Transcript for the June 9, 2015 Meeting of the Endocrinologic and Metabolic Drugs Advisory ...
October 30, 2017 | Author: Anonymous | Category: N/A
Short Description
.. mortality in a large cardiovascular outcomes trial. 21. Janet Evans-Watkins 06-09-15 FDA EMDAC - Revised 07 ......
Description
1
1
FOOD AND DRUG ADMINISTRATION
2
CENTER FOR DRUG EVALUATION AND RESEARCH
3 4 5
ENDOCRINOLOGIC AND METABOLIC DRUGS
6
ADVISORY COMMITTEE (EMDAC)
7 8 9 10 11
Tuesday, June 9, 2015
12
8:01 a.m. to 5:17 p.m.
13 14 15 16
Hilton Washington DC North/Gaithersburg
17
620 Perry Parkway
18
Gaithersburg, Maryland
19 20 21 22
A Matter of Record (301) 890-4188
2
Meeting Roster
1 2
DESIGNATED FEDERAL OFFICER (Non-Voting)
3
Philip A. Bautista, PharmD
4
Division of Advisory Committee and
5
Consultant Management
6
Office of Executive Programs, CDER, FDA
7 8
ENDOCRINOLOGIC AND METABOLIC DRUGS ADVISORY
9
COMMITTEE MEMBERS (Voting)
10
David W. Cooke, MD
11
Associate Professor of Pediatrics
12
Clinical Director, Division of Pediatric
13
Endocrinology
14
Director, Pediatric Endocrine Fellowship Training
15
Program
16
Johns Hopkins University School of Medicine
17
Baltimore, Maryland
18 19 20 21 22
A Matter of Record (301) 890-4188
3
1
Brendan M. Everett, MD, MPH
2
Assistant Professor of Medicine
3
Harvard Medical School
4
Director, General Cardiology Inpatient Service
5
Brigham and Women’s Hospital
6
Boston, Massachusetts
7 8
William R. Hiatt, MD, FACP, FAHA
9
Professor of Medicine
10
Division of Cardiology
11
University of Colorado School of Medicine
12
President, Colorado Prevention Center
13
Clinical Research
14
Aurora, Colorado
15 16
Robert J. Smith, MD
17
(Chairperson)
18
Professor of Medicine (Endocrinology)
19
Alpert Medical School of Brown University
20
Ocean State Research Institute
21
Providence Veterans Administration Medical Center
22
Providence, Rhode Island
A Matter of Record (301) 890-4188
4
1
Charles A. Stanley, MD
2
Professor of Pediatrics
3
Perelman School of Medicine University of
4
Pennsylvania
5
Division of Endocrinology & Diabetes
6
Children’s Hospital of Philadelphia
7
Philadelphia, Pennsylvania
8 9
Peter W. F. Wilson, MD
10
Director
11
Epidemiology and Genomic Medicine
12
Atlanta Veterans Administration Medical Center
13
Professor of Medicine and Public Health
14
Emory University
15
Emory Clinical Cardiovascular Research Institute
16
Atlanta, Georgia
17 18 19 20 21 22
A Matter of Record (301) 890-4188
5
1
TEMPORARY MEMBERS (Voting)
2
Michael J. Blaha, MD, MPH
3
Director of Clinical Research
4
Ciccarone Center for Prevention of Heart Disease
5
Assistant Professor
6
Johns Hopkins University School of Medicine
7
Baltimore, Maryland
8 9
Daniel Budnitz, MD, MPH
10
CAPT, US Public Health Service
11
Director, Medication Safety Program
12
Division of Healthcare Quality Promotion
13
Centers for Disease Control and Prevention
14
Atlanta, Georgia
15 16
Kenneth D. Burman, MD
17
Professor of Medicine
18
Georgetown University and Uniformed Services
19
University of the Health Sciences
20
Director, Section of Endocrinology
21
MedStar Washington Hospital Center
22
Washington, District of Columbia
A Matter of Record (301) 890-4188
6
1
Nancy Geller, PhD
2
Chief, Biostatistics Research Branch
3
Director, Office of Biostatistics Research
4
National Heart, Lung, and Blood Institute (NHLBI)
5
NIH
6
Bethesda, Maryland
7 8
Debra G. McCall, MBA
9
(Patient Representative)
10
Murrieta, California
11 12
Martha Nason, PhD
13
Mathematical Statistician
14
Biostatistics Research Branch
15
Division of Clinical Research
16
National Institute of Allergy and Infectious
17
Diseases
18
National Institutes of Health (NIH)
19
Bethesda, Maryland
20 21 22
A Matter of Record (301) 890-4188
7
1
Michele Orza, ScD
2
(Acting Consumer Representative)
3
Senior Advisor to the Executive Director
4
Patient-Centered Outcomes Research Institute
5
Washington, District of Columbia
6 7
Philip Sager, MD, FACC, FAHA
8
Consulting Professor of Medicine
9
Stanford University School of Medicine
10
Chair, Scientific Programs Committee
11
Cardiac Safety Research Consortium
12
San Francisco, California
13 14
Robert D. Shamburek, MD
15
Clinician/Scientist
16
Lipoprotein Metabolism Section
17
Cardiovascular and Pulmonary Branch
18
NHLBI, NIH
19
Bethesda, Maryland
20 21 22
A Matter of Record (301) 890-4188
8
1
Abraham Thomas, MD, MPH
2
Senior Vice President and Chair
3
Department of Medicine
4
New York University Lutheran
5
Brooklyn, New York
6 7
ACTING INDUSTRY REPRESENTATIVE TO THE COMMITTEE
8
(Non-Voting)
9
Helmut H. Albrecht, MD, MS, FFPM
10
(Acting Industry Representative)
11
Principal, H2A Associates, LLC
12
Miami, Florida
13 14 15
FDA PARTICIPANTS (Non-Voting)
16
Mary H. Parks, MD
17
Deputy Director
18
Office of Drug Evaluation II (ODE II)
19
Office of New Drugs (OND), CDER, FDA
20 21 22
A Matter of Record (301) 890-4188
9
1
James P. Smith, MD, MS
2
Deputy Director
3
DMEP, ODE II, OND, CDER, FDA
4 5
Mary Dunne Roberts, MD
6
Clinical Reviewer
7
DMEP, ODE II, OND, CDER, FDA
8 9
Jean-Marc Guettier, MD
10
Director
11
Division of Metabolism and Endocrinology
12
Products (DMEP)
13
ODE II, OND, CDER, FDA
14 15
Julie Golden, MD
16
Medical Officer
17
DMEP, ODE II, OND, CDER, FDA
18 19 20 21 22
A Matter of Record (301) 890-4188
10
C O N T E N T S
1 2
AGENDA ITEM
3
Call to Order and Introduction of Committee
4 5 6
PAGE
Robert Smith, MD Conflict of Interest Statement Philip Bautista, PharmD
7
FDA Introductory Remarks
8
James Smith, MD, MS
9 10 11
Jay Edelberg, MD, PhD
13
Management Options
16 17 18 19 20 21
21
Introduction
Clinical Need for Additional Cholesterol
15
17
Applicant Presentations – Sanofi Aventis
12
14
12
Christopher Cannon, MD
42
50
Study Design and Efficacy Bill Sasiela, PhD
57
Alirocumab Safety Ned Braunstein, MD
74
Benefit-Risk Summary Robert Eckel, MD Clarifying Questions
22
A Matter of Record (301) 890-4188
95 100
11
C O N T E N T S (continued)
1 2
AGENDA ITEM
3
FDA Presentations
4
Clinical Efficacy Review
5 6
PAGE
Julie Golden, MD Clinical Safety Review
7
Mary Dunne Roberts, MD
8
Risk/Benefit Considerations
9
135
Julie Golden, MD
152
179
10
Clarifying Questions
181
11
Open Public Hearing
213
12
Clarifying Questions (continued)
260
13
Questions to the Committee and Discussion
298
14
Adjournment
416
15 16 17 18 19 20 21 22
A Matter of Record (301) 890-4188
12
P R O C E E D I N G S
1 2
Call to Order
3
Introduction of Committee DR. SMITH:
4
Good morning.
I'd first before
5
starting here like to remind everyone to please
6
silence your cell phones, smartphones, or other
7
devices that make noise if you've not already done
8
so.
9
contact, Eric Pahon.
10 11
I'd also like to identify the FDA press If you are here, Eric, please
stand; waving his hand in the back there. My name is Robert Smith.
I'm the
12
chairperson for the Endocrinologic and Metabolic
13
Drugs Advisory Committee.
14
meeting of the Endocrinologic and Metabolic Drugs
15
Advisory Committee to order.
16
going around the table and asking the members of
17
the advisory panel and FDA participants to
18
introduce yourselves, and we'll start down here on
19
the right.
20
DR. ALBRECHT:
I will now call this
And I will start by
Helmut Albrecht.
I'm the
21
industry representative.
I work for H2a
22
Associates, LLC, which is a consulting company for
A Matter of Record (301) 890-4188
13
1
the pharmaceutical industry.
2
faculty appointment at Florida International
3
University. DR. COOKE:
4
I have an adjunct
I'm David Cooke.
I'm an
5
associate professor of pediatrics at Johns Hopkins,
6
and I'm in the division of pediatric endocrinology
7
there.
8 9 10 11
DR. STANLEY:
Charles Stanley.
I'm a
pediatric endocrinologist at Children's Hospital of Philadelphia. CAPT BUDNITZ:
Dan Budnitz.
I'm a medical
12
epidemiologist with the Division of Healthcare
13
Quality Promotion at the Centers for Disease
14
Control and Prevention.
15
DR. ORZA:
I'm Michelle Orza.
I'm with the
16
Patient-Centered Outcomes Research Institute, and
17
I'm the acting consumer representative.
18 19 20
MS. MCCALL:
I'm Debra McCall.
I'm the
patient representative. DR. BURMAN:
Ken Burman, chief of
21
endocrinology at MedStar Washington Hospital Center
22
and professor at Georgetown University.
A Matter of Record (301) 890-4188
14
DR. THOMAS:
1
Abraham Thomas, senior vice
2
president and chairman of medicine at NYU Lutheran
3
in Brooklyn, New York. DR. BLAHA:
4
I'm Mike Blaha, director of
5
clinical research, Hopkins Ciccarone Center for
6
Prevention of Heart Disease.
7
DR. BAUTISTA:
8
Philip Bautista.
9
officer of EMDAC.
My name is
I'm the designated federal
DR. SMITH:
10
Good morning.
And I'm Robert Smith.
I'm
11
professor of medicine and professor of public
12
health at Brown University in Providence, Rhode
13
Island.
14
DR. HIATT:
William Hiatt.
I'm a professor
15
of medicine in the division of cardiology at the
16
University of Colorado School of Medicine.
17
specialize in vascular medicine.
18
DR. WILSON:
I
Peter Wilson, professor of
19
medicine at Emory University, professor of public
20
health, endocrinology, preventive medicine, and
21
epidemiology.
22
DR. GELLER:
I'm Nancy Geller.
A Matter of Record (301) 890-4188
I'm the
15
1
director of the Office of Biostatistics Research at
2
the National Heart, Lung, and Blood Institute. DR. EVERETT:
3
I'm Brendan Everett.
I'm the
4
director of general inpatient cardiology at Brigham
5
and Women's Hospital and Harvard Medical School in
6
Boston. DR. SHAMBUREK:
7
I'm Bob Shamburek with the
8
lipoprotein metabolism section in the NHLBI,
9
intramural of NIH.
10
DR. SAGER:
Philip Sager.
I'm a consulting
11
professor of medicine at Stanford University.
12
a cardiologist and also have expertise in drug
13
safety.
14
DR. NASON:
Martha Nason.
I'm
I'm a
15
mathematical statistician at the National Institute
16
for Allergy and Infectious Diseases.
17
DR. ROBERTS:
My name is Mary Roberts.
I'm
18
a medical officer in the Division of Metabolism and
19
Endocrinology Products.
20 21 22
DR. GOLDEN:
Julie Golden, medical officer,
metabolism and endocrinology, FDA. DR. SMITH:
James Smith, deputy division
A Matter of Record (301) 890-4188
16
1
director, Division of Metabolism and Endocrinology
2
Products.
3
DR. GUETTIER:
Jean-Marie Guettier, division
4
director of the Division of Metabolism and
5
Endocrinology Products.
6 7
DR. PARKS:
Mary Parks, deputy director,
Office of Drug Evaluation II.
8
DR. SMITH:
Thank you.
9
Now, for topics such as those being
10
discussed at today's meeting, there are often a
11
variety of opinions, some of which are quite
12
strongly held.
13
will be a fair and open forum for discussion of
14
these issues and that individuals can express their
15
views without interruption.
16
reminder, individuals will be allowed to speak into
17
the record only if recognized by the chairperson.
18
We look forward to a productive meeting.
19
Our goal is that today's meeting
Thus, as a gentle
In the spirit of the Federal Advisory
20
Committee Act and the Government in the Sunshine
21
Act, we ask that the advisory committee members
22
take care that their conversations about the topic
A Matter of Record (301) 890-4188
17
1
at hand take place in the open forum of the
2
meeting.
3
are anxious to speak with the FDA about these
4
proceedings.
5
discussing the details of this meeting with the
6
media until its conclusion.
7
reminded to please refrain from discussing the
8
meeting topic during breaks or lunch.
9
We are aware that members of the media
However, FDA will refrain from
Also, the committee is
Thank you.
Now, I will pass the microphone to Phil
10
Bautista, who will read the conflict of interest
11
statement.
12
Conflict of Interest Statement
13
DR. BAUTISTA:
Thank you.
14
The FDA is convening today's meeting of the
15
EMDAC under the authority of FACA of 1972.
16
the exception of the industry representative, all
17
members and temporary voting members of the
18
committee are SGEs or regular federal employees
19
from other agencies and are subject to the federal
20
conflict of interest laws and regulations.
21 22
With
The following information on the status of this committee's compliance with federal ethics and
A Matter of Record (301) 890-4188
18
1
conflict of interest laws, covered by but not
2
limited to those found at 18 U.S.C., Section 208,
3
is being provided to participants in today's
4
meeting and to the public.
5
FDA has determined that members and
6
temporary voting members of this committee are in
7
compliance with the federal ethics and conflict of
8
interest laws.
9
authorized FDA to grant waivers to SGEs and regular
Under Section 208, Congress has
10
federal employees who have potential financial
11
conflicts, when it is determined that the agency's
12
need for a particular individual's services
13
outweighs his or her potential financial conflict
14
of interest.
15
Related to the discussion of today's
16
meeting, members and temporary voting members of
17
this committee have been screened for potential
18
financial conflicts of interest of their own, as
19
well as those imputed to them, including those of
20
their spouses or minor children, and for the
21
purposes of Section 208, their employers.
22
interests may include investments, consulting,
A Matter of Record (301) 890-4188
These
19
1
expert witness testimony, contracts, grants,
2
CRADAs, teaching, speaking, writing, patents and
3
royalties, and primary employment.
4
Today's agenda involves the discussion of
5
the safety and efficacy of BLA 12559, proposed
6
trade name Praluent, established name alirocumab,
7
for injection, submitted by Sanofi Aventis USA as
8
an adjunct to diet for long-term treatment of adult
9
patients with primary hypercholesterolemia,
10
non-familial and heterozygous familial, or mixed
11
dyslipidemia, including patients with type 2
12
diabetes mellitus, to reduce LDL cholesterol, total
13
cholesterol, non-HDL cholesterol, apolipoprotein B,
14
triglycerides, and lipoprotein A, and increased HDL
15
cholesterol, and apolipoprotein A1, either in
16
combination with a statin or as monotherapy,
17
including patients who cannot tolerate statins.
18
This is a particular matters meeting during
19
which specific matters related to Sanofi's
20
alirocumab will be discussed.
21
for today's meeting and all financial interests
22
reported by the committee members and temporary
A Matter of Record (301) 890-4188
Based on the agenda
20
1
voting members, no conflict of interest waivers
2
have been issued in connection with this meeting.
3
To ensure transparency, we encourage all
4
standing committee members and temporary voting
5
members to disclose any public statements that they
6
have made concerning the topic at issue.
7
With respect to FDA's invited industry
8
representative, we would like to disclose that
9
Dr. Helmut Albrecht is participating in this
10
meeting as a non-voting industry representative,
11
acting on behalf of regulated industry.
12
Dr. Albrecht's role at this meeting is to represent
13
industry in general and not any particular company.
14
Dr. Albrecht is employed by H2a Associates and
15
Alitair Pharmaceuticals.
16
We would like to remind members and
17
temporary members that if the discussions involve
18
any other products or firms not already on the
19
agenda for which an FDA participant has a personal
20
or imputed financial interest, the participants
21
need to exclude themselves from such involvement,
22
and their exclusion will be noted for the record.
A Matter of Record (301) 890-4188
21
FDA encourages all other participants to
1 2
advise the committee of any other financial
3
relationships they may have with the firm at issue.
4
Thank you. DR. SMITH:
5
Thank you, Phil.
We'll now
6
proceed with the FDA's introductory remarks from
7
Dr. Jim Smith. FDA Introductory Remarks – James Smith
8
DR. J. SMITH:
9
Good morning.
My name is Jim
10
Smith, and I would like to take this opportunity to
11
thank all of you for being here and for all the
12
time that you've already spent reviewing the
13
hundreds of pages of materials that we've sent to
14
you.
15
this application.
16
We are really looking forward to discussing
In my introductory remarks over the next 15
17
to 20 minutes, I'd like to provide you with just a
18
bit of history regarding the regulatory use of LDL
19
cholesterol as a surrogate endpoint and what that
20
means, and some of the regulatory issues that we
21
need to deal with and think about.
22
to walk through the discussion points that we had
A Matter of Record (301) 890-4188
Then I'd like
22
1
provided you as well as the voting question.
2
Now, although the approval of lipid-
3
modulating drugs on the basis of changes in
4
biomarkers such as serum cholesterol dates back
5
more than 50 years, I'm only going to take you back
6
to 1987, when only eight advisors of this committee
7
recommended approval of Mevacor, or lovastatin,
8
after presentations and deliberations that together
9
lasted less than four hours.
10
At that time, bile acid binding resins such
11
as cholestyramine were the drugs of choice for the
12
treatment of hypercholesterolemia, and clinicians
13
were desperate for more effective therapies and
14
tolerable therapies to reduce the risk of
15
cardiovascular disease.
16
Lipid Research Clinics trial had just finished a
17
few years earlier.
18
controlled cardiovascular outcomes trial in which
19
the average baseline LDL cholesterol exceeded
20
200 milligrams per deciliter.
21
a much different time.
22
You may recall that the
And that was a placebo-
Certainly, that was
At that meeting, little direct attention was
A Matter of Record (301) 890-4188
23
1
given to the fact that efficacy was based solely on
2
a biomarker.
3
deliberated earlier that morning about whether
4
studies ought to be required prior to the approval
5
to evaluate atherosclerosis by various imaging
6
methods.
7
But the same committee had
Although the committee had felt that such
8
information would be desirable, it appears that
9
they believe that the methodology was not
10
sufficiently developed.
11
discussed lovastatin later that day, they certainly
12
recognized the true clinical benefit remained
13
uncertain and the target population could be very
14
large, leading to an in-depth discussion regarding
15
the safety of the product.
16
When the committee
The chairman, Dr. Frederick Singer, closed
17
the meeting with, "I think our recommendation is to
18
approve this drug and hope that it does as much as
19
we think it might, but that it should be used
20
carefully."
