials
to enhance detection of a cardiovascular safety
signal prior to drug approval? - Independent, blinded cardiovascular adjudication?
- Meta-analysis of safety data from phase 2/3
trials? - Changes to sample sizes/duration of exposures?
29Points for Consideration and Discussion
- Should a long-term cardiovascular trial be
required to show benefit or rule out an
unacceptable increase in cardiovascular risk? - Conclusive evidence of cardiovascular benefit has
not been established for any drug for type 2
diabetes despite several large, long-term trials - If ruling out harm, what non-inferiority margin
do you recommend?
30Points for Consideration and Discussion
- In the absence of a concerning cardiovascular
safety signal during Phase 2/3, should there
still be a requirement to conduct a long-term
cardiovascular trial? - When should such a study be conducted?
- What about already marketed therapies?
31Large Cardiovascular Trial Discussion Points
- Benefit vs. rule out harm (if harm, what
magnitude)? - Patient population?
- Comparator(s)?
- Primary endpoint?
- What should the HbA1c target be?
- How to define and manage deteriorating glycemia?
- How to manage other cardiovascular risk factors?
- Whether to ensure comparability of glycemia and
cardiovascular risk factors across treatment
groups? - Trial size and duration?
32Large Cardiovascular Trial Discussion Points
- Patient population?
- Pre-diabetes? New-onset diabetes? Longstanding
diabetes? - Generalizability?
- Statistical power?
- DREAM (pre-diabetes without cardiovascular
disease) - 1 with MACE over median 3 year follow-up
- ADOPT (new-onset diabetes)
- Up to 25 with inadequate glycemic control over
4-6 years
33Large Cardiovascular Trial Discussion Points
- What should the comparator be?
- Drug X vs. Placebo
- Drug X vs. Placebo as add-on to standard therapy
- Drug X vs. Drug Y as add-on to standard therapy
- How should deteriorating glycemia (which may vary
between treatment groups) be defined, managed? - How do we interpret the effects of one drug
within a multidrug regimen? - If ruling out harm in an active-controlled trial,
how much do we need to know about the
cardiovascular effects of the comparator?
34Large Cardiovascular Trial Discussion Points
- What should the primary endpoint be? Composite?
- Cardiovascular death (or all-cause mortality)
- Nonfatal myocardial infarction
- Nonfatal stroke
- Include other events in the primary endpoint?
- Coronary revascularization?
- Lower-extremity amputations?
- Predefine, justify, accurately capture and analyze
352008 ADA Treatment Goals in Diabetes
36Large Cardiovascular Trial Discussion Points
- Should investigators be encouraged to manage
blood pressures, lipid profiles, aspirin use, and
other cardiovascular factors to current
guidelines (which will not necessarily ensure
comparability across treatment groups)? - OR
- Should algorithms be used post-randomization with
the intent of equalizing these risk factors
across treatment groups?
37Total Sample Sizes
a0.05 90 power 5-year trial 2-year
recruitment
38Questions to the Committee
- It should be assumed that an anti-diabetic
therapy with a concerning CV safety signal during
Phase 2/3 development will be required to conduct
a long-term cardiovascular trial. For those
drugs or biologics without such a signal, should
there be a requirement to conduct a long-term
cardiovascular trial ? (vote yes/no
requested) - If yes, please discuss when such a study should
be conducted? - Pre-approval?
- Post-approval? If a long-term cardiovascular
trial is required post-approval, please discuss
whether this study should be ongoing at the time
of approval (i.e., trial already initiated at
time of approval)
39Questions to the Committee
- No currently marketed therapy for type 2 diabetes
has established conclusive evidence of
macrovascular benefit - Most marketed therapies for type 2 diabetes have
not been tested for lack of cardiovascular harm - Therefore, please discuss how any suggestion for
a requirement for a long-term cardiovascular
trial should be applied to existing anti-diabetic
therapies
40Acknowledgements
- Ilan Irony, M.D
- Karen Mahoney, M.D.
- Robert Misbin, M.D.
- Valerie Pratt, M.D.
- Joanna Zawadzki, M.D.
- Joy Mele, M.S.
- Todd Sahlroot, Ph.D.
- Mary Parks, M.D. Director, Division of Metabolism
and Endocrinology Products - Curtis Rosebraugh, M.D., M.P.H. Director, Office
of Drug Evaluation II - Robert Temple, M.D. Associate Director, Office of
Medical Policy - John Jenkins, M.D. Director, Office of New Drugs
- Gerald Dal Pan, M.D., M.H.S. Director, Office of
Surveillance Epidemiology
FDA clinical diabetes team
Statistics