21
first of eight statins approved, seven of which
22
remain on the market today.
Of course, lovastatin was only the
And each statin has
A Matter of Record (301) 890-4188
24
1
been approved on the basis of effects on LDL
2
cholesterol primarily.
3
It was not until the 1994 publication of
4
what is commonly referred to as the 4S trial,
5
however, the lowering of LDL cholesterol by a
6
statin was first shown to reduce the risk of
7
mortality in the setting of secondary prevention.
8
Although it's beyond the scope of my introductory
9
remarks to review all the trials that have been
10
performed since then, suffice it to say that this
11
was the first of multiple cardiovascular outcomes
12
trials that have robustly demonstrated that statins
13
reduce cardiovascular events in a variety of
14
clinical use scenarios and patient populations.
15
In a meta-analysis of individual patient
16
data from 26 randomized controlled trials that
17
investigated statins, the Cholesterol Treatment
18
Trialist Collaboration reported, "The often quoted
19
relationship that each approximately 40-milligram
20
per-deciliter reduction in LDL cholesterol reduces
21
the risk for major cardiovascular events by
22
approximately 22 percent."
A Matter of Record (301) 890-4188
25
1
LDL cholesterol is often referred to as a
2
validated surrogate endpoint since statin-induced
3
reductions in LDL cholesterol have been shown time
4
and time again to reduce cardiovascular risk.
5
what is a surrogate?
6
So
FDA has defined a surrogate endpoint as a
7
marker such as a laboratory measurement or physical
8
sign that is used in therapeutic trials as a
9
substitute for a clinically meaningful endpoint
10
that is a direct measure of how a patient feels,
11
functions, or survives and is expected to predict
12
the effective therapy.
13
Note that a drug's effect on the surrogate
14
itself is of no value to the patient.
15
it may simply reflect a change in a number on a
16
laboratory report.
17
surrogate only if the effect on the surrogate does
18
lead to a clinical benefit.
19
For example,
It is considered a valid
The regulatory use of a surrogate depends on
20
the evidence that a drug's effect on the surrogate
21
predicts clinical benefit.
22
surrogate endpoint that is known to predict
A marker may be a
A Matter of Record (301) 890-4188
26
1
clinical benefit; that is a validated surrogate
2
that could be used for traditional or full
3
approval, or one that is reasonably likely to
4
predict a drug's intended clinical benefit, for
5
example, in the setting of accelerated approval.
6
I would like to also emphasize that risk
7
factor and surrogate are not synonymous.
There are
8
numerous examples of where a drug leads to
9
favorable changes in a risk factor, but those
10
changes have not translated into the expected
11
clinical benefits. Now, I have mentioned accelerated approval,
12 13
and there are invariably questions regarding this
14
regulatory pathway at these meetings where we
15
discuss the potential approval of lipid-modulating
16
drugs.
17
approval pathway for a drug that treats a serious
18
life-threatening disease or condition, generally
19
provides a meaningful advantage over available
20
therapies, and demonstrates an effect on a
21
surrogate endpoint that is reasonably likely to
22
predict clinical benefit.
Accelerated approval is a potential
And for drugs granted
A Matter of Record (301) 890-4188
27
1
accelerated approval, postmarketing confirmatory
2
trials have been required to verify and describe
3
the anticipated effect on clinical outcomes.
4
FDA has never used this regulatory pathway
5
for the approval of lipid-modulating drugs.
6
Lovastatin and pravastatin were approved before
7
this pathway even came into existence.
8
cardiovascular outcomes trials that were conducted
9
to confirm the clinical benefits of the statins
The
10
were not regulatory requirements, although the
11
agency strongly encouraged them.
12
Now, as you know, this has not been without
13
controversy.
14
two drugs for the treatment of homozygous familial
15
hypercholesterolemia, which I'll touch on in a
16
moment, the last approval of a first-in-class LDL-
17
lowering drug was Zetia in October 2002.
18
you probably recall, this approval became quite
19
controversial, especially after the publication of
20
the ENHANCE trial in 2008 and the SEAS trial six
21
months later.
22
Aside from the recent approvals of
And as
In ENHANCE, the addition of ezetimibe to a
A Matter of Record (301) 890-4188
28
1
simvastatin was not shown to reduce the progression
2
of atherosclerosis, as measured by carotid intima
3
media thickness among patients with heterozygous
4
familial hypercholesterolemia.
5
combination of simvastatin to ezetimibe was not
6
shown to reduce the risk of the primary composite
7
endpoint of cardiovascular events among patients
8
with asymptomatic aortic stenosis.
In SEAS, the
9
The SEAS trial also raised a concern that
10
active therapy was associated with cancer-related
11
events.
12
trials involving simvastatin alone, many implicated
13
ezetimibe.
And since this had not been observed in
14
Now, following a review of each of these
15
trials, FDA issued drug safety communications in
16
2009, stating that the results of the ENHANCE trial
17
did not change our position that an elevated LDL
18
cholesterol is a risk factor for cardiovascular
19
disease and that lowering LDL cholesterol reduces
20
the risk for cardiovascular disease.
21 22
Related to SEAS, FDA found it unlikely that Vytorin or Zetia increased the risk of cancer or
A Matter of Record (301) 890-4188
29
1
cancer-related death.
But despite FDA's
2
conclusions, the lack of cardiovascular outcomes
3
data for ezetimibe remained a highly controversial
4
topic. This emphasizes an important point.
5 6
Approval based on a surrogate endpoint always
7
leaves uncertainty regarding true clinical benefit,
8
which can create challenges when safety concerns
9
arise.
That is, what's the net clinical benefit in
10
the face of either known risks or the potential
11
risks that might be identified in the future after
12
millions of patients are chronically exposed to a
13
new treatment?
14
Now, I'm sure that you're aware that the
15
long-awaited IMPROVE-IT trial, which studied the
16
effect of adding ezetimibe to simvastatin on
17
cardiovascular outcomes among more than 18,000
18
patients with acute coronary syndrome, has been
19
completed.
20
the New England Journal of Medicine.
21 22
In fact, it was published last week in
According to the publication, adding ezetimibe to simvastatin led to a statistically
A Matter of Record (301) 890-4188
30
1
significant reduction in the risk of cardiovascular
2
events of a magnitude that the investigators
3
expected based on the degree of LDL lowering
4
achieved.
5
and so while I acknowledge that the results may
6
sound promising, we will be making a regulatory
7
decision regarding the application being discussed
8
today before we will have had an opportunity to
9
fully review the results from IMPROVE-IT and their
10 11
This trial has not been reviewed by FDA,
potential implications. Regardless of the results of IMPROVE-IT, I
12
do wonder what the ezetimibe controversy says about
13
the community's comfort with LDL cholesterol as a
14
valid surrogate, that is, a substitute for clinical
15
outcomes, for LDL lowering drugs other than
16
statins.
17
recent cholesterol guidelines from the ACC and AHA
18
have placed a much greater emphasis on treatments
19
that have proven benefits with regard to clinical
20
outcomes, moving away from specific biomarker
21
targets.
22
And along these lines, I note that the
Now, we recognize that this is controversial
A Matter of Record (301) 890-4188
31
1
and that not all professional societies have
2
followed suit, and we are not here to debate the
3
pros and cons of various practice guidelines.
4
only raise this because, with the statins, we now
5
have a class of drugs that not only effectively
6
lowers LDL cholesterol, but also has proven
7
benefits on the very outcomes for which we're using
8
LDL cholesterol as a surrogate.
9
I
It is worth noting that relying on
10
surrogates sometimes fails us.
11
effects could lead to net clinical harm, even if we
12
were correct about the relationship between the
13
surrogate and the clinical benefit that it was
14
expected to predict.
15
might get the causal relationship wrong entirely.
16
Off-target adverse
Alternatively, sometimes we
As just one example, despite an approximate
17
25 percent reduction in LDL cholesterol and a
18
70 percent increase in HDL, the CETP inhibitor
19
torcetrapib increased the risk of cardiovascular
20
events and increased the risk of all-cause
21
mortality in a large cardiovascular outcomes trial.
22
Although many believed that off-target effects were
A Matter of Record (301) 890-4188
32
1
responsible for this, this humbling example should
2
serve as a reminder that it sometimes takes an
3
unexpected result to make us aware of either the
4
presence or the severity of such off-target
5
effects.
6
One might ask, then, why aren't we asking
7
you to consider approval of this application being
8
discussed today under an accelerated approval
9
paradigm, where confirmation of benefit on clinical
10
outcomes would be required in the postmarketing
11
setting?
12
applicant is currently conducting a cardiovascular
13
outcomes trial.
14
After all, as I'm sure you've read, the
The answer in part is that we've never used
15
this regulatory pathway before for a
16
lipid-modulating drug, and there would be several
17
regulatory considerations that we would have to
18
work through to determine the feasibility of using
19
this approval pathway, and today is not the
20
appropriate forum for those discussions.
21 22
We have determined that it would be most helpful to us for you to focus your discussion on
A Matter of Record (301) 890-4188
33
1
whether the LDL cholesterol lowering induced by
2
alirocumab is sufficient to substitute for
3
demonstrating its effect on clinical outcomes in
4
one or more populations.
5
This is a question that we must answer when
6
considering whether to approve alirocumab using the
7
regulatory pathway that we have used for
8
LDL-lowering drugs in the past, and we'd like to
9
hear your thoughts about it.
10 11
And I want to
emphasize two points here. First, by alirocumab, we're not asking you
12
to discuss the benefits of lowering LDL cholesterol
13
by any drug that could potentially be developed.
14
Each drug may raise different issues that may
15
affect whether or not we'd consider approval based
16
on LDL cholesterol as a surrogate endpoint.
17
Second, it is critical that you note the
18
words "substitute for."
You'll hear today that the
19
applicant is conducting a cardiovascular outcomes
20
trial in the setting of acute coronary syndrome.
21
We commend them for conducting such a trial, and we
22
believe that it will provide valuable information.
A Matter of Record (301) 890-4188
34
1
However, if this drug is approved for one or
2
more patient populations on the basis of its
3
effects on LDL cholesterol, it is likely that the
4
approval would remain barring a serious safety
5
signal regardless of the results of the trial.
6
I'm sure the company will do everything in
7
their power to maintain the integrity of this
8
placebo-controlled trial.
9
the availability of the drug in the marketplace
10
leads to patients discontinuing from the trial,
11
refusing to enroll, or starting to take alirocumab
12
outside of the trial's protocol, we will have no
13
control over that.
14
But if an approval and
Even without these challenges, if the trial
15
ultimately does not demonstrate the expected
16
reduction in the risk for cardiovascular events, I
17
imagine that the applicant could argue that there
18
are unique aspects to the pathophysiology of the
19
acute coronary syndrome that should be considered
20
and that the drug still serves the needs of other
21
patients such as those with genetic disorders of
22
LDL metabolism like heterozygous familial
A Matter of Record (301) 890-4188
35
1 2
hypercholesterolemia. A few years ago, with the input of this
3
advisory committee, we determined that a reduction
4
in LDL cholesterol was a sufficient basis to
5
establish benefit for two first-in-class
6
LDL-lowering drugs for the rare, life-threatening
7
disease of homozygous familial
8
hypercholesterolemia.
9
These patients have absent or severely
10
dysfunctional LDL receptors, leading to
11
extraordinarily high LDL cholesterol and premature
12
aggressive cardiovascular disease that often
13
manifests in childhood.
14
cholesterol were especially compelling as evidence
15
of benefit for this population in which the
16
phenotype is a direct result of abnormal LDL
17
metabolism.
18
So reductions in LDL
Some of our advisors, however, emphasized
19
that the position of LDL cholesterol as a surrogate
20
may be context dependent and that the committee's
21
discussion revolving around the acceptance of LDL
22
cholesterol to establish benefit for those drugs
A Matter of Record (301) 890-4188
36
1
should not be extrapolated to other clinical
2
scenarios.
3
So to embellish on this, acceptance of using
4
LDL cholesterol as the basis for approval may
5
depend on the population and/or the mechanism of
6
action of the drug under investigation.
7
we are asking that you consider the various
8
potential uses for alirocumab specifically and
9
discuss for what patient populations, if any, you
So today,
10
are convinced that the drug's effect on LDL
11
cholesterol is sufficient by itself to establish a
12
clinical benefit.
13
Many of you have been involved in
14
discussions regarding the approval of drugs for
15
diabetes or obesity, for which postmarketing
16
cardiovascular outcomes trials have been required.
17
You might be wondering whether the paradigm could
18
apply to this application as well.
19
situations, the regulatory requirement to conduct a
20
cardiovascular outcomes trial is to characterize
21
safety, not to confirm efficacy.
22
In those
Since the passage of FDAAA, FDA can now
A Matter of Record (301) 890-4188
37
1
require postmarketing clinical trials to assess a
2
known serious risk related to the use of the drug,
3
to assess signals of a serious risk related to the
4
use of the drug, or to identify an unexpected
5
serious risk when available data indicate the
6
potential for a serious risk.
7
There are other statutory provisions that we
8
must consider before requiring postmarketing
9
clinical trials as well.
But the important
10
distinction here is that these trials that are
11
being conducted under the diabetes paradigm are
12
intended to exclude a certain degree of
13
cardiovascular risk.
14
being designed to confirm clinical benefit with
15
regard to microvascular complications of diabetes.
16
They are not, for example,
To me, it would make little sense to require
17
a safety trial to exclude increased cardiovascular
18
risk, for example accepting non-inferiority to
19
placebo by some margin, for a drug intended for use
20
solely to reduce cardiovascular risk.
21 22
Of course, it's possible that you or we may identify safety concerns with alirocumab that
A Matter of Record (301) 890-4188
38
1
warrant further evaluation in a postmarketing
2
trial, and then we would consider whether the
3
applicant's ongoing cardiovascular outcomes trial
4
might provide an acceptable platform to evaluate
5
those concerns.
6
than requiring an outcomes trial to confirm the
7
benefit of a drug on clinical outcomes after
8
approval.
9
But this is a different situation
Now, I'll turn to the discussion points that
10
we are going to ask you to discuss today, so you
11
can keep them in mind as you hear the
12
presentations.
13
The first discussion point asks you discuss
14
the safety of alirocumab as observed in the
15
clinical development program, and we've listed
16
several topics that were investigated as part of
17
the safety review for you to consider.
18
should feel free to raise any other concerns that
19
you may have as well.
But you
20
We're also asking you to discuss the
21
adequacy of the current clinical database to
22
characterize safety.
And last, as you know, we
A Matter of Record (301) 890-4188
39
1
have seen unprecedented drug-induced levels of LDL
2
cholesterol when some patients are treated with
3
PCSK9 inhibitors such as a alirocumab.
4
The applicant and our reviewers have
5
attempted to look for safety signals among patients
6
who achieve these very low levels, but these
7
analyses have their limitations.
8
we'd like you to comment on your level of concern,
9
which may even include your thoughts regarding the
Nevertheless,
10
adequacy or limitations of the current safety
11
database to answer this question.
12
The second discussion point is central to
13
the discussion of benefit.
14
lowering LDL cholesterol is sufficient to
15
substitute for a demonstration of the effect of
16
alirocumab on clinical outcomes, that is, to
17
substitute for investigation in a cardiovascular
18
outcomes trial in one or more populations.
19
The question is whether
We don't want you to feel restricted by
20
precedent here or what the applicant has proposed
21
as an indication.
22
regarding the patient populations for whom you
We'd like to hear a discussion
A Matter of Record (301) 890-4188
40
1
believe that lowering LDL cholesterol with
2
alirocumab would be convincing enough evidence of a
3
clinical benefit that you wouldn't need to see data
4
from a cardiovascular outcomes trial to confirm
5
that benefit. We understand that this will be challenging,
6 7
but that's the question that we face when making a
8
regulatory decision regarding approval.
9
that's why you're here, to provide us with some
10 11
And so
advice. Last, we have the voting question.
Has the
12
applicant sufficiently established that the LDL
13
cholesterol-lowering benefit of alirocumab exceeds
14
its risks to support approval in one or more
15
patient populations?
16
reminding you here that under the current
17
regulatory pathway, it would not be required to
18
successfully demonstrate an effect of alirocumab on
19
cardiovascular outcomes after such an approval.
20
And once again, we're
Notice that we're not asking you to vote on
21
the exact indication that the applicant has
22
proposed.
And I don't want anyone reading too much
A Matter of Record (301) 890-4188
41
1
into that as a reflection of our thinking.
2
applicant's indication contains a lot of points
3
with which you may agree or disagree.
4
didn't feel that it would be appropriate in this
5
case to make you vote all or none.
6
The
And we
Along similar lines, I cannot emphasize
7
enough that your discussion and your comments
8
following your vote are going to be absolutely
9
critical to our decision-making.
10
So with that, I'll conclude, and I'll turn
11
the podium over to the applicant.
And once again,
12
thank you all very much for your time today and
13
your dedication to the public health by joining us.
14
We really look forward to the discussion.
15
DR. SMITH:
Thank you, Dr. Smith.
16
Both the Food and Drug Administration, the
17
FDA, and the public believe in a transparent
18
process for information-gathering and
19
decision-making.
20
the advisory committee meeting, FDA believes that
21
it's important to understand the context of an
22
individual's presentation.
To ensure such transparency at
A Matter of Record (301) 890-4188
42
1
For this reason, FDA encourages all
2
participants, including the sponsor's non-employee
3
presenters, to advise the committee of any
4
financial relationships that you may have with the
5
firm at issue such as consulting fees, travel
6
expenses, honoraria, and interests in the sponsor,
7
including equity interests and those based upon the
8
outcome of the meeting.
9
you, at the beginning of your presentation, to
Likewise, FDA encourages
10
advise the committee if you do not have any such
11
financial relationships.
12
If you choose not to address this issue of
13
financial relationships at the beginning of your
14
presentation, it will not preclude you from
15
speaking.
16
So we'll now proceed with Sanofi Aventis's
17
presentations and, Dr. Edelberg, I guess you're
18
going to lead off.
19 20
Applicant Presentations – Jay Edelberg DR. EDELBERG:
21
Edelberg.
22
PCSK9 unit.
Good morning.
My name is Jay
I'm a cardiologist and head of Sanofi's We'd like to thank the FDA and this
A Matter of Record (301) 890-4188
43
1
advisory committee for your time and the review of
2
alirocumab, which was co-developed by Sanofi and
3
Regeneron Pharmaceuticals.
4
Today, I'd like to start with a review of
5
our proposed indication, the role of PCSK9 in lipid
6
metabolism, the rationale for developing a drug
7
that marks PCSK9, and then an overview of our
8
clinical program.
9
colleagues.
I'll then hand over to my
We'll review the unmet medical need in
10
lipid-lowering therapy, the clinical efficacy,
11
safety, and the overall benefit/risk assessment of
12
alirocumab.
13
Alirocumab is proposed for use as an adjunct
14
to diet for long-term treatment in adult patients
15
with primary hypercholesterolemia, non-familial and
16
heterozygous familial, or mixed dyslipidemia,
17
including in patients with type 2 diabetes.
18
Alirocumab can be used either in combination with a
19
statin or other lipid-lowering agents, or as
20
monotherapy for patients who cannot tolerate
21
statins.
22
The rationale for developing alirocumab was
A Matter of Record (301) 890-4188
44
1
based on the unmet need for additional LDL-lowering
2
in specific patient populations to help them get to
3
their guideline-directed goal and lessons learned
4
from genetics suggesting that an inhibitor of PCSK9
5
could meet this need.
6
In 2003, even before we understood its
7
mechanism of action, genetic studies showed that
8
ganglia-function mutations in PCSK9 were associated
9
with marked increases in LDL and premature
10
cardiovascular disease.
11
demonstrated that loss of function mutations in
12
PCSK9 were associated with low LDL levels and
13
reduced cardiovascular risk.
14
Later, multiple studies
The ERIC study in particular showed that
15
people with mutations causing the greatest loss of
16
function in PCSK9 had 28 percent lower LDLs and
17
88 percent fewer cardiovascular events over the
18
course of 15 years.
19
drug that blocks PCSK9 could both lower LDL and
20
reduce cardiovascular events.
21 22
These data suggested that a
Animal studies using alirocumab to block PCSK9 support the observations seen in the genetic
A Matter of Record (301) 890-4188
45
1
and epidemiologic studies.
2
experimental mouse model, we saw significant
3
atherosclerotic lesions after 8 weeks of a
4
hyperlipidemic diet.
5
In a well-established
In mice treated with alirocumab, we saw
6
almost a 90 percent lowering of the development of
7
these lesions.
8
with atorvastatin and treatment alone.
9
alirocumab was added to atorvastatin, we saw nearly
10 11
This was twice the change absorbed When
a complete absence of atherosclerosis. Let me briefly show you how PCSK9 limits LDL
12
clearance from the body to understand how
13
alirocumab inhibition to PCSK9 lowers LDL.
14
As Brown and Goldstein discovered, LDL is
15
taken up by LDL receptors, found primarily in the
16
liver, where it is internalized and cleared from
17
the circulation.
18
recycled back to the cell surface or it combined to
19
more LDL.
20
lower LDL in large part by increasing the number of
21
LDL receptors.
22
The LDL receptor can then be
Statins, which inhibit HMG co-reductase
PCSK9 binds to and is a key regulator of LDL
A Matter of Record (301) 890-4188
46
1
receptors.
2
complex, including LDL and its receptor, is
3
targeted for lysosomal degradation, which prevents
4
the recycling of the LDL receptors.
5
the number of recycled receptors, PCSK9 lowers the
6
ability of the liver to clear LDL particles, thus
7
increasing LDL levels.
8 9
When PCSK9 is bound, the entire
By reducing
Alirocumab is a fully human monoclonal antibody that binds to PCSK9 with high affinity and
10
specificity.
11
degradation of LDL receptors.
12
for the LDL receptors that bind to and are
13
internalized with LDL to be recycled.
14
effect is to lower LDL levels.
15
It prevents PCSK9-mediated This blocking allows
The net
We designed a comprehensive program to
16
evaluate whether alirocumab could meet the
17
significant unmet medical need for additional
18
LDL-lowering therapy, specifically focusing on the
19
patients with HeFH, high-risk patients without FH,
20
and patients with statin intolerance.
21
we wanted to explore dosing regimens that could
22
provide the required flexibility to more precisely
A Matter of Record (301) 890-4188
In addition,
47
1
target individual patients' LDL levels while
2
reducing the potential for unnecessarily low levels
3
of LDL.
4
Given that this therapy allows patients to
5
achieve LDL values lower than what was previously
6
possible, we also designed our program with an
7
additional objective to assess the safety of
8
treating patients to very low LDL levels with
9
studies of up to 18 months in duration.
10
With this in mind, we designed our phase 3
11
program with two distinct doses to allow therapy
12
that can be tailored to meet the treatment goals of
13
individual patients.
14
meant to be the usual starting dose to provide
15
approximately a 45- to 50-percent reduction in LDL
16
on top of the benefits of statin and other
17
lipid-lowering therapies.
18
provides maximal efficacy.
19
The 75-milligram dose is
The 150-milligram dose
The higher dose is meant to be an option for
20
up-titration for patients who need additional
21
lowering to achieve their goal as well as a
22
starting dose for patients who need greater LDL
A Matter of Record (301) 890-4188
48
1 2
reduction. The phase 3 clinical development program
3
supporting the efficacy and safety of alirocumab is
4
extensive.
5
double-blind, randomized controlled trials, which
6
were between 6 and 24 months in length, focusing on
7
the high-risk populations with unmet need for
8
additional LDL-lowering.
9
The program consisted of 10
In the program, patients with HeFH, treated
10
with maximally tolerated doses of statin and often
11
other lipid-lowering therapies, had an average LDL
12
of over 150.
13
the majority with established cardiovascular
14
disease, had a mean LDL of 110 despite maximally
15
tolerated dosing of statins.
16
High-risk non-FH patients enrolled,
The average LDL in patients with statin
17
intolerance was 191.
Almost half of these patients
18
had established cardiovascular disease and
19
15 percent had familial hypercholesterolemia.
20
Overall, the program included nearly 5300
21
patient-years in double-blind controlled studies,
22
the patients most likely to use and benefit from
A Matter of Record (301) 890-4188
49
1
alirocumab.
In addition, we have an ongoing
2
cardiovascular outcome trial, which will enroll
3
18,000 patients. All 10 phase 3 studies met the primary
4 5
endpoint.
Alirocumab can provide up to a
6
63 percent dose-dependent mean reduction in LDL
7
cholesterol in patients with high cardiovascular
8
risk who are not well-controlled with current
9
therapy, including maximally tolerated doses of
10
statin.
11
clearly demonstrates that alirocumab has a
12
favorable safety profile and is well-tolerated.
13 14
In addition, the data from this program
With this background in mind, I'd like to review the agenda for the rest of our presentation.
15
Dr. Chris Cannon will discuss the clinical
16
need for alirocumab, followed by Dr. Bill Sasiela,
17
who will present the phase 3 study design and
18
efficacy results.
19
the safety results.
20
the presentation, and I will return to answer your
21
questions.
22
Dr. Ned Braunstein will present Dr. Robert Eckel will conclude
We're also joined by additional experts.
A Matter of Record (301) 890-4188
50
1
All of our external experts have been compensated
2
for their time and travel expenses, and none have
3
an equity interest in today's outcome.
4
invite Dr. Cannon to the lectern.
5
I'll now
Applicant Presentation – Christopher Cannon DR. CANNON:
6
Thank you very much for having
7
me be a part of this proceeding.
8
Cannon.
9
Hospital in Boston.
My name is Chris
I'm a cardiologist at Brigham and Women's I've been the principal
10
investigator of the PROVE-IT and IMPROVE-IT trials
11
and have served on the steering committee for the
12
alirocumab phase 3 trials.
13
discuss our perspective as cardiologist and
14
physicians on the need for additional LDL-lowering
15
options for our patients.
16
Today, I'm here to
As we all know, cardiovascular disease is
17
the leading cause of death in the United States,
18
and as such is a top public health priority.
19
than 83 million Americans have cardiovascular
20
disease, which is responsible for 1 in 3 deaths.
21
More specifically, coronary heart disease affects
22
about 15 million Americans and accounts for 1 in
A Matter of Record (301) 890-4188
More
51
1 2
every 6 deaths each year. Now, listed here are four broad areas of
3
research that have supported LDL as a well
4
validated marker of cardiovascular risk.
5
reviewed, there's considerable evidence supporting
6
LDL as a valid marker of risk in all of these
7
areas, but I'll focus my presentation on the
8
clinical outcomes trials of both statins and
9
non-statin agents.
10
As just
This slide gives an overview, beginning with
11
4S, of many of the large secondary prevention
12
trials that have been conducted over the last
13
15 years.
14
rate plotted against the achieved LDL cholesterol
15
within the statin groups in orange and the placebo
16
groups in gray.
17
Here, we have the cardiovascular event
This figure shows that across these trials,
18
the lower and lower LDL levels seen across the
19
trials were consistently associated with lower and
20
lower event rates.
21
data supporting the concept that lowering LDL is a
22
key target in the management of coronary heart
And these have been some of the
A Matter of Record (301) 890-4188
52
1
disease. In 2010, as also mentioned, the Cholesterol
2 3
Treatment Trialist did a pooled analysis of 26
4
statin trials involving over 170,000 patients.
5
results of the larger trials with more than 1,000
6
events are plotted here, blue indicating statin
7
versus placebo trials, and those in the white,
8
high-dose versus regular-dose statin.
The
9
Now, we see here that the amount of LDL
10
lowering with statins is closely related to the
11
relative risk reduction in cardiovascular events.
12
It's also noted, specifically, about a
13
40-milligram-per-deciliter reduction in LDL led to
14
a 22 percent reduction in major cardiovascular
15
events. Now, until recently, no other
16 17
lipid-modifying therapies had been shown to reduce
18
cardiovascular risk beyond statins, as illustrated
19
here.
20
the IMPROVE-IT trial in the New England Journal of
21
Medicine.
22
of a non-statin agent, in this case ezetimibe, two
But last week, we published the results of
And this trial found that the addition
A Matter of Record (301) 890-4188
53
1
statins, produced additional LDL reduction and a
2
significant reduction in cardiovascular events.
3
In this 18,000-patient trial, those on
4
simvastatin alone achieved an LDL of 70 milligrams
5
per deciliter, and when adding ezetimibe, the mean
6
LDL was 53 milligrams per deciliter.
7
difference in LDL translated into a significant
8
6.4 percent relative risk reduction in
9
cardiovascular events.
10
And this
There are two major findings from IMPROVE-IT
11
that are worth considering with regard to the
12
alirocumab program.
13
found that the degree of cardiovascular benefit was
14
directly proportional to the degree of LDL
15
lowering.
16
on top of the meta-analysis from the CTT
17
collaboration, and you can see that it falls
18
directly on the line, relating LDL and
19
cardiovascular benefit.
20
with this non-statin agent is very consistent with
21
the degree of benefits seen with statin therapy.
22
First, the IMPROVE-IT trial
I have plotted here the IMPROVE-IT trial
And as such, the efficacy
Now, the second take-away is actually the
A Matter of Record (301) 890-4188
54
1
bad news, that despite these very intensive
2
LDL-lowering therapies, nearly a third of patients
3
had a recurring cardiovascular event over the
4
subsequent seven years.
5
we call a residual risk of cardiovascular events.
6
And this represents what
Now, beyond the patients illustrated here,
7
there are many patients who can't get to the levels
8
that we were able to achieve in either the statin
9
alone or the combination arm, and so those are
10 11
patients who need additional therapies. A final point I'd like to consider this
12
morning are the three groups of patients where we
13
see the greatest unmet need for additional
14
lipid-lowering therapy.
15
group that we've just reviewed.
16
I'll begin with the middle
Patients at high cardiovascular risk,
17
despite maximally tolerated statin therapy, are
18
generally defined as those who have had a prior MI,
19
or stroke, or other cardiovascular event, or
20
patients with multiple coronary risk factors,
21
including diabetes.
22
population was defined in the alirocumab clinical
And this is how the patient
A Matter of Record (301) 890-4188
55
1
program.
2
difficulty achieving their LDL target goals.
3
And many of these patients often have
In a recent review of the VA database, about
4
20 percent of these high-risk patients with
5
coronary disease couldn't get an LDL below
6
100 milligrams per deciliter, and 59 percent of the
7
high-risk patients couldn't get below 70.
8 9
The second group of patients are those with familial hypercholesterolemia or FH.
This is a
10
genetic disorder of lipid metabolism causing very
11
high LDL levels, which predisposed these patients
12
to premature coronary heart disease.
13
prevalence of heterozygous FH is estimated to be
14
between 1 in 200 and 250.
15
these individuals have LDL levels of between 200
16
and 400.
17
The
And when untreated,
A recent cross-sectional study found that
18
79 percent of FH patients were unable to achieve an
19
LDL goal of less than 100, despite the use of
20
high-dose statins and additional therapies like
21
ezetimibe.
22
The final population where there's an unmet
A Matter of Record (301) 890-4188
56
1
need for additional LDL-lowering therapies are
2
patients with statin intolerance.
3
estimated that between 10 and 15 percent of
4
patients who need statins show some sign of
5
intolerance, which usually manifests as muscle pain
6
or weakness.
7
patients thought to be statin intolerant can
8
actually be restarted on low doses of statins.
9
It's been
But we have seen that many of these
We as clinicians are always trying to work
10
with these patients to get them onto a statin.
11
despite our efforts, there are patients, we find,
12
who can't tolerate any dose of statin.
13
Unfortunately, nearly all of these patients cannot
14
achieve the 50 percent reduction in LDL that is
15
recommended by current guidelines.
16
But
To summarize, cardiovascular disease is the
17
leading cause of death in the United States and
18
clearly remains one of our largest public health
19
problems.
20
have been shown to reduce cardiovascular morbidity
21
and mortality.
22
remained several key groups of patients with large
Statin use and lifestyle intervention
However, as I showed, there
A Matter of Record (301) 890-4188
57
1
residual risk for cardiovascular events due to high
2
LDL levels.
3
For these patients who are unable to achieve
4
sufficient LDL lowering despite the best aggressive
5
treatment with current therapies, there is an
6
urgent need for additional lipid-lowering
7
therapies.
8
I want to thank you for your attention, and
9
I'll turn the lectern over to Dr. Bill Sasiela, who
10
will present the data on the efficacy of
11
alirocumab.
12
Applicant Presentation – Bill Sasiela
13
DR. SASIELA:
Good morning.
My name is Bill
14
Sasiela, and I'm a vice-president of program
15
direction for cardiovascular metabolism at
16
Regeneron.
17
extensive clinical development program, which
18
demonstrates that treatment with alirocumab is
19
effective in reducing LDL cholesterol with or
20
without statins or other lipid-modifying therapies.
21 22
Today, I'll review the results from our
To start, I'd like to show some representative data that shed light on the dynamic
A Matter of Record (301) 890-4188
58
1
relationship between PCSK9 inhibition and LDL
2
reduction in rationale for the dose selection and
3
design of our phase 3 studies.
4
First, let's look at what happens with a
5
single subcutaneous injection of alirocumab.
This
6
graph shows plasma alirocumab concentrations after
7
a single 150-milligram dose in healthy subjects
8
with LDL greater than 100 milligrams per deciliter.
9
As you can see, alirocumab concentration reaches
10
its maximum level within days after administration,
11
and this causes a rapid and near-immediate
12
reduction in levels of free circulating PCSK9,
13
shown by the pink line.
14
The 150-milligram dose of alirocumab blocked
15
PCSK9 sufficiently to drive free PCSK9 levels below
16
the level of detection for 10 days.
17
antibody is consumed and eliminated over time, free
18
PCSK9 levels increase to baseline levels.
19
As the
This PCSK9 blockade results in robust
20
reductions in LDL cholesterol, as depicted by the
21
light blue line.
22
PCSK9 is observed rapidly and results in a peak
The maximum reduction of free
A Matter of Record (301) 890-4188
59
1
reduction in LDL by day 14.
As would be expected,
2
the rise of free PCSK9 levels after day 10 results
3
in a corresponding increase in LDL back to
4
baseline. Now, these studies also show that background
5 6
lipid-lowering therapies could impact the
7
pharmacokinetics and pharmacodynamics of
8
alirocumab.
9
LDL-lowering drugs, and particularly statins, can
10
We and others have demonstrated that
increase the production of PCSK9. In our phase 1 and 2 studies, the impact of
11 12
this increased PCSK9 production on alirocumab's
13
efficacy was seen in two ways.
14
efficacy of alirocumab was greater when it was used
15
with other LDL-lowering drugs than when it was used
16
alone.
17
4-week interval, clearances accelerated, leading to
18
a shorter half-life and reduced efficacy at the end
19
of the dosing period.
20
First, the peak
And second, when alirocumab was dosed at a
Therefore, we selected every-2-week dosing
21
for the initial clinical development to ensure
22
consistent efficacy over the dosing interval
A Matter of Record (301) 890-4188
60
1
regardless of background lipid-lowering therapies. We chose two doses with distinct efficacy to
2 3
give physicians flexibility in dosing and treatment
4
based on phase 2 data and modeling.
5
PK/PD modeling to identify 75 milligrams every
6
2 weeks as a dose that would provide 45 to
7
50 percent reductions in LDL.
8
150 milligrams every 2 weeks to provide maximal LDL
9
efficacy.
We used the
We then selected
The availability of two doses will allow
10 11
prescribers to select a starting dose based on the
12
degree of LDL lowering needed.
13
75-milligram dose to be the usual starting dose,
14
but prescribers can choose to start with
15
150 milligrams if a patient needs larger reductions
16
in LDL.
17
the patient's response to treatment, will be
18
possible if needed.
19
We intend the
Additionally, dose adjustments, based on
From here, I summarize the design of our
20
phase 3 program.
Our phase 3 program enrolled
21
patients with hypercholesterolemia who were at high
22
or very high cardiovascular risk based on clinical
A Matter of Record (301) 890-4188
61
1
guidelines at the time the studies were designed.
2
The objective of the studies was to reach
3
patient-specific LDL goals based on baseline
4
cardiovascular risk.
5
the goal was LDL cholesterol less than
6
100 milligrams per deciliter.
7
high risk, the goal was less than 70 milligrams per
8
deciliter.
9
For patients at high risk,
For patients at very
The primary efficacy endpoint of all of our
10
phase 3 studies was the percent change in LDL
11
cholesterol at week 24.
12
conducted on an intention-to-treat basis, which
13
analyzed all lipid values, regardless of adherence
14
to study treatment.
15
be presenting today.
16
Our primary analysis was
The ITT analyses are what I'll
We had several secondary endpoints, which
17
were ranked to account for multiplicity.
18
include an analyses of efficacy at week 12 prior to
19
potential up-titration and at week 52.
20
examined the proportion of patients who achieve
21
various LDL treatment goals, as well as changes in
22
other lipid parameters.
A Matter of Record (301) 890-4188
These
We also
62
1
Our phase 3 patients represented three core
2
populations with unmet need for additional LDL
3
cholesterol lowering:
4
heterozygous familial hypercholesterolemia, high
5
cardiovascular risk patients, and patients not on
6
statins, including patients who are statin
7
intolerant.
8 9
patients on statins with
Five studies compared alirocumab to placebo, shown in green and blue.
Five studies compared
10
alirocumab to ezetimibe, shown in gray and yellow.
11
In gray are the two ezetimibe controlled studies
12
that enrolled patients not receiving background
13
statin therapy, including the ALTERNATIVE study,
14
which enrolled patients with documented statin
15
intolerance.
16
The high FH in long-term studies, shown in
17
blue, used the 150-milligram dosing regimen.
18
other eight studies evaluate 75 milligrams with the
19
option of up-titration to 150 milligrams if the
20
patient's LDL goal had not been achieved.
21 22
The
On this slide, I'll provide a profile to patients enrolled in the phase 3 program that
A Matter of Record (301) 890-4188
63
1
represents the three core patient populations and
2
demonstrates a significant unmet medical need for
3
additional lipid lowering in these populations.
4
We had over 1300 heterozygous familial
5
hypercholesterolemic patients in our program, and
6
despite nearly 80 percent of them taking
7
high-intensity statins and more than 60 percent
8
receiving additional therapies beyond the statin,
9
their mean LDL was over 150 milligrams per
10
deciliter.
11
had prior cardiovascular disease in their lifetime.
12
Nearly half of these patients already
There were over 3500 patients at high or
13
very high risk for cardiovascular disease.
Over
14
80 percent of these patients experienced prior
15
cardiovascular disease and 40 percent had diabetes
16
as well.
17
tolerated statins and nearly a quarter were also
18
getting additional lipid therapies.
19
baseline LDL cholesterol levels were 110 milligrams
20
per deciliter.
These patients were receiving maximally
And yet, their
21
Finally, among the statin-intolerant
22
patients, nearly half of these patients had already
A Matter of Record (301) 890-4188
64
1
experienced a prior cardiovascular disease, and
2
over 40 percent were getting treated with non-
3
statin lipid therapies, yet their baseline LDL
4
levels were the highest in our program at just over
5
190 milligrams per deciliter.
6
some of these patients also had familial
7
hypercholesterolemia.
8 9
Not surprisingly,
This slide summarizes the enrollment and duration of the phase 3 studies.
The five
10
placebo-controlled studies randomized nearly 3500
11
total patients and were 12 or 18 months in
12
duration.
13
I want to point out that the LT study, which
14
we call the long-term study in the briefing book
15
and in our New England Journal of Medicine
16
publication, is just one of four 18-month trials
17
that were placebo-controlled.
18
controlled trials enrolled nearly 1800 patients.
19
These trials were 6 to 24 months in length.
20
The five ezetimibe
Let me describe the study designs of these
21
two dosing regimens in more detail.
22
with the two studies that use continuous
A Matter of Record (301) 890-4188
I'll start
65
1
150-milligram dosing, LT, and high FH.
2
patients in these studies were treated with diet
3
and a statin at maximum-tolerated dose, with or
4
without other lipid-modifying therapies.
5
were randomized 2 to 1 to receive subcutaneous
6
injections of either alirocumab, 150 milligrams, or
7
placebo every other week for 78 weeks.
Patients
The primary endpoint was assessed at
8 9
All
week 24.
After the end of the double-blind period,
10
patients were followed for an additional 8 weeks or
11
eligible to enroll in an open-label extension
12
study.
13
In the eight studies that used the
14
up-titration regimen, patients were randomized to
15
receive alirocumab, 75 milligrams, or controlled.
16
All patients received placebo injections for
17
alirocumab, and in the ezetimibe-controlled
18
studies, alirocumab patients received placebo for
19
ezetimibe to maintain the blind.
20
The double-blind periods ranged from
21
24 weeks to 2 years.
At week 12, patients in the
22
alirocumab group were blindly up-titrated from 75
A Matter of Record (301) 890-4188
66
1
to 150 milligrams if they had not achieved their
2
individual LDL goal at the week 8 visit.
3
prior trial designs, we assessed the primary
4
endpoint at week 24.
As in the
I'd also like to take you through the design
5 6
of the ALTERNATIVE study, which enrolled patients
7
with a history of statin intolerance.
8
trial, statin intolerance was defined as having
9
failed at least two statins due to side effects,
For this
10
one of which must have been at the lowest approved
11
dose.
This is consistent with NLA guidelines. Any patients who exhibit a skeletal muscle
12 13
adverse event during a 4-week placebo run-in period
14
were excluded from the double-blind treatment
15
period.
16
treatments:
17
up-titration, ezetimibe, or statin rechallenge for
18
atorvastatin, 20 milligrams.
19
objective compared alirocumab to ezetimibe at
20
week 24.
21 22
Patients were randomized to 1 of 3 alirocumab with potential
The primary efficacy
All the 10 phase 3 studies met their primary efficacy objectives.
The first two studies I'll
A Matter of Record (301) 890-4188
67
1
discuss are the high FH NLT studies, which compared
2
alirocumab, 150 milligrams, every other week, to
3
placebo.
4
tolerated dose of a statin.
5
All patients also receive maximum
Alirocumab met the primary efficacy
6
objective in both the LT and high FH studies.
7
you can see, the reductions in LDL cholesterol were
8
significantly greater with alirocumab than placebo.
9
As
In a pooled analysis of these studies,
10
maximum efficacy was reached by 4 weeks and was
11
sustained for one year.
12
60 percent LDL reduction from baseline at week 24
13
and a 56 percent reduction from baseline at one
14
year, with a similar 60 percent difference compared
15
to placebo at this time point.
16
Alirocumab achieved a mean
The next three studies I'll review are FH I,
17
FH II, and COMBO I.
These trials studied the
18
up-titration dosing, with all patients on
19
maximum-tolerated dose of a statin and many with
20
additional lipid-modifying therapies.
21
FH I, all patients had heterozygous FH.
22
COMBO I, all patients were at high cardiovascular
A Matter of Record (301) 890-4188
In FH I and And in
68
1
risk. Alirocumab met the primary efficacy
2 3
objective in all of these studies as well.
4
average, patients treated with alirocumab achieved
5
LDL reductions of 48 to 49 percent at week 24.
6
This dosing regimen also produced consistent and
7
sustained efficacy through one year. Three ezetimibe-controlled studies in our
8 9
On
program evaluated the up-titration regimen with
10
patients on background statin therapy.
COMBO II
11
enrolled patients at high cardiovascular risk on
12
maximum-tolerated dose of a statin.
13
II enrolled patients who were not at goal and
14
monitored doses of either atorvastatin or
15
rosuvastatin at study entry.
OPTIONS I and
Alirocumab was superior to all comparators
16 17
in these three studies with the exception of the
18
rosuvastatin 20-milligram arms, where the
19
multiplicity-adjusted level for significance was
20
0.01.
21
superiority over ezetimibe in at least one
22
treatment comparison, so all studies met the
Overall, each study showed alirocumab
A Matter of Record (301) 890-4188
69
1 2
primary efficacy objective. Finally, two ezetimibe-controlled studies
3
examined the efficacy of alirocumab among patients
4
who are not on background statin therapy.
5
studies used the up-titration regimen.
6
was superior to ezetimibe in both the ALTERNATIVE
7
study among statin-intolerant patients and in the
8
MONO study, where alirocumab was used as
9
monotherapy.
10
Those
Alirocumab
Eight of the 10 phase 3 studies we're
11
discussing today used up-titration from
12
75 milligrams to 150 milligrams every 2 weeks.
13
evaluated the impact of up-titration on efficacy
14
and examined the two doses at week 12 prior to
15
up-titration.
16
We
The FH I study illustrates the effect of
17
up-titrating alirocumab in patients receiving
18
background statin.
19
of patients who were up-titrated.
20
shows those who remained on 75 milligrams
21
throughout the study.
22
patients who up-titrated tend to have higher
The red line depicts a subgroup The green line
As we would expect, the
A Matter of Record (301) 890-4188
70
1
baseline LDL levels and a lesser response to
2
alirocumab at the 75-milligram dose.
3
In this study, patients who up-titrated saw
4
on average a 17-percentage-point additional
5
reduction in LDL.
6
75 milligrams maintained a stable level of efficacy
7
through one year.
8 9
The patients who remained on the
As previously mentioned, we conducted the ALTERNATIVE study in statin-intolerant patients,
10
and, therefore, these patients were using no
11
background statin, no ezetimibe.
12
study that I just showed you, the titration
13
criteria was selected for patients both with higher
14
baseline LDL levels as well as a lesser response to
15
alirocumab.
16
As in the FH I
Titration in these patients provided an
17
additional 6 percent mean reduction in LDL, and 25
18
percent of these patients achieved at least a
19
10 percent reduction in LDL cholesterol.
20
As seen before, those that remained on
21
75 milligrams maintained efficacy over the 24-week
22
treatment period.
Overall, data from the eight
A Matter of Record (301) 890-4188
71
1
titration studies reinforced the utility of
2
up-titration regardless of background statin usage.
3
Because the trials of the up-titration
4
dosing regimen and the 150-milligram regimen
5
enrolled similar patients, we can compare the
6
relative efficacy of the two doses by assessing
7
efficacy at week 12 prior to up-titration.
8
the studies I reviewed, the mean LDL reduction at
9
week 12 was 45 to 49 percent with the 75-milligram
10
Across
dose and 63 percent with the 150-milligram dose.
11
Based on these results, we will recommend
12
75 milligrams as the usual starting dose and that
13
physicians should select 150 milligrams as the
14
starting dose for patients needing a 60 percent or
15
greater reduction in LDL cholesterol.
16
flexibility will allow physicians to achieve the
17
desired reductions in LDL while minimizing the
18
propensity for very low LDL levels.
19
This dosing
I'll now briefly review the findings from
20
additional secondary endpoints, which showed
21
favorable effects of alirocumab on other lipids and
22
apolipoproteins.
Your briefing document provides
A Matter of Record (301) 890-4188
72
1 2
details of these analyses. Both non-HDL cholesterol and apo B represent
3
a broader spectrum of atherogenic lipoproteins than
4
LDL and are sometimes used as a secondary treatment
5
goal in lipid guidelines.
6
studies, we observed mean reductions of these
7
parameters, ranging from 32 to 52 percent, with 75-
8
and 150-milligram doses.
9
Across our phase 3
Similarly, we observed mean reductions of 25
10
to 38 percent in total cholesterol with the two
11
doses of alirocumab.
12
Lipoprotein(a) is a similar but distinct
13
atherogenic lipoprotein from LDL.
14
statin therapy has no effect on Lp(a).
15
contrast, we observed mean reductions in the range
16
of 17 to 30 percent with alirocumab.
17
therapy also induced favorable changes in fasting
18
triglycerides, HDL, and apo A1.
19
Interestingly, In
Alirocumab
In summary, the clinical program
20
demonstrated that alirocumab is effective in
21
reducing LDL cholesterol with or without statins or
22
other lipid-modifying therapies across all the
A Matter of Record (301) 890-4188
73
1
patient populations studied. Today, I've summarized results from 10
2 3
phase 3 clinical trials, which studied over 5,000
4
patients.
5
followed in double-blind studies for at least one
6
year.
7
Nearly 80 percent of the patients were
All 10 studies met their primary LDL
8
cholesterol endpoints.
In the studies that used
9
up-titration, most patients reached their LDL
10
treatment goal on the lower 75-milligram dose.
11
Starting with up-titration to 150 milligrams
12
provided additional efficacy for patients who
13
required additional LDL lowering.
14
robust reductions in LDL were absorbed across
15
patient subgroups and background lipid treatments.
16
Significant and
Alirocumab also had a positive overall
17
impact across a spectrum of lipid parameters,
18
suggesting that alirocumab has a potentially strong
19
anti-atherosclerotic profile.
20
I'll now turn the presentation over to
21
Dr. Ned Braunstein to review the clinical safety of
22
alirocumab.
A Matter of Record (301) 890-4188
74
1
Applicant Presentation – Ned Braunstein
2
DR. BRAUNSTEIN:
Thank you, Bill.
3
I'm Ned Braunstein, senior vice-president at
4
Regeneron, and I'll be presenting our safety data.
5
Our strategy was to study alirocumab in long-term,
6
double-blind controlled studies in the patients
7
most likely to use and benefit from the product.
8
My presentation is outlined here, and I'll start by
9
discussing our prespecified safety analysis plan.
10
We combined data from the phase 2 and
11
phase 3 studies into three pools:
12
placebo-controlled pool, an ezetimibe-controlled
13
pool, and a global pool.
14
pool evaluates alirocumab or placebo on top of
15
statin and, in the phase 3 studies, on top of a
16
maximally tolerated dose of statin.
17
a
The placebo-controlled
Four phase 3 placebo-controlled studies were
18
18 months long.
One was 12 months.
19
3,000 continuing patients have completed their
20
one-year visit at the time of data cutoff for the
21
application, and over 800 have completed 18 months.
22
We considered these placebo-controlled data to be
A Matter of Record (301) 890-4188
More than
75
1
primary in our assessment of safety.
The ezetimibe
2
pool evaluated five studies with differing duration
3
and statin use.
We consider these data supportive.
Now, the placebo- and ezetimibe-controlled
4 5
pools are separate and non-overlapping, so finally,
6
we combined these two pools into a global pool with
7
all the alirocumab data in one treatment group and
8
all the control data in a second group. Overall, we pooled over 5200 patient-years
9 10
of double-blind safety data in the phase 2 and
11
phase 3 studies.
12
smaller than those shown in the efficacy section
13
because the safety pools don't include the statin
14
comparator arms from the OPTIONS or ALTERNATIVE
15
studies.
Now, these numbers are a little
We evaluated adverse events based on two
16 17
tiers.
18
no prespecified hypotheses were screened using the
19
Cox model, and then we explored in greater detail
20
those whose 95 percent confidence interval of the
21
hazard ratio excluded 1.
22
Common adverse events for which there were
Adverse events of special interest were
A Matter of Record (301) 890-4188
76
1
those for which there were prespecified hypotheses
2
based on the safety profile of other drugs that
3
lower LDL or theoretical concerns related either to
4
low LDL or to the administration of monoclonal
5
antibodies.
6
event terms to group and analyze together using a
7
variety of statistical techniques.
8 9
For these, we prespecified adverse
Now, as discussed both in the FDA's and our briefing books, sensitivity analyses of the
10
individual doses showed no dose-related trends, so
11
I will focus on analyses in which the 75- and the
12
150-milligram doses are combined.
13
an overview of the adverse event data.
14
I'll start with
I'll be showing a lot of slides like this,
15
so let me orient you.
The placebo-controlled pool
16
is shown in the middle columns, and the ezetimibe
17
pool is shown in the columns to the right.
18
because of the approximately 2 to 1 randomization
19
ratio in the placebo-controlled pool, it is
20
important to focus on the incidence rates, as shown
21
here, and not the numbers of patients with events.
22
In both the placebo- and ezetimibe-
A Matter of Record (301) 890-4188
Now,
77
1
controlled pools, the overall incidence rates of
2
treatment-emergent adverse events, serious adverse
3
events, deaths, and adverse events leading to
4
discontinuation were similar between the alirocumab
5
and controlled treatment groups.
6
these findings in more detail in the next several
7
slides.
I'll describe
8
We use the Cox model to screen for adverse
9
events that might be higher in one treatment group
10
or the other.
11
terms occurring in at least 1 percent of patients
12
were more common in the placebo group and 4 were
13
more common in the alirocumab group, so clearly the
14
screening approach was sensitive but not specific.
15
In the placebo-controlled pool, 7
Among the adverse events more common in the
16
alirocumab group, injection site reaction and
17
pruritus consistently occurred more often in
18
patients taking alirocumab than controlled.
19
Now, with respect to serious adverse events,
20
the incidence was similar with alirocumab or
21
control.
22
examined by organ system, as shown here.
There were no meaningful differences when
A Matter of Record (301) 890-4188
I'll get
78
1
to some of the details when I present the adverse
2
events of special interest.
3
Deaths were also balanced across the
4
treatment groups.
All deaths were adjudicated as
5
to their cause by an independent adjudication
6
committee that was blinded to treatment assignment.
7
Twenty-six of 37 were due to cardiovascular disease
8
and of these, 21 were due to coronary heart
9
disease, consistent with our having enrolled a
10
patient population at high or very high
11
cardiovascular risk.
12
The percentage of patients who discontinued
13
due to any adverse event was comparable between
14
alirocumab and control groups.
15
events were the most frequent overall, and these
16
occurred mainly in the ezetimibe-controlled pool
17
and specifically in the ALTERNATIVE study,
18
conducted in patients with a history of statin
19
intolerance.
20
Musculoskeletal
I will turn now to our analyses of adverse
21
events of interest.
The briefing book discusses
22
many topics shown here.
In my presentation, I will
A Matter of Record (301) 890-4188
79
1
summarize a subset of these that we consider of
2
particular interest to this committee, and I'll
3
begin with injection site reactions. Local injection site reactions are commonly
4 5
observed in patients receiving injectable biologic
6
agents.
7
have reported local injection site reactions were
8
6 percent in the alirocumab groups and 4 percent in
9
the pooled control groups.
Overall, the percentages of patients who
The majority were mild
10
and transient.
11
none was reported as a serious adverse event.
12
Injection site reactions only rarely led to
13
discontinuation of treatment.
14
Only 1 was reported as severe, and
Now, patients who developed treatment-
15
emergent anti-drug antibodies were more likely to
16
have a local injection site reaction compared to
17
those who tested negative.
18
risk factors for local injection site reactions.
19
We identified no other
Turning to general allergic events, these
20
occurred at similar rates in patients treated with
21
alirocumab compared to placebo with an exposure-
22
adjusted hazard ratio of 1.1.
A Matter of Record (301) 890-4188
We observed a
80
1
similar trend in the ezetimibe-controlled pool.
2
The small difference between treatment groups was
3
mostly due to pruritus.
4
occurred in no more than 0.4 percent of patients in
5
any group.
6
anti-drug antibodies did not have an increased
7
incidence of allergic events.
Serious allergic events
And patients with treatment-emergent
8
Now, although discontinuations due to a
9
general allergic adverse event were infrequent,
10
they did occur in a higher percentage of
11
alirocumab-treated patients compared to control; in
12
the alirocumab groups, allergic events leading to
13
discontinuation, including flushing,
14
hypersensitivity, hypersensitivity vasculitis,
15
nummular eczema, pruritus, and rash.
16
though, resolved after stopping treatment.
These all,
17
Now, the patient with angioedema, that
18
patient had a history of recurrent angioedema.
19
other terms such as contact dermatitis and
20
interstitial lung disease had other identifiable
21
causes.
22
Although not described in their clinical
A Matter of Record (301) 890-4188
And
81
1
trials data, statin-labeling notes that
2
neurocognitive events have been reported in
3
postmarketing use.
4
is not dependent on LDL because it synthesizes all
5
the cholesterol it needs.
6
of statins may be through some unique mechanism
7
rather than through LDL lowering, per se.
So this putative effect
With regard to alirocumab, monoclonal
8 9
Now, the central nervous system
antibodies do not typically cross the blood-brain
10
barrier.
So a priori, one would not anticipate an
11
effect of alirocumab on neurocognitive function.
12
Nonetheless, given this effective statin,
13
neurocognitive events were considered an adverse
14
event of special interest in our program. We prespecified two ways to look at
15 16
neurocognitive events.
17
of adverse event terms that we refer to as the
18
sponsor's MedDRA query.
19
focused set of terms that was proposed to us by the
20
FDA.
21 22
First, we used a broad set
The second way used a more
I'll focus first on the LT study, so looking first at the sponsor's MedDRA query, there were
A Matter of Record (301) 890-4188
82
1
more neurocognitive events in the alirocumab group
2
compared to placebo.
3
that produced a lower-risk estimate.
4
hazard ratios for both analyses have 95 percent
5
confidence intervals, so that's span one, so we
6
can't draw definitive conclusions.
7
look at all five 12- to 18-month placebo-controlled
8
studies or at the ezetimibe-controlled studies, we
9
see hazard ratios of around 1 for both MedDRA
10
In the query produced by FDA, However, the
And when we
queries. Importantly, no patient treated with
11 12
alirocumab in the entire phase 2 or 3 program
13
discontinued due to a neurocognitive event.
14
overall, the incidence of neurocognitive events
15
with alirocumab groups is similar to control in our
16
pools.
17
of a class effect.
18
entire phase 2/3 double-blind program, these data
19
are not definitive, especially considering that we
20
are trying to assess an effect of a drug over the
21
background incidence of events.
22
So
But we recognize that there is a question And with only 44 events in the
So our pharmacovigilance plan for these is
A Matter of Record (301) 890-4188
83
1
thorough.
We have designated neurocognitive events
2
and adverse events of special interest in our
3
ongoing studies and have enlisted independent
4
experts to assist in their evaluation.
5
50,000 patient-years of exposure expected in our
6
outcomes trial, we will have enough data to exclude
7
a 30 percent difference assuming an event rate of
8
1 percent.
With over
9
Because of the well-known association of
10
statins with musculoskeletal events, we included
11
this category in our analyses of adverse events of
12
interest.
13
terms representing the musculoskeletal events most
14
commonly reported by statin-intolerant patients.
15
In the placebo-controlled pool, the incidence of
16
musculoskeletal events was comparable in the
17
alirocumab and placebo groups.
18
that these were patients who were taking and known
19
to tolerate statins.
We pre-specified a grouping of MedDRA
Remember, though,
20
In the ezetimibe-controlled studies, the
21
majority of musculoskeletal events were from the
22
ALTERNATIVE study, which enrolled patients with
A Matter of Record (301) 890-4188
84
1
documented statin intolerance.
2
those data shortly.
3
that excluded the ALTERNATIVE study data, and we
4
can see that the incidence of musculoskeletal
5
events was comparable in the alirocumab and
6
ezetimibe groups.
7
I will show you
On this slide are analyses
The ALTERNATIVE study data are shown here.
8
Patients in the ALTERNATIVE study were randomized
9
to alirocumab, ezetimibe, or to atorvastatin,
10
20 milligrams.
11
rechallenge arm, we had to exclude patients with a
12
history of severe musculoskeletal reactions.
13
Nonetheless, fewer patients in the alirocumab
14
group, shown in blue, reported musculoskeletal
15
adverse events than in the atorvastatin group.
16
similar trend was seen versus ezetimibe.
17
Now, because of the atorvastatin
A
The hazard ratio in comparison to
18
atorvastatin was 0.6 and 0.7 in comparison to
19
ezetimibe.
20
who discontinued therapy for skeletal muscle-
21
related adverse events.
22
associated with fewer musculoskeletal complaints
Similar trends were seen for patients
Overall, alirocumab was
A Matter of Record (301) 890-4188
85
1
than atorvastatin. Increases in serum transaminases have been
2 3
reported with statins, and we looked at hepatic
4
effects in two ways, based on investigator's
5
assessment of adverse events and using laboratory
6
data.
7
investigators were similar between treatment
8
groups.
9
to changes in hepatic enzymes.
10
Overall, hepatic adverse events reported by
Most hepatic adverse events were related
These occurred at a slightly higher
11
incidence in the alirocumab group compared to the
12
placebo group and a slightly lower incidence in the
13
alirocumab group compared to the ezetimibe group.
14
Now, potential hepatotoxicity may be
15
signaled by a set of findings called Hy's law.
16
Hy's law criteria are a three-fold elevation above
17
upper limit of normal in transaminases and a
18
two-fold elevation above upper limit of normal in
19
bilirubin in the absence of cholestasis or some
20
other identifiable cause.
21 22
So there were three patients with these elevations in transaminase and bilirubin levels in
A Matter of Record (301) 890-4188
86
1
the program, but none of them met Hy's law criteria
2
because of alternative etiologies, as shown here.
3
All of these events were reported as serious
4
adverse events, and all the patients recovered.
5
Statin labeling also notes the potential for
6
increased blood glucose leading to an increased
7
risk of diabetes.
8
we looked at this in detail by evaluating shifts in
9
glycemic control, the new use of anti-diabetic
As shown in our briefing book,
10
medications, adverse events of diabetes, and mean
11
changes in hemoglobin A1c or fasting blood glucose.
12
I will discuss one of these today, the incidence of
13
new onset diabetes.
14
In this prespecified analysis, we looked at
15
the subgroup of patients not diabetic at baseline
16
and determined the incidence of nuance at diabetes
17
or clinically meaningful changes in blood glucose,
18
as determined by the investigator.
19
in the alirocumab groups is similar to that in each
20
of the control groups.
21
multiple analyses that we performed indicate that
22
alirocumab does not meaningfully impact glycemic
The incidence
Overall, the data from
A Matter of Record (301) 890-4188
87
1 2
control. To assess the cardiovascular safety of
3
alirocumab, an independent cardiovascular events
4
adjudication committee reviewed all cardiovascular
5
events and deaths in the phase 3 clinical program.
6
This slide shows the possible results of
7
adjudication.
8
events consists of several hard outcomes shown on
9
this slide.
10
MACE or major adverse cardiovascular
Now, since there were only three patients
11
with unstable angina in our program, almost all of
12
the MACE events in our program meet the strict MACE
13
definition as defined previously by this committee.
14
MACE is the primary endpoint in our ongoing
15
cardiovascular study, and in my presentation today,
16
I will focus on MACE.
17
Events could also be adjudicated as
18
congestive heart failure, coronary
19
revascularization procedures, hemorrhagic strokes,
20
TIAs, or cardiovascular deaths.
21
number of events that we have so far are low, so
22
the data should be interpreted cautiously.
A Matter of Record (301) 890-4188
Overall, the
88
1
As initially submitted in the application,
2
85 MACE events were confirmed in the global pool of
3
phase 3 studies.
4
upper bound of the 95 percent confidence interval
5
was 1.25.
6
most of these events were nonfatal myocardial
7
infarctions.
The hazard ratio was 0.81.
The
Typical for this patient population,
8
Other cardiovascular events included
9
congestive heart failure, requiring hospitalization
10
and coronary revascularization procedures.
11
latter was reported with a somewhat higher
12
frequency in the alirocumab group than in the
13
comparator group, so including these events, the
14
hazard ratio was 1.08 and the upper bound of the
15
95 percent confidence interval was 1.5.
16
The
Most of the adjudicated events were in the
17
LT study, which was the largest of the 78-week
18
placebo-controlled phase 3 studies.
19
enrolled a mix of FH and non-FH patients.
20
baseline LDL cholesterol in this study was
21
approximately 120 milligrams per deciliter.
22
Forty-eight percent of patients had a history of
A Matter of Record (301) 890-4188
The LT study The mean
89
1
myocardial infarction or stroke, and about
2
77 percent had existing cardiovascular disease.
3
Twenty percent of patients had diabetes plus at
4
least two additional risk factors.
5
These are the types of patients described by
6
Dr. Cannon who are at risk for major adverse
7
cardiovascular events, have elevated LDL
8
cholesterol despite maximally tolerated statins,
9
and who therefore have an unmet need for additional
10 11
LDL cholesterol lowering. In prespecified summaries of the LT study
12
data over the entire study period, both adjudicated
13
MACE and a broader cardiovascular endpoint, that
14
included congestive heart failure and
15
revascularizations, had a lower incidence in the
16
alirocumab plus statin group compared to the
17
placebo plus statin group.
18
A post hoc analysis investigated these
19
endpoints using the Cox model.
20
for MACE based on the interim data in the
21
application was 0.46, and in the final data that we
22
reported in the New England Journal of Medicine,
A Matter of Record (301) 890-4188
The hazard ratio
90
1
the hazard ratio was 0.52. But the Kaplan-Meier plot illustrates
2 3
another important point, that approximately 2 and
4
half percent of patients in the statin plus placebo
5
group had hard endpoint MACE events over one year,
6
consistent with the unmet need of these patients
7
with elevated LDL cholesterol and cardiovascular
8
risk factors. So our phase 3 cardiovascular event data are
9 10
reassuring, and the positive trends support the
11
cardiovascular safety of alirocumab in this
12
population of patients, but they are insufficient
13
to draw conclusions about cardiovascular benefit. To answer this important question, we have
14 15
an ongoing placebo-controlled, cardiovascular
16
outcomes trial.
17
will enroll approximately 18,000 high-risk patients
18
with recent acute coronary syndrome who are being
19
treated with a maximally tolerated dose of a potent
20
statin.
21
prespecified number of events has accrued and the
22
last patient enrolled has been followed for two
The study began in 2012, and it
The study will continue until a
A Matter of Record (301) 890-4188
91
1 2
years. The primary endpoint is adjudicated major
3
adverse cardiovascular events or MACE, and the
4
primary analysis is intention to treat.
5
will have adequate statistical power to evaluate
6
the potential benefit of alirocumab, to reduce the
7
incidence of MACE, and it will also provide a
8
greater-than-50,000 patient-year safety database
9
for ongoing pharmacovigilance.
10
This study
Finally, we evaluated the safety of LDL
11
values less than 25.
12
LDL cholesterol values is limited, so our
13
controlled studies were designed specifically to
14
evaluate the safety of alirocumab in patients who
15
achieved LDL values below 25.
16
Now, our experience with low
Now, to ensure we had an adequate number of
17
patients who achieved these values, we initiated
18
alirocumab at the highest 150-milligram dose every
19
two weeks in the 2400-patient LT study.
20
patients were not allowed to down-titrate in that
21
study or in any of our studies.
22
24 percent of patients on alirocumab had two or
A Matter of Record (301) 890-4188
Moreover,
So as a result,
92
1
more consecutive LDL cholesterol values less than
2
25 milligrams per deciliter, and 9 percent had two
3
or more values less than 15, providing a reasonable
4
subgroup to study.
5
The overall rates of treatment-emergent
6
adverse events, serious adverse event events,
7
deaths, or adverse events leading to treatment
8
discontinuation were similar in patients with low
9
LDL cholesterol values compared to the overall
10
population.
11
adverse events are in your briefing book.
12
Detailed analyses of individual
It is important, though, to recognize that
13
the comparison that we did and that's shown here is
14
performed in post-randomization subgroups, and,
15
therefore, there are differences in baseline
16
characteristics in patients who did or did not
17
experience low LDL cholesterol in our studies.
18
So while no adverse effects were identified,
19
the safety data on low LDL cholesterol from
20
double-blind trials are limited beyond a year.
21
considering that alirocumab is the first in a new
22
class of drugs, because the additional benefit of
A Matter of Record (301) 890-4188
So
93
1
such low LDL cholesterol values is unknown, and
2
because we don't want patients to stop or lower
3
their statin dose, we recommend a dosing strategy
4
designed to achieve patients' LDL goals while
5
avoiding these very low levels of LDL.
6
In our studies, the strongest predictors of
7
achieving two consecutive LDL cholesterol values
8
less than 25 milligrams per deciliter were baseline
9
LDL and starting dose.
Patients who started
10
therapy on the 75-milligram dose of alirocumab are
11
shown in the orange bars.
12
therapy on the 150-milligram dose is shown on the
13
blue bars.
14
Patients who started
Baseline LDL cholesterol is shown on the
15
X-axis, and the percent of patients with LDL
16
cholesterol less than 25 milligrams on two or more
17
consecutive occasions is shown on the Y-axis.
18
So for example, in patients with LDL
19
cholesterol below 150 at baseline, treating with
20
the 150-milligram dose resulted in LDL cholesterol
21
values less than 25 in as many as 20 to 70 percent
22
of patients, whereas starting therapy with the
A Matter of Record (301) 890-4188
94
1
lower 75-milligram dose resulted in these low
2
values much less often.
3
to limit patients achieving these low LDL
4
cholesterol values is to start most patients with a
5
75-milligram dose of alirocumab and limit the
6
150-milligram starting dose to the minority of
7
patients who need a greater than 60 percent
8
reduction from baseline to achieve a desirable
9
LDL-C level.
Therefore, a practical way
We believe this approach maximizes the known
10 11
benefit/risk of the product for all the patient
12
groups that we have studied while minimizing the
13
unknown risk that might be associated with
14
long-term very low levels of LDL. So to summarize the safety data, alirocumab
15 16
was well tolerated with a favorable safety profile
17
in over 5200 patient-years of data from
18
double-blind studies conducted mostly in high-risk
19
patients on a background of maximally tolerated
20
statin.
21 22
The only imbalances were injection site reactions and pruritus.
Some rare and sometimes
A Matter of Record (301) 890-4188
95
1
serious allergic events did lead to treatment
2
discontinuation.
3
neurocognitive adverse events compared to control
4
and no effect on glycemic control.
There was no increase in
5
In contrast to statins, we saw no increase
6
in musculoskeletal events with alirocumab compared
7
to placebo, and we actually saw fewer
8
musculoskeletal events compared to atorvastatin in
9
patients with a history of statin intolerance.
10
there were no apparent safety concerns in our
11
studies with LDL values less than 25.
12
And
Finally, the phase 3 data support the
13
cardiovascular safety of alirocumab and the ongoing
14
outcomes trial will evaluate cardiovascular benefit
15
and provide a large database for continuing
16
pharmacovigilance.
17
to Dr. Eckel, who will conclude the presentation.
I will now turn the podium over
18
Applicant Presentation – Robert Eckel
19
DR. ECKEL:
20
I'm Bob Eckel.
Thanks very much, Ned. I'm a professor of medicine
21
at the University of Colorado and director of the
22
Lipid Clinic at the University of Colorado
A Matter of Record (301) 890-4188
96
1
Hospital.
I coauthored the 2013 ACC AHA
2
cholesterol guideline, and I actually co-chaired
3
the lifestyle guideline for cardiovascular disease
4
risk reduction.
5
Throughout my career, my research has
6
focused on the relationship between metabolic
7
disorders and cardiovascular health.
8
here today know, current lipid-lowering medications
9
fail to meet the needs of many of my patients.
10
the presentation today began by defining three
11
patient groups who need additional treatment
12
options for LDL cholesterol lowering.
13
As many of us
So
As you heard from Drs. Sasiela and
14
Braunstein, alirocumab provided additional LDL
15
cholesterol lowering for these patient groups in 10
16
phase 3 clinical trials with an overall favorable
17
safety profile.
18
the patients I evaluate and treat come from each of
19
these categories, many of whom remain inadequately
20
treated.
21 22
When I'm in the clinic, most of
The most important benefit observed in the alirocumab program was a substantial LDL
A Matter of Record (301) 890-4188
97
1
cholesterol reduction.
2
of approximately 45 to 60 percent in the alirocumab
3
groups were not only superior to the comparators,
4
they're also much greater than can be achieved with
5
any currently approved lipid-modifying agent that
6
would be added to statin therapy.
7
LDL cholesterol reductions
In these studies, the benefit of alirocumab
8
extended beyond lowering LDL cholesterol.
9
Alirocumab also reduced other pro-atherogenic
10
biomarkers as compared with placebo or ezetimibe.
11
And the initial analyses of cardiovascular outcomes
12
in the LT study and in the pooled analysis are very
13
promising, showing not just lack of harm but the
14
potential for benefit.
15
Alirocumab offers two doses, so we can
16
tailor alirocumab treatment to the specific needs
17
of our patients.
18
75 milligrams administered every other week.
19
this dose provides about a 45 percent LDL
20
cholesterol lowering on average, which is going to
21
be sufficient efficacy for many of my patients.
22
Moreover, the 150 dose is available for
The usual starting dose is
A Matter of Record (301) 890-4188
And
98
1
patients who do not get sufficient LDL cholesterol
2
lowering with the 75-milligram dose and need to
3
up-titrate; and this may include patients at
4
particularly high CVD risk or patients with very
5
high LDL cholesterol levels.
6
options will allow physicians like myself the
7
flexibility we need to choose the most appropriate
8
dose for individual patients.
9
Having the two dose
With respect to safety, alirocumab was
10
associated with few specific adverse events.
As
11
you've heard from Dr. Braunstein, the safety and
12
tolerability were comparable to the controls among
13
more than 3,000 patients in the clinical trials.
14
And while no clinical program can definitively
15
characterize the safety profile of a medication,
16
I'm reassured by the sponsor's ongoing efforts to
17
expand the safety database through their large
18
cardiovascular disease outcomes trial.
19
Considering the balance of alirocumab's
20
potential benefits and risks, it's important to
21
acknowledge that LDL cholesterol is a powerful
22
predictor of cardiovascular disease events.
A Matter of Record (301) 890-4188
99
1
Findings from studies in genetics, epidemiology,
2
and outcomes trials consistently show that lower
3
LDL cholesterol levels are associated with lower
4
risks for CVD events. In addition, and as Dr. Cannon showed you,
5 6
reducing LDL cholesterol can reduce the risk of CVD
7
events, and this effect appears to be independent
8
of the medication used.
9
consistency of the evidence has led to the
The breadth and
10
widespread adoption of LDL cholesterol to be a
11
major target for LDL reduction and reduction of
12
cardiovascular disease risk. Although a specific target for LDL that
13 14
provides maximum cardiovascular disease benefit has
15
not yet been defined, we're hopeful from additional
16
work that ultimately this level can be achieved.
17
An ongoing cardiovascular outcomes trial with
18
alirocumab will soon establish whether the LDL
19
cholesterol lowering achieved with alirocumab can
20
also reduce the risk of cardiovascular disease
21
events.
22
In closing, I'll remind you that alirocumab
A Matter of Record (301) 890-4188
100
1
produced unprecedented LDL cholesterol lowering in
2
clinical trials and excellent tolerability.
3
on observations from studies showing that LDL
4
cholesterol reduction lowers the risk of major
5
cardiovascular disease events, it's reasonable to
6
predict that the additional lowering achieved with
7
alirocumab will further decrease this risk.
8 9
Based
Considering its favorable overall benefit/risk profile, I feel very strongly that
10
alirocumab should be made available for the
11
patients with the greatest need for additional LDL
12
cholesterol lowering, particularly patients with
13
heterozygous FH, patients at high or very high CVD
14
risk, and patients who are intolerant of statins.
15
And this is the patient group I see routinely.
16
Thank you very much for your time, and I'm
17
going to ask Dr. Jay Edelberg to return to the
18
lectern to answer any of your questions.
19 20
Jay?
Clarifying Questions DR. SMITH:
I'd like to thank the sponsor
21
for those presentations.
And so we now have time
22
for clarifying questions from members of the
A Matter of Record (301) 890-4188
101
1
committee.
2
signal me or Philip Bautista on my right, we'll
3
generate a list.
4
now, we'll have opportunity later.
5
particularly, we'd like to direct these questions
6
to the sponsor and the sponsor presentation.
7
have more time for the FDA later.
8 9
The way we'll handle this is if you
If we don't get to your questions And I think
We'll
Finally, I would ask if you would identify yourself for the record.
When you ask your
10
question, it would be helpful if you can direct it
11
to a specific individual, though not necessary.
12
we'll start with Dr. Hiatt.
13
DR. HIATT:
Thank you.
William Hiatt.
So
I
14
have a question for Dr. Cannon on your slide, 21.
15
This is a well-known relationship between reduction
16
in LDL cholesterol and reduction in major
17
cardiovascular events.
18
gratifying to see the results of the IMPROVE-IT
19
trial that you presented at the American Heart
20
Association.
21
of drugs that, for the lowering of LDL cholesterol,
22
it fell along that line
And I felt that it was
Demonstrated with a different class
A Matter of Record (301) 890-4188
102
1
My question to you is -- and this is
2
somewhat rhetorical -- if you were to plot the
3
results of LDL lower cholesterol with other drug
4
classes such as CETP inhibitors, niacin, fibrates,
5
omega 3 fatty acids, bile acid binders, would you
6
expect to see the same relationship?
7
DR. CANNON:
This is an excellent question.
8
I have not done that exercise other than the Heart
9
Protection 2 study, in part because the changes in
10
LDL by the agents you've mentioned are very small.
11
And so the need for lots of events is to give
12
appropriate power to make that assessment is what's
13
necessary.
14
So in IMPROVE-IT, we had 5,300 events in
15
order to adequate precision to plot.
If one does
16
do the Heart Protection 2 study, it falls but has
17
very wide confidence intervals, so it's roughly on
18
the line.
19
possible simply by power, but we did with
20
IMPROVE-IT.
So that exercise really hasn't been
21
DR. HIATT:
Well, let me ask you the
22
follow up on that.
Do you then believe that any
A Matter of Record (301) 890-4188
103
1
drug by any mechanism that lowers the LDL
2
cholesterol will retain that same clinical benefit,
3
yes or no?
4
DR. CANNON:
The lowering of LDL should have
5
benefit, I think, here.
Other components of that
6
are obviously safety.
7
and then be able to show effects on LDL and its
8
consequences.
9
drugs -- and torcetrapib is the example.
And so a drug has to be safe
So I think that's where other It had
10
many other effects, adverse safety effects, so that
11
obscures the relationship of LDL to outcomes.
12
DR. SMITH:
Dr. Sager?
13
DR. SAGER:
One of the issues we'll grapple
14
with today, I'm sure, is statin intolerance in
15
patients who can't tolerate statins.
16
ALTERNATIVE trial, where patients were randomized
17
to one of the arms to atorvastatin, 20 milligrams,
18
69.8 percent of them completed a trial, obviously,
19
could tolerate it.
20
In your
How sure are you that these subjects were
21
really statin intolerant?
You used a reasonable
22
definition, but is it maybe that our definitions
A Matter of Record (301) 890-4188
104
1
aren't really the best definitions or there's some
2
issue around how physicians maybe define patients
3
as statin intolerant?
4
DR. EDELBERG:
In the ALTERNATIVE trial,
5
which studied statin intolerance, we did use the
6
definition by the NLA of patients having previously
7
been intolerant to two statins, one at the lowest
8
approved dose.
9
trial was that the patients had to be willing to be
10
rechallenged with a statin during the blinded phase
11
of the program, which obviously limited the ability
12
of some of the patients to even consider the trial.
But the important component of the
DR. SAGER:
13
Okay.
That's a good point.
14
However, one could still take away from this that a
15
large number of individuals that are considered
16
statin intolerant truly do tolerate statins. DR. EDELBERG:
17
Absolutely.
The ability of
18
many patients with a previous history of statin
19
intolerance to be able to be successfully
20
rechallenged with a statin was not only shown in
21
the ALTERNATIVE trial, but in previous trials after
22
that.
A Matter of Record (301) 890-4188
105
I'd ask Dr. Eckel, an expert in this area,
1 2
to actually comment on this and how difficult it is
3
to treat these patients overall. DR. ECKEL:
4
Yes.
This is my most common
5
referral.
I would say two out of three patients
6
referred to the Lipid Clinic at the University of
7
Colorado Hospital are statin-intolerant patients by
8
referral. Now, this is an art of medicine, not a
9 10
science of medicine, and definitions really have
11
been many, but none are succinct because the
12
science here is difficult to define.
13
rhabdomyolysis is a complication of statin therapy
14
we rarely see, fortunately.
15
define themselves.
16
intolerance, the definition used here is one of
17
many.
18
intolerance and have an algorithm about how to
19
treat these patients.
20
Clearly,
Those patients clearly
But in terms of statin
And in fact, I've written on statin
So most of us in the field realize that we
21
can give a low-dose potent statin once a week in up
22
to 80 percent of people, and have them tolerate
A Matter of Record (301) 890-4188
106
1
that, and escalate the dosage slowly until
2
intolerance develops.
3
people are high-risk patients of either events or
4
at high risk for events.
However, almost all of these
So the art of medicine is really working
5 6
with the patient, her or him, to try to understand
7
how serious the side effect is and whether we can
8
ultimately get them on the maximum statin therapy
9
before we would consider other agents.
And by the
10
way, the other agents, back then to Dr. Hiatt's
11
question, just aren't sufficient to really get the
12
LDL cholesterol down to acceptable levels. So that's a long answer, but I hope I
13 14
clarified the difficult patient with statin
15
intolerance in terms of defining them in the
16
clinic.
17 18 19
DR. SMITH:
Dr. Wilson, did you have a
follow-up question related to that? DR. WILSON:
Yes.
Peter Wilson.
So you
20
have considered a narrow definition of statin or
21
what we would call partial statin intolerance.
22
Many of us in lipidology agree with Dr. Eckel's
A Matter of Record (301) 890-4188
107
1
approach.
2
statins -- and we go through three or four.
3
you only took comers with one or two failures.
4
Right?
5
We've got six or seven statins -- six
DR. EDELBERG:
That's right.
The definition
6
for entry was previously two failures that
7
demonstrate the trial.
8
did have more than two previous failures.
9
DR. WILSON:
But
But some patients actually
Then I think important for
10
those not accustomed to lipids is the challenge was
11
predicated on a single agent as a challenge and not
12
a statin for which these patients were naive for a
13
challenge.
14
intolerance, but it was within the confines of what
15
you were undertaking.
So this is an issue for statin
16
Then just to end, in our experience in the
17
VA, about 70 percent of patients can be on statins
18
after two years, and they have 30 to 40 percent,
19
and the committee should be aware of that.
20
DR. EDELBERG:
And we agree that patients
21
who can take statins should take statins.
22
they can get to their goal with a statin, that's
A Matter of Record (301) 890-4188
And if
108
1
the optimal situation.
2
can't take statins or can't take enough statin to
3
get to their goal, there's an unmet medical need.
4
DR. SMITH:
5
DR. BURMAN:
But for the patients who
Dr. Burman? This is Ken Burman.
Either
6
Dr. Edelberg or Dr. Eckel, on slide 100, you
7
mentioned the pro-atherogenic biomarkers, but you
8
left out CRP.
9
not decreased with PCSK9 inhibitors.
And my understanding is that CRP is I couldn't
10
find great data to support that, but that's my
11
impression.
12
Would you like to comment on that?
DR. EDELBERG:
In our program, the majority
13
of the patients treated were on statins.
14
those on statins, they were on a maximally
15
tolerated dose of statin.
16
decrease in CRP in our patients in the program.
17
DR. SMITH:
And of
We did not observe a
Did you want to see data if
18
that's available, Dr. Burman, or is that an
19
adequate answer?
20
DR. BURMAN:
That would be great if the
21
panel would allow it.
22
DR. EDELBERG:
Dr. Sasiela will review these
A Matter of Record (301) 890-4188
109
1 2
data with you. DR. SASIELA:
I'll show you the slide here.
3
This is pooled data looking at patients within the
4
placebo-controlled trials on the left, so if you
5
look at the pink box and the blue box in the
6
left-hand set, you'll see the pink is placebo; the
7
blue is alirocumab.
8
ezetimibe-controlled studies, those with background
9
statins in the middle, and those without statins on
10 11
The next set will be the
the right. So you can see across all these studies,
12
really, the mean change in CRP is sitting right
13
around zero.
14
DR. SMITH:
Dr. Cooke?
15
DR. COOKE:
For Dr. Braunstein, with regard
16
to the long-term diabetes risk, were there fasting
17
insulin levels examined in any of these trials?
18
And if so, what were the results?
19
DR. ECKEL:
So we looked at mean-fasting
20
glucose levels over time in patients who had
21
different categories that we assigned to them at
22
baseline, who were normal or were diabetic.
A Matter of Record (301) 890-4188
And if
110
1
we looked at that over time, we saw that there was
2
no change.
3 4 5
I actually have the slide now for you.
DR. COOKE:
So fasting insulin levels, not
glucose. DR. ECKEL:
Oh, I'm sorry.
I apologize.
6
No, we did not do fasting insulin levels.
7
studies were not designed with that in mind, so no,
8
we did not look at that.
9 10
DR. SMITH:
No, the
Dr. Gellar?
DR. GELLER:
Nancy Gellar.
I have a concern
11
about the LDLs below 25.
12
that by treatment, but I assume that the LDLs
13
didn't go below 25 in the non-drug group.
14
secondly, you mentioned -- I think this was
15
Dr. Braunstein -- that there was no down-titration
16
in the trials.
17
the ongoing outcomes trial?
18
you recommend in clinical practice when LDL falls
19
below 25 and based on what?
20
First, you didn't show
And
Is that true in outcome as well in
DR. EDELBERG:
And secondly, what do
In our clinical program, we
21
did not do down-titration.
We specifically wanted
22
to examine the safety of patients getting to LDL's
A Matter of Record (301) 890-4188
111
1
below 25.
And as Dr. Braunstein pointed out, we
2
didn't see any safety signal. In our outcomes study, patients will be
3 4
down-titrated when reaching below 25 milligrams per
5
deciliter.
6
milligrams, on 150 milligrams, they'll be moved
7
down to the 75-milligram dose. For patients who get below 15 milligrams,
8 9
So if a patient reaches below 25
they actually can be withdrawn from therapy from
10
the 75-milligram dose.
11
our program is that, with the flexible dosing
12
regimen, the usual starting dose of 75 milligrams,
13
we're able to decrease the number of patients who
14
get to below the 25 threshold and, moreover, be
15
able to decrease that below 15 milligrams per
16
deciliter. Actually, Dr. Braunstein is going to review
17 18
An important component of
those data with you. DR. BRAUNSTEIN:
19
Can I have EL-50?
So in
20
our cardiovascular outcomes trial, we set a lower
21
limit of 15 for where patients would remain in the
22
study.
But it turns out, with our dosing regimen,
A Matter of Record (301) 890-4188
112
1
that's going to be a very rare occurrence because
2
if you look at our experience in our phase 3
3
program, this is a very similar slide that I showed
4
in my presentation.
5
percentage of patients who achieve a 25, this is
6
the percentage of patients who achieve a level of
7
15 or less on two or more consecutive occasions.
8
And although we still see that at the high dose, in
9
patients treated, or started at the low dose, it's
10 11
But instead of showing you the
a very, very rare occurrence. So we expect to be able to retain patients
12
in our cardiovascular outcomes trial because at the
13
low dose, we're really rarely seeing patients
14
getting below 15.
15
DR. GELLER:
So if you get below 25 and
16
you're on the lower dose, do you stop treatment,
17
and if so, for how long?
18
DR. EDELBERG:
In the outcomes study, we
19
would continue on therapy.
And outside of the
20
outcomes study, this is really based on the
21
clinical judgment.
22
physicians will want to continue their patients
We recognize that some
A Matter of Record (301) 890-4188
113
1
with LDLs below 25, whereas other physicians may
2
want to suspend therapy.
3
clinical judgment, we think.
4
DR. SMITH:
5
DR. SHAMBUREK:
But this is left to
Dr. Shamburek? Two questions on mechanism.
6
With the PCSK9 having its effect on the LDL
7
receptors, you have the group of one study with all
8
heterozygous FH, where they're all predominantly
9
fewer -- there's a heterogeneous of the receptors,
10 11
but they're presumably half or a big variance. When you looked at that group compared to
12
the combination or others, any reason why you
13
didn't see either a lack of effect since they're
14
starting out presumably with less?
15
information?
16
number, the PCSK9 concentration, or does that group
17
maybe get up-titrated and you lose it in the
18
background noise?
19
Do you have any
Is that due to the LDL receptor
DR. EDELBERG:
We saw a comparable efficacy
20
with both the 75- and the 150-milligram dose in
21
non-FH and FH patients.
22
review the other aspects.
Dr. Sasiela will actually
A Matter of Record (301) 890-4188
114
DR. SASIELA:
1
I just want to make sure I
2
understand your questions to clarify.
3
your question was in regards to why we saw
4
comparable efficacy between FH and non-FH? DR. SHAMBUREK:
5
Yes.
So I think
You would think
6
sometimes rare genetic disorders can give you input
7
to mechanism and things, so almost by definition
8
heterozygous FH are half the amount of receptors.
9
Your other group could be a mixture that could have
10
similar pathology, but they may be hyper-secreting
11
LDL or something like that, and you might see a
12
difference. So does it tell you something about is it
13 14
the PCSK9 concentration?
15
number?
Any information you can tease out. DR. SASIELA:
16
Is it the LDL receptor
Yes.
There are a couple
17
things, I think.
18
we looked clinically in our studies, both phase 2
19
and confirmed in phase 3, we really didn't see any
20
difference in efficacy between the FH and the
21
non-FH.
22
As Dr. Edelberg mentioned, when
We did some very interesting preclinical
A Matter of Record (301) 890-4188
115
1
studies with some outside collaborators, where we
2
looked at a spectrum -- we took cells from patients
3
with a spectrum of FH, both patients who were
4
non-FH, patients who were heterozygous FH, and
5
various forms of homozygous FH, and looked at the
6
impact of both PCSK9 on LDL receptor numbers, as
7
well as alirocumab impacting the effect of PCSK9
8
and LDL receptor numbers. What we saw was that, in every case, except
9 10
for patients who are homozygous null-null LDL
11
receptors defects, that PCSK9 would influence the
12
number of -- would down-regulate the number of LDL
13
receptors.
14
could restore those LDL receptor numbers back
15
higher.
And if we block it with alirocumab, we
I mean, if you want to see those, it's --
16
DR. SHAMBUREK:
17
DR. SASIELA:
No. Okay.
That's fine. So I think we've got a
18
pretty good sense that the bottom line is, I think,
19
regardless of the genetic defect for heterozygous
20
FH, these patients are going to see a similar
21
response to non-FH patients.
22
DR. SHAMBUREK:
And one other follow-up with
A Matter of Record (301) 890-4188
116
1
the mechanism, I think two speakers mentioned the
2
reduction in Lp(a) and something between 17 to
3
30 percent.
4
50 milligrams per deciliter.
5
of the mechanism and why it's different, or, number
6
two, were the patients who had the highest Lp(a)
7
the ones who had the reductions, or was it across
8
the board?
9
And the baseline was about
DR. SASIELA:
Do you have any idea
No, absolutely.
This was I
10
think a really interesting and somewhat exciting
11
finding from the program, given that statins have
12
shown no reduction in Lp(a).
13
this in a number of ways.
14
working to tease out the exact mechanism.
15
got some recent preclinical data that shows that
16
there may be a reduction in LP apo(a) production
17
that's impacted, that PCSK9 has a direct influence
18
on that, though I am also aware there are also some
19
recent papers that still show some influence of the
20
LDL receptor.
21 22
We have looked at
I think we're still We've
So I think we're still working this out, and we've got an ongoing lipoprotein kinetic study that
A Matter of Record (301) 890-4188
117
1
should hopefully shed some additional light on
2
this.
3
baseline, reductions in Lp(a) seemed to be, for the
4
most part, independent of baseline Lp(a).
In looking at your other question about
5
DR. SMITH:
6
DR. ORZA:
Dr. Orza? I have several questions, but I
7
first have to make a disclosure.
In the early
8
2000s, I was on the staff of the American College
9
of Cardiology.
And my staff and I supported the
10
activities of the college members such as guideline
11
development, but we did not have any control over
12
them.
13
Also in that capacity, I supported several
14
committees which Dr. Cannon, one of the sponsor's
15
experts, served.
16
going to bias my thinking in any way, but I think
17
it's fair to have it in the record.
18
And I don't think that that is
I have several questions, and some of them
19
have very short answers, but if it gets to be too
20
many, I'll ask to be put back in the queue.
21 22
The first one is probably for Dr. Sasiela, and it's if you could provide more information
A Matter of Record (301) 890-4188
118
1
about the diet component of the trial treatments,
2
especially in relationship to your findings about
3
BMI.
4
background materials, it seemed that the majority
5
of the patients were overweight or obese, according
6
to their BMI.
7
were any changes in that.
It wasn't in the slides, but in the
And I was just wondering if there
8
The second question is about --
9
DR. SMITH:
I would suggest maybe let's go
10
one question at a time.
11
going after each one is answered.
12 13 14
DR. EDELBERG:
Okay?
And then you keep
I'll ask Dr. Sasiela to come
up and address the first question there. DR. SASIELA:
To address your question about
15
diet, as is the hallmark of pretty much all lipid
16
trials, we inform and guide patients on appropriate
17
diet for hyperlipidemia and hypercholesterolemia.
18
And we try and check and make sure that they're
19
maintaining that throughout the course of the
20
study, nothing with diaries or anything, but
21
through questionnaires.
22
good following of the diet.
And we saw that we had a
A Matter of Record (301) 890-4188
119
1
We also tracked BMI and didn't see really
2
any significant change in BMI or weight over the
3
course of the studies.
4
DR. ORZA:
So how do you think that's
5
related to the effectiveness of the diet component
6
and/or do you think it has any relationship to the
7
effects of the LDL-lowering medications?
8
DR. SASIELA:
I'm sorry.
9
regards to your question.
10
DR. ORZA:
I'm not sure in
Well, if the diet component of
11
the treatment regimen were successful, we would
12
expect to see people's BMIs going down, presumably.
13
DR. SASIELA:
These were -- sorry.
14
DR. EDELBERG:
Let's ask Dr. Eckel to
15
actually address this.
16
able to actually comment on diet and LDL, too.
17
might be helpful here.
18
DR. SASIELA:
I think that he might be
Yes.
I mean, just they
19
weren't counseled on weight loss.
20
a lipid heart-healthy diet.
21 22
DR. ECKEL: a good question.
It
It was just for
I have a couple comments.
It's
I think the false perception is
A Matter of Record (301) 890-4188
120
1
that with weight loss, LDL cholesterol falls.
2
does during active weight reduction, but once
3
people have reached the lower body weight and are
4
sustaining the weight loss, LDL comes back to
5
baseline unless the kind of recommendations we put
6
forth and the lifestyle recommendations of the ACC
7
and AHA are heeded to:
8
trans-fat, et cetera.
9
great therapy for LDL long term.
10
It
low saturated fat, low So weight loss is not a
The second point is, if you look at the
11
statin-related diabetes incidence, it's interesting
12
to look at the fact that the weight gain that
13
occurs on statin therapy, which is modest.
14
on that phase of the curve.
15
upward, which may in part relate to some of that
16
weight gain.
17
But is
In fact, it goes
So there's no question that we should have
18
patients on a good lifestyle.
19
balance between healthy nutrition and also physical
20
activity.
21
in a proactive way.
22
And that includes a
So I'm going to build on your question
DR. ORZA:
The second question is -- I'm not
A Matter of Record (301) 890-4188
121
1
sure who it's for, but it's whether you have any
2
information about patient preferences of patients'
3
tolerance of the self-injection every two weeks and
4
how that relates to how they feel about taking a
5
pill every day, just information around the method
6
of administration of these drugs.
7
DR. EDELBERG:
I'm going to ask Dr. Sasiela
8
to review.
9
pre-filled syringe and the pre-filled pen
10 11
We had very good compliance in both the
injections throughout the long-term trials. DR. SASIELA:
Yes, I think this was one of
12
the questions we had in terms of providing this, a
13
relatively new injectable for treatment of
14
hypercholesterolemia.
15
could show the compliance data over the course of
16
the 10 studies.
17
And I was wondering if I
if I can get that up, I'll show you that in
18
general -- I have seen it enough times.
19
you that, overall, between both the placebo
20
injections and the alirocumab injections, the
21
compliance was very high, ranging in 97, 98 percent
22
across both groups.
A Matter of Record (301) 890-4188
I can tell
122
DR. ORZA:
1
Then this is a quick one.
When
2
do you expect enrollment on your outcomes study to
3
end?
4
DR. EDELBERG:
We look forward to our
5
outcomes study being fully enrolled by the end of
6
the year, and we'll be looking forward to having
7
the trial complete by the end of 2017.
8
DR. ORZA:
Then in the statin intolerance
9
group on slide 40, these were people who were
10
intolerant of a statin but were willing to be
11
randomized to maybe getting a statin again.
12
DR. EDELBERG:
13
DR. ORZA:
That's right.
So can you speak to how that
14
group might be different than the larger group of
15
statin-intolerant patients or those who are having
16
problems with statins and how that might affect the
17
results?
18
DR. EDELBERG:
Yes.
We recognize it.
19
did enroll a subset of patients with statin
20
intolerance because they were willing to be
21
rechallenged.
22
person to comment on how this fits into, more
I think Dr. Eckel is the right
A Matter of Record (301) 890-4188
We
123
1
broadly, the clinical picture. DR. ECKEL:
2
This is a very select group, and
3
I think you have to interpret the data for what
4
it's worth.
5
statins again.
6
on three or four statins already in a variety of
7
doses and are absolutely unwilling to go back on
8
statins.
It's worth people willing to go on Many of the people I see have been
So again, like I said earlier, this is an
9 10
art of medicine more than it is science.
11
statin therapy.
12
drug that has major LDL-lowering capabilities is
13
something I need, and I think we all need in this
14
group of patients.
15
the art is right now.
16
really a population that needs special attention, I
17
think.
18 19 20
They need
So the alternative to have another
How poorly we define it is what But nevertheless, this is
I'm not sure I answered your question, but I can't any better than what I did. DR. ORZA:
I'm just wondering, the people
21
who were really having a problem on statins and
22
such that they would not be willing to be
A Matter of Record (301) 890-4188
124
1
challenged again with a statin, we really don't
2
know about those people and the problems that they
3
were having with statins, what kinds of problems
4
they might be having with -- I can't pretend I know
5
how to pronounce alirocumab.
6
DR. ECKEL:
Is that correct?
That's why these patients are
7
referred, because many of us who are in the
8
trenches of dealing with more complicated patients
9
see a tremendous amount of statin intolerance.
We
10
have to do our best to define it clinically and try
11
to re-enter statins when and if possible.
12
sometimes, we just can't.
13
within a week I have a phone call.
14
me know the patient, again, is having problems.
15
DR. SMITH:
16
DR. WILSON:
And
Or when I do, it's My nurse lets
Dr. Wilson? The first simple question
17
is -- I have two.
The simple one is, is there any
18
change in tender xanthomas in patients with
19
extremely high cholesterol in those findings with
20
statins?
21
then the xanthomas go away or improve?
22
information?
And then you add this new molecule, and
A Matter of Record (301) 890-4188
Any
125
1
DR. EDELBERG:
2
comment specifically.
3
heterozygous patients, and those were patients who
4
were already on maximally tolerated doses of
5
statin.
6
DR. SASIELA:
7
DR. WILSON:
I'll ask Dr. Sasiela to We enrolled only
So the answer is not needed. The second one is related to
8
very low levels of LDL cholesterol or
9
apolipoprotein B.
And this condition is called
10
hypobetalipoproteinemia, and they're susceptible to
11
neurologic diseases, retinitis pigmentosa,
12
acanthosis.
13
be at the heart of all this, vitamins K, A, D, and
14
E, and partial are very severe deficiencies.
15
those patients get vitamin supplements.
16
and vitamin disorders are thought to
And
So the question is, what have been the
17
strategies and what data you might have related to
18
vitamin levels in patients with LDLs less than 50
19
in these trials?
20
DR. EDELBERG:
Dr. Braunstein's going to
21
review the data on vitamin valuation in the
22
program.
A Matter of Record (301) 890-4188
126
DR. BRAUNSTEIN:
1
So we measured vitamin A,
2
D, E, and K levels in the LT study, And we saw no
3
changes in A and D and only some very small changes
4
in K.
5
similar in pattern to what you see with the statin.
6
And that is, vitamin E went down, but the vitamin E
7
to LDL ratio increased.
8 9
Now, the vitamin E decreases were very
But importantly, when we looked at the adverse events in those patients, in the patients
10
who had the lowest levels of LDL -- the patients
11
who had low levels of vitamin E were almost
12
exclusively in the group of patients who had low
13
levels of LDL.
14
So I put it up on this slide.
We had 31
15
patients in the study who actually had low levels
16
of vitamin E, and of those, 21 were patients with
17
LDLs less than 25.
18
events, like the retinal abnormalities, the ataxia,
19
the neurologic findings, none of those patients had
20
any of those types of adverse events that you might
21
be concerned about.
22
DR. SMITH:
But when we look for adverse
Dr. Hiatt?
A Matter of Record (301) 890-4188
127
DR. HIATT:
1
I have a question on drug
2
safety.
3
be asked a series of questions about drug safety
4
based on 5,200 patient-years of exposure, which I
5
would just comment is not a lot of exposure for
6
therapy that's designed to be long term and
7
potentially millions of patients, though I
8
recognize the design of the trials to target LDL
9
cholesterol levels would not have a lot of exposure
10
This is William Hiatt.
And we're going to
at this stage of development.
11
So I'd like to just ask a question.
12
sponsor's background document, table 35, which you
13
don't have to pull up -- I'll just read -- this is
14
a shift plot, patients with impaired glucose
15
tolerance at baseline, shifting to diabetes.
16
that's in placebo-controlled trials, 3.8 percent
17
versus 5.7 percent.
18
closer to 3.3 versus 3.9.
19
In the
And
In ezetimibe-controlled, it's
So I recall several years ago reviewing the
20
JUPITER trial, and at that time, I think the risk
21
of diabetes with statin therapy became
22
obvious -- though I think if you look back at the
A Matter of Record (301) 890-4188
128
1
adverse events and some of the other large
2
cardiovascular outcomes trials with other statins,
3
we did see diabetes risk at that time as well.
4
So I want to know from the sponsor's
5
perspective, you've sort of presented this as a
6
no-risk kind of thing.
7
confident you are of saying that this drug could
8
not be associated with an increased risk of
9
developing diabetes.
But I would wonder how
And the reason I'm asking
10
that obviously is that I'm not sure we have enough
11
information to rule that risk out specifically.
12
I guess I'd like to know what your position
13
is on this in that, do you feel for our review
14
today that we should assume that the absence of
15
evidence is evidence of absence?
16
what I'm asking.
17
DR. EDELBERG:
I guess that's
We reviewed glycemic control
18
several different ways, and found no impact overall
19
on glycemic control.
20
review this with you.
21 22
Dr. Braunstein is going to
DR. BRAUNSTEIN:
So as Dr. Edelberg said, we
haven't seen what we consider to be a signal in our
A Matter of Record (301) 890-4188
129
1
data, but I do want to show you those data.
But to
2
your more specific pointed question, we would never
3
assume that the absence of evidence is the proof of
4
absence.
5
we're not seeing that evidence.
6
would start first with the figure that you were
7
referring to.
But in the large program that we have, And I thought I
And that is --
8
DR. HIATT:
Table 35 in your background.
9
DR. BRAUNSTEIN:
Yes.
We also have that
10
same -- right.
So we can put that up.
11
saw here is when we look at patients --
12
DR. HIATT:
13
DR. BRAUNSTEIN:
So what we
That's it. -- yes.
And we actually
14
have one that's easier to read, so I'm just going
15
to put up this.
16
button, but there we go.
17
version of that.
18
I think I thought I hit the This is an easier-to-view
So this is patients who were impaired
19
glucose control at baseline based on some criteria
20
that we identified.
21
diabetes at any time during the study, then they
22
were considered to have shifted.
And if they worsened to
A Matter of Record (301) 890-4188
And as you
130
1
pointed out, there was a small imbalance in the
2
numbers in the alirocumab versus placebo group.
3
But also in the same analysis, you see there's also
4
an increased number of patients who have regressed
5
to the normal group.
6
So we see regression basically in both
7
directions in this analysis, so it's hard to
8
understand or interpret -- it's hard to know for
9
sure exactly what that means.
So when we look at
10
all of our analyses, for example, here, what we
11
also looked at is an analysis of first use of new
12
diabetic medications.
13
impaired glucose control generally wouldn't be on
14
medication, we were concerned about patients
15
developing more frank diabetes that would be
16
clinically recognized.
17
Because patients with
When we look at that -- when we see that,
18
we're seeing that there's no evidence for increased
19
risk of diabetic use.
20
hemoglobin A1C over time across the studies -- I
21
think we can get that slide up for you quickly.
22
DR. HIATT:
And also, if we look at
Actually, I'm sorry.
A Matter of Record (301) 890-4188
Don't move
131
1
on from this slide. DR. BRAUNSTEIN:
2 3
I'm sorry.
Why don't we go
back to this slide? DR. HIATT:
4
Well, where you are right now,
5
the hazard ratio of .70 is reassuring.
6
upper bound is 1.73. DR. BRAUNSTEIN:
7 8 9
Oh, yes.
But the
No, no, no.
Right. DR. HIATT:
We look at safety data
10
differently than we look at efficacy data.
11
We're not testing a hypothesis here.
12
asking what the upper bound of risk could be.
13
Similar questions are asked in diabetes drugs.
14
What's the upper bound of MACE?
15
is 1.73 to 2.52.
16
Right?
We're just
Upper bound here
So you're trying to present a case that sort
17
of says there's nothing here.
18
challenging your data.
19
go from 5,000 patient-years to 50,000
20
patient-years, are you sort of telling us that we
21
shouldn't worry about this?
22
DR. BRAUNSTEIN:
I see that.
I'm not
I'm just saying, when you
Well, I can tell you that,
A Matter of Record (301) 890-4188
132
1
with 50,000 patient-years, that's what we're
2
actually going to have in our outcomes study.
3
this remains an adverse event of interest in our
4
outcomes study.
So
Now, we recognize with the statins the
5 6
question or the evidence for abnormalities in
7
glycemic control weren't really recognized until
8
the very large databases.
9
looking at this in our 50,000 patient-year outcomes
So we're committed to
10
study.
11
we have -- we can presumably come back to this
12
committee and would be much more definitive on this
13
point.
14
And that I think would be clearly -- there
DR. HIATT:
Sure.
I mean, there have been
15
editorials written about this risk.
16
benefit of lowering LDL far exceeds the risk of new
17
onset diabetes.
18
here, are you telling us that there's not a
19
diabetes risk today for our review of the safety
20
question?
21 22
I think the
But I am just asking the question
DR. BRAUNSTEIN:
Well, all I'm saying is,
for your review today, we don't find evidence for
A Matter of Record (301) 890-4188
133
1
diabetes risk in a 5,300 patient-year database.
2
That's all I can tell you today.
3
recognize we are committed to continuing to look at
4
this.
5
which will have much more --
7
10
Which would be much more
definitive and is a terrific opportunity. DR. BRAUNSTEIN:
8 9
We'll have a 50,000 patient-year database,
DR. HIATT:
6
But please also
But I think what we can say
is, if there was a large risk, we would have seen it by now.
Right?
11
DR. HIATT:
Sure.
12
DR. BRAUNSTEIN:
So whatever risk, if there
13
is anything, we're talking about something that
14
would be rather small.
15
that in mind, too.
16
risks.
17
DR. HIATT:
So we should at least keep
We're not talking about large
Agreed, but we didn't understand
18
that risk with statins until the large outcome
19
trials and the more mature ones, quite frankly,
20
came along.
21 22
DR. SMITH:
Because of time, we're going to
take a break for 15 minutes.
I am well aware that
A Matter of Record (301) 890-4188
134
1
there are several panel members who have questions,
2
who didn't yet have an opportunity to ask them, and
3
I will come back to you later during the
4
discussions.
5
those concerns.
There will be opportunity to cover
So we'll now take a 15-minute break.
6
Panel
7
members, I ask that you please remember there
8
should be no discussion of the meeting topics
9
during the break, among yourselves, or with any
10
member of the audience.
11
meeting at 10:30. (Whereupon, at 10:16 a.m., a recess was
12 13 14
We're going to resume this
taken.) DR. SMITH:
Please head for your seats so we
15
can get the next part of this meeting today
16
started.
17
(Pause.)
18
DR. GOLDEN:
19 20
Good morning, Chairman Smith
and members of the committee. DR. SMITH:
I'm sorry.
Just one moment -- we are going
21
to come back to the questions that I know are
22
lingering from panel members, later in the morning,
A Matter of Record (301) 890-4188
135
1
subsequent to the FDA presentations, so we'll now
2
proceed with FDA presentations. FDA Presentation – Julie Golden
3 4
DR. GOLDEN:
Thank you.
5
Good morning, Chairman Smith and members of
6
the committee.
7
with the Division of Metabolism and Endocrinology
8
Products.
9
today, and I'm pleased to provide you with our
10 11
I'm Julie Golden, a medical officer
I am the first of two FDA speakers
perspective of alirocumab's efficacy. Here is an outline of the FDA talks.
First,
12
I will provide a brief background of the topic and
13
an overview of some of the key features of the
14
phase 3 program.
15
controlled trials first, focusing on the primary
16
LDL cholesterol endpoint and some dosing issues.
17
will then briefly discuss some of the other lipid
18
parameters.
19
I will discuss the placebo-
With respect to the active controlled
20
trials, I will mention some clinical
21
considerations, then separately discuss the trial
22
in patients with statin intolerance.
A Matter of Record (301) 890-4188
Finally, I
I
136
1
will briefly discuss neutralizing antibodies and
2
potential loss of efficacy.
3
follow with a discussion of the safety review of
4
alirocumab, then I will return with a brief
5
summary.
6
Dr. Mary Roberts will
Historically, LDL cholesterol has been used
7
as a surrogate for cardiovascular risk.
Statins
8
are now considered first-line therapy to treat
9
hypercholesterolemia.
Cardiovascular benefit has
10
been established, and the safety profile is well-
11
characterized.
12
By contrast, cardiovascular outcomes trials
13
with non-statin lipid therapies have shown mixed
14
results.
15
alirocumab is approved based on LDL lowering as a
16
validated surrogate endpoint, a cardiovascular
17
outcomes trial to confirm benefit on clinical
18
outcomes will not be a regulatory requirement.
19
As you heard from Dr. Smith, if
Issues regarding the use of LDL as a
20
surrogate were raised with the sponsor at the
21
end-of-phase-2 meeting.
22
that if the data raised concern about alirocumab's
FDA informed the sponsor
A Matter of Record (301) 890-4188
137
1
potential efficacy to reduce cardiovascular events
2
or about its safety, results from a completed
3
cardiovascular outcomes trial may be required prior
4
to approval.
5
Because at that time the IMPROVE-IT trial
6
was still ongoing, FDA also stated that the results
7
of IMPROVE-IT may influence the decision on whether
8
the cardiovascular outcomes trial needs to be
9
completed prior to approval.
FDA also informed the
10
sponsor that the advisory committee would be asked
11
whether alirocumab would be approved prior to
12
completion of the cardiovascular outcomes trial.
13
Therefore, this is in essence the question
14
we are posing to you today.
15
approved based on LDL cholesterol before the
16
cardiovascular outcomes trial has been completed?
17
Should alirocumab be
Here is a schematic of some key aspects of
18
the phase 3 program.
The five placebo-controlled
19
trials are FH I and II, HIGH FH, long-term, and
20
COMBO I.
21
COMBO II, OPTIONS I and II, ALTERNATIVE, and MONO.
22
All patients in FH I, FH II, and HIGH FH, and
The five active control trials are
A Matter of Record (301) 890-4188
138
1
approximately 18 percent of patients in the
2
long-term trial were identified as having
3
heterozygous familial hypercholesterolemia.
4
Patients in the high FH trial were only eligible if
5
they had an LDL cholesterol at screening greater
6
than 160 milligrams per deciliter. Long-term and COMBO I enrolled patients at
7 8
high or very high risk of atherosclerotic
9
cardiovascular disease without HeFH.
All of the
10
placebo-controlled trials enrolled patients who
11
were on maximally tolerated statin with or without
12
other lipid-modulating therapy not at their LDL
13
goal.
14
Two of the trials utilized the higher
15
150-milligram dose administered every two weeks
16
throughout the trial.
17
enrolled a patient population with high baseline
18
LDL cholesterol and a large long-term trial.
19
HIGH FH, which is noted,
The other eight trials started all patients
20
at the lower 75-milligrams every-two-weeks dosing
21
regimen and titrated patients to 150 milligrams
22
every two weeks at week 12 if they had not achieved
A Matter of Record (301) 890-4188
139
1
a pre-specified LDL goal at week 8.
2
ages were 51 to 53 years in the HeFH trials and 61
3
to 63 years in the long-term and COMBO I trials.
4
Forty-four to 47 percent of patients in the HeFH
5
trials and 34 to 38 percent in long-term and COMBO
6
I were female.
7
trials were white.
8 9
The average
The majority of the patients in the
The COMBO I trial only included U.S. sites. The other trials enrolled zero to 25 percent of
10
patients from the U.S.
11
the HeFH trials and particularly in the HIGH FH
12
trial by design.
13
HeFH was mentioned previously.
14
patients in long-term had documented HeFH.
15
Baseline LDL was higher in
The proportions of patients with Eighteen percent of
The majority of patients in the long-term
16
and COMBO I trials had documented coronary heart
17
disease, as it is a substantial percentage of
18
patients in the HeFH trials.
19
86 percent of patients were on high-intensity
20
statin at baseline, defined as 40 or 80 milligrams
21
of atorvastatin or 20 or 40 milligrams of
22
rosuvastatin daily.
Forty-four 44 to
A Matter of Record (301) 890-4188
140
1
The primary efficacy endpoint was percent
2
change in LDL cholesterol at week 24.
3
trials that utilized the up-titration regimen, mean
4
percent LDL lowering from baseline was similar in
5
the alirocumab-treated patients, approximately
6
48 percent.
7
treatment difference in LDL ranged from 46 to 58
8
percent.
9
In the three
The overall placebo-subtracted
Here are the two trials that utilize the
10
150-milligram dose throughout.
11
lowering from baseline and the placebo-subtracted
12
treatment effect were lower in the HIGH FH trial
13
than in the long-term trial.
14
Note that the LDL
Cross-trial comparisons have limitations.
15
Nevertheless, this could suggest an attenuation of
16
effect in patients who have higher LDL cholesterol.
17
LDL change by baseline LDL subgroups in the long-
18
term and HIGH FH trials was assessed to explore
19
this finding further.
20
In the long-term trial, there is an apparent
21
trend toward attenuated treatment effect with
22
increasing baseline LDL cholesterol.
A Matter of Record (301) 890-4188
141
1
Interestingly, this pattern wasn't seen with the
2
other phase 3 trials and to some extent may be a
3
reflection of a regression to the mean phenomenon
4
in the placebo group since the maximal
5
pharmacodynamic effect was achieved in all
6
alirocumab subgroups. The other thing to point out is that the
7 8
treatment effect in the subgroup from long-term
9
with the highest LDL, which was 160 milligrams per
10
deciliter or greater, was similar to the treatment
11
effect in HIGH FH, in which patients were
12
specifically recruited with an LDL of greater than
13
160.
14
Because the 75- and 150-milligram doses were
15
not studied in parallel randomized arms in phase 3
16
trials, dose response was explored in a few ways.
17
The week 12 LDL analysis allows for a cross-study
18
comparison of 75-milligram and 150-milligram doses,
19
given that dose adjustment in these trials could
20
not occur until week 12.
21 22
Results from FH I, FH II, and COMBO I represent the efficacy achieved with a 75-milligram
A Matter of Record (301) 890-4188
142
1
dose.
2
the HeFH trial subsequently up-titrated, whereas
3
only 17 percent of patients from COMBO I
4
up-titrated.
5
FH and long-term represent the range of efficacy
6
achieved with the 150-milligram dose across
7
populations.
8 9
Of note, about 40 percent of patients from
Again, the variable results from HIGH
The solid line in this figure illustrates mean LDL percent change over time in alirocumab-
10
treated patients from FH I who did not up-titrate,
11
meaning they remained on 75 milligrams throughout.
12
This red line represents the mean LDL percent
13
change over time in patients who had not achieved
14
their LDL goal at week 8 and therefore up-titrated
15
to 150 at week 12.
16
These are the week 12 results for the
17
titrated and untitrated groups when everyone was at
18
75 milligrams, and these are the results of the
19
primary endpoint.
20
fairly similar pattern.
21 22
FH II and COMBO I demonstrated a
There are limitations to assessing dose in phase 3.
As noted before, because no phase 3 trial
A Matter of Record (301) 890-4188
143
1
included parallel 75- and 150-milligram dose arms,
2
dose evaluation is not a randomized comparison and
3
rather relies on either a cross-trial comparison or
4
post-randomization factors.
5
For example, patients who required
6
up-titration were different than those who did not
7
in that they, not surprisingly, had a higher LDL
8
cholesterol at baseline.
9
limitations of these analyses, it appears that
Nevertheless, despite the
10
there may be patients for whom the higher
11
150-milligram dose will be useful for additional
12
LDL lowering.
13
Now, I will discuss some of the secondary
14
lipid endpoints.
Apo B, non-HDL cholesterol, and
15
total cholesterol are biomarkers of cardiovascular
16
risk that reflect LDL cholesterol as well as other
17
putative atherogenic lipoproteins such as VLDL.
18
previous guidelines, non-HDL was considered a
19
secondary target when triglycerides were high.
20
However, decreases in apo B, non-HDL, and total
21
cholesterol with alirocumab essentially reflect LDL
22
decreases and are therefore expected.
A Matter of Record (301) 890-4188
In
144
Regarding drug-mediated changes in
1 2
triglycerides and HDL, recent trials have not
3
supported a cardiovascular benefit, particularly
4
when non-statin lipid-altering drugs are added to a
5
statin.
6
These figures illustrate the placebo-
7
subtracted differences in percent changes in apo B,
8
non-HDL, and total cholesterol.
9
expected, mirror the LDL cholesterol results.
The results, as Four
10
of 5 trials demonstrated a significant effect on
11
HDL and 3 of 5 trials demonstrated a significant
12
effect on triglycerides.
13
are directionally favorable, as noted before, the
14
clinical significance is uncertain.
15
Although these changes
The active control trials, COMBO II,
16
OPTIONS I and II, ALTERNATIVE, and MONO evaluated
17
alirocumab versus ezetimibe.
18
and II, and ALTERNATIVE were conducted in patients
19
primarily at high or very high cardiovascular risk,
20
whereas the MONO trial was conducted in patients at
21
moderate cardiovascular risk.
22
COMBO II, OPTIONS I
COMBO II enrolled patients on a maximally
A Matter of Record (301) 890-4188
145
1
tolerated dose of statin, whereas, by design, the
2
OPTIONS trials specifically enrolled patients at a
3
moderate statin dose and included additional
4
comparator arms of statin up-titration or statin
5
switch.
6
patients who were not on background statin.
7
ALTERNATIVE trial, which I will discuss separately,
8
specifically screened and randomized patients who
9
are identified as statin intolerant.
The ALTERNATIVE and MONO trials enrolled The
Across all the active control trials with
10 11
and without add-on statin, alirocumab significantly
12
lowered LDL by 24 to 36 percent versus ezetimibe.
13
Note that alirocumab was not evaluated for LDL
14
lowering as monotherapy versus high-intensity
15
statin.
16
been conducted, the clinical implications of any
17
LDL differences between alirocumab and statins are
18
unknown.
19
However, even if such a comparison had
Although the OPTIONS trials demonstrated a
20
numeric 20 to 49 percent difference in LDL lowering
21
versus various regimens of statin up-titration or
22
switch to a higher-intensity statin, not all
A Matter of Record (301) 890-4188
146
1
comparisons were statistically significant
2
utilizing a multiple testing strategy.
3
Furthermore, there is evidence that
4
intensifying statin is an effective strategy to
5
reduce cardiovascular risk.
6
clinical implications of the differences in LDL
7
changes seen in these trials remains unclear.
8 9
Therefore, the
FDA conveyed these concerns about active control comparisons at the end-of-phase-2 meeting,
10
noting that while statistically significant and
11
clinically meaningful improvements in lipid
12
parameters may be allowed to be described in the
13
clinical trial section of the label, we would not
14
consider adding these data to the labeling until
15
the CVOT was completed and provided a very robust
16
assessment of the long-term safety and efficacy
17
profiles of alirocumab.
18
So now, I'll turn my attention to statin
19
intolerance.
It has been reported that many
20
patients stop taking statins because of side
21
effects, most commonly muscle related, although a
22
causal relationship is not often clear.
A Matter of Record (301) 890-4188
A recent
147
1
retrospective cohort study found that among
2
patients in a clinical setting who had a statin-
3
related event and discontinued statin, 59 percent
4
were rechallenged with a statin.
5
rechallenged patients, 92 percent were still taking
6
a statin 12 months later.
7
And of these
ALTERNATIVE was a trial designed to evaluate
8
the effect of alirocumab in patients with statin
9
intolerance.
We believe it was conducted in a
10
rigorous way and fairly addresses a number of key
11
issues in the statin-intolerant population beyond
12
the comparative LDL-lowering efficacy of alirocumab
13
and ezetimibe.
14
The patient population included those with
15
statin intolerance and moderate to very high
16
cardiovascular risk.
17
defined as the inability to tolerate at least two
18
statins, including one at the lowest-approved dose
19
due to skeletal muscle-related symptoms.
20
sponsor and FDA mutually agreed on this definition
21
for the trial.
22
Statin intolerance was
The
Eligible patients who are willing to be
A Matter of Record (301) 890-4188
148
1
rechallenged with atorvastatin entered a 4-week
2
single-blind placebo-run-in period.
3
who did not experience skeletal muscle-related
4
adverse events, other than those due to strain or
5
trauma during the run-in period, were eligible to
6
enter the double-blind treatment period.
7
were then randomized for 24 weeks into alirocumab,
8
ezetimibe, or atorvastatin groups.
Only patients
Patients
Of the 361 patients screened, 314 patients
9 10
completed the single-blind placebo-run-in period,
11
while 47 patients did not and were therefore not
12
randomized.
13
were the greatest source of run-in failure.
Skeletal muscle-related adverse events
Patients were randomized into the treatment
14 15
groups.
One patient randomized to ezetimibe was
16
never treated.
17
alirocumab group, 25 percent of patients each in
18
the ezetimibe and atorvastatin groups prematurely
19
discontinued during the treatment period due to
20
adverse events.
21
leading to discontinuation in all treatment groups
22
was musculoskeletal.
Eighteen percent of patients in the
The most common adverse event
A Matter of Record (301) 890-4188
149
Eighty-one percent of patients in the
1 2
alirocumab group and 70 percent of patients in the
3
ezetimibe as well as atorvastatin groups completed
4
22 weeks of drug therapy and attended the 24-week
5
visit.
6
Alirocumab's LDL-lowering efficacy was
7
similar in this trial as compared to the other
8
phase 3 trials and likewise compared favorably with
9
ezetimibe.
It was not formally tested for
10
statistical significance versus atorvastatin, but
11
the unadjusted changes from baseline are included
12
here for completeness.
13
Similar to other regulatory comments shared
14
with the sponsor during alirocumab development
15
mentioned previously, FDA noted that it would be a
16
review issue, whether we would include data from a
17
statin-intolerant trial in the labeling before the
18
cardiovascular outcomes trial was completed.
19
So why would FDA be so cautious about statin
20
intolerance?
After all, FDA acknowledges that
21
there are some patients unable to take statins,
22
which is of concern.
However, prevalence of true
A Matter of Record (301) 890-4188
150
1
pharmacological intolerance might be lower.
2
Notably, 70 percent of statin-intolerant patients
3
as defined for this trial in fact tolerated
4
20 milligrams of atorvastatin daily.
5
some concern that patients may consider abandoning
6
statins for a new drug that does not have
7
established long-term safety and cardiovascular
8
benefit.
FDA does have
Perhaps language that indicates use in
9 10
combination with maximally tolerated statin therapy
11
would acknowledge that statins remain first-line
12
therapy, but also recognize that for some patients,
13
maximally tolerated statin might mean no statin at
14
all.
15
Finally, I would like to briefly mention
16
neutralizing antibodies and potential loss of
17
efficacy.
18
1.2 percent of alirocumab-treated patients and no
19
patients treated with control.
20
neutralizing antibodies appear transient and were
21
not associated with loss of efficacy, where there
22
were confounding factors that challenged the
Neutralizing antibodies were observed in
A Matter of Record (301) 890-4188
Most events of
151
1
interpretation.
2
with neutralizing antibodies that appeared to be
3
associated temporally with transient or prolonged
4
loss of efficacy.
5
However, there were a few patients
In the setting of a persistent neutralizing
6
response, it's unclear if continuing therapy will
7
have any safety implications or if an LDL-lowering
8
effect with alirocumab can be restored over time.
9
I'm going to show a few examples of
10
increases in LDL, which is represented by the red
11
line, coincident with the neutralizing antibodies,
12
represented by the black squares.
13
patient 1, LDL lowering was initially seen with the
14
alirocumab treatment, and there was an increase to
15
baseline after the neutralizing activity was
16
documented.
17
In this figure,
By the end of the trial, neutralizing
18
activity was no longer observed, although LDL had
19
not returned to its nadir.
20
patient had continued on the drug longer or if the
21
dose had been increased, if efficacy could
22
ultimately have been restored.
It is unknown if the
A Matter of Record (301) 890-4188
152
Here is an example of LDL increasing above
1 2
baseline coincident with neutralizing antibodies.
3
LDL cholesterol for patient 2 in fact doubled from
4
73 milligrams per deciliter at baseline to 153 at
5
week 12.
6
confounded by non-compliance of either study drug
7
and/or background statin, particularly later in the
8
trial.
9
in this case cannot be determined with certainty,
10
However, this case was possibly
The contribution of neutralizing antibodies
but neither can it be dismissed. A third patient's profile demonstrates that
11 12
even in the face of high-titer neutralizing
13
antibodies, the effect on LDL efficacy was
14
transient.
15
Roberts, who will discuss her review of alirocumab
16
safety.
17 18
So now, let me introduce Dr. Mary
FDA Presentation –Mary Dunne Roberts DR. ROBERTS:
Good morning, Chairman Smith,
19
members of the committee.
My name is Mary Roberts.
20
I reviewed the safety profile of alirocumab and
21
plan to discuss the overall safety findings, and
22
specific adverse events, and topics of special
A Matter of Record (301) 890-4188
153
1
interest listed here.
2
risk associated with alirocumab treatment is formed
3
from an integrated summary of safety of 14 trials,
4
4 phase 2 trials and 10 phase 3 trials, which
5
includes 3,340 patients treated with alirocumab as
6
of the application cutoff date of August 31st,
7
2014.
8 9
The division's assessment of
The safety database was divided into two main safety pools based on the comparator used of
10
either placebo or ezetimibe.
11
events of interest, such as deaths and injection
12
site reactions, a global pool of all studies,
13
regardless of phase or control, was used to
14
evaluate these events.
15
events were evaluated in phase 3 trials as was an
16
exploratory analysis of glycemic shifts.
17
For some adverse
Adjudicated cardiovascular
Within the placebo-controlled pool, there
18
were nine trials, 4 phase 2 trials contributing
19
approximately 6 percent of alirocumab-exposed
20
patients and 5 phase 3 trials contributing the
21
remainder of patients.
22
patients, 81 percent were treated for at least one
Of 2,476 alirocumab-treated
A Matter of Record (301) 890-4188
154
1
year and 23 percent retreated for at least one and
2
a half years, yielding approximately 2800 patient-
3
years.
4
Due to the 2 to 1 randomization strategy
5
used in the phase 3 trials, the placebo-treated
6
patients contributed roughly 1400 patient-years to
7
this pool.
8
maximally tolerated statin therapy.
9
All patients within this pool were on
In the ezetimibe-controlled pool, there were
10
5 phase 3 trials of 864 alirocumab-treated
11
patients, yielding approximately 700 patient-years
12
and 618 ezetimibe-treated patients with 400
13
patient-years of exposure.
14
had a double-blind phase of 24 weeks, contributing
15
to the drop-off observed in the figure.
16
trials is 104 weeks in duration and provides the
17
majority of patients after 24 weeks of exposure.
18
The applicant has developed two doses of
Four of the 5 trials
One of the
19
alirocumab, 75 milligrams or 150 milligrams,
20
administered subcutaneously every two weeks for
21
marketing.
22
doses have been pooled.
For the safety analysis, the alirocumab Support for this approach
A Matter of Record (301) 890-4188
155
1
is based on review of exploratory analyses in
2
phase 3 trials.
3
adverse events in patients remaining on
4
75 milligrams were up-titrated to 150 milligrams
5
after week 12 was conducted.
6
for adverse events after week 12 was similar
7
between these groups, including no increase in
8
injection site reactions at a higher dose.
9
In these trials, an evaluation of
The overall profile
We acknowledge that these are post-
10
randomization comparisons.
11
direct parallel comparison of the 75-milligram and
12
150-milligram doses within a phase 3 trial to
13
evaluate a dose response.
14
a review of the exploratory analyses of phase 3
15
trials did not identify consistent dose-dependent
16
relationship with incidence of adverse events that
17
would prohibit pooling of the doses for the safety
18
analyses.
19
However, there is no
Given these limitations,
Treatment groups within the placebo-
20
controlled pool, alirocumab versus placebo, an
21
ezetimibe-controlled pool, alirocumab versus
22
ezetimibe, were relatively well matched for
A Matter of Record (301) 890-4188
156
1
demographics and baseline characteristics.
2
placebo-controlled pool, the mean age was 59 years,
3
60 percent were men, 90 percent were Caucasian, and
4
30 percent participated at sites within the United
5
States.
6
age was 62.
7
87 percent Caucasian, and 50 percent participated
8
at U.S. sites.
9
In the
In the ezetimibe-controlled pool, the mean Approximately 65 percent were men,
The majority of patients in both the phase 3
10
placebo-controlled and ezetimibe-controlled pools
11
had a history of cardiovascular disease or
12
cardiovascular risk factors.
13
had coronary heart disease.
14
of patients reported a previous myocardial
15
infarction.
16
Sixty to 70 percent Approximately a third
In both of the main safety pools,
17
approximately 70 percent reported a history of
18
hypertension and an estimated 30 percent reported a
19
history of diabetes.
20
patients within the placebo-controlled pool were on
21
a background statin, with 54 percent on a high-
22
intensity statin and approximately a quarter of
At randomization, almost all
A Matter of Record (301) 890-4188
157
1
patients were taking ezetimibe. In comparison, more patients, 20 to 27
2 3
percent, were not receiving background statin
4
therapy and fewer were on a high-intensity statin
5
in the ezetimibe-controlled pool, which reflects
6
the inclusion criteria and objectives of particular
7
studies within this pool.
8
This table provides a summary of the
9
treatment-emergent events, which were defined as an
10
event occurring within 70 days of the last study
11
injection in the placebo- and ezetimibe-controlled
12
pools.
13
reported adverse event, serious adverse event, or
14
discontinuation due to an adverse event was similar
15
in the alirocumab versus the comparator groups.
16
The percentage of patients with any
As of the application cutoff date, there
17
were a total of 37 on-study deaths, 17 in the
18
control group and 20 deaths in the alirocumab
19
group, which occurred after the start of treatment
20
and before the last planned protocol visit.
21
were no deaths that occurred in the phase 1 or
22
phase 2 trials.
A Matter of Record (301) 890-4188
There
158
1
The majority of deaths occurring in the
2
phase 3 trials were adjudicated as cardiovascular,
3
which is expected in a population at high
4
cardiovascular risk.
5
few fatal events to determine the effect of
6
alirocumab on mortality.
7
Importantly, there are too
A similar proportion or lower proportion of
8
alirocumab-treated patients permanently
9
discontinued treatment due to a treatment-emergent
10
adverse event compared to the comparator-treated
11
patients.
12
leading to discontinuation that occurred in at
13
least two or three patients and with greater
14
incidence in the alirocumab-treated group.
15
The tables below list the adverse events
Based on theoretical or identified concerns
16
about PCSK9 inhibition, or therapeutic protein
17
products in general, or alirocumab specifically,
18
several adverse events of special interest were
19
prespecified for potential additional monitoring
20
and reporting requirements.
21
these adverse events, the applicant utilized
22
prespecified standardized MedDRA queries, SMQs, or
In order to evaluate
A Matter of Record (301) 890-4188
159
1
company MedDRA queries, CMQs, which were developed
2
when no appropriate SMQ was available.
3
SMQs are groupings of MedDRA terms usually
4
at the preferred term level that relate to a
5
defined medical condition or area of interest.
6
Adverse events of special interest to be discussed
7
are listed here.
8 9
The first adverse event of special interest is diabetes-related events.
The role of PCSK9 and
10
the LDL receptor in glucose homeostasis is
11
developing as more data is accrued from nonclinical
12
and clinical investigations.
13
To date, there has not been a consistent
14
metabolic profile characterized in animal or human
15
studies of PCSK9 variance.
16
supportive data in the literature suggesting that
17
the LDL receptor may play a role in the risk of
18
developing type 2 diabetes, possibly through
19
increased cholesterol transfer into pancreatic beta
20
cells, which may adversely impact its function in
21
glycemic control.
22
LDL receptors, this theoretical concern was
However, there is some
As the PCSK9 inhibitors increase
A Matter of Record (301) 890-4188
160
1
explored within the LDL-lowering development
2
program. Nonclinical studies of alirocumab did not
3 4
show an effect on plasma glucose, or pancreas
5
structure or function in monkeys and rats.
6
phase 3 placebo and ezetimibe-controlled pools,
7
mean change from baseline in fasting glucose and
8
hemoglobin A1C over time did not demonstrate
9
meaningful differences between treatment groups
10
In the
overall or by baseline glycemic status. In addition, when evaluating patients
11 12
without diabetes at baseline, the incidence of
13
diabetes-related adverse events were comparable
14
between treatment groups.
15
measures of central tendency, and laboratory
16
values, and adverse events independently may not
17
convey clinically significant changes in glycemic
18
status.
19
planned exploratory analysis evaluating shifts in
20
glycemic status in phase 3 placebo in ezetimibe
21
trials.
22
However, looking at
Therefore, the applicant performed a
The following table describes the baseline
A Matter of Record (301) 890-4188
161
1
glycemic control categories as defined by the
2
applicant in the phase 3 placebo-controlled trials
3
and ezetimibe-controlled trials involved in this
4
analysis.
5
either impaired glucose control or diabetes at
6
baseline as per medical history and/or laboratory
7
values.
8
similar between treatment groups.
9
Approximately 70 percent of patients had
Distribution of glycemic categories was
This table shows the shifts and the
10
applicant-defined glycemic control categories in
11
patients on background statin therapy with normal
12
or impaired glucose control at baseline in the
13
phase 3 placebo-controlled pool.
14
A slightly higher proportion of alirocumab-
15
treated patients, 31.2 percent, compared with
16
26.6 percent of placebo-controlled-treated patients
17
with normal glucose control at baseline, met the
18
criteria for impaired glucose control at least once
19
during the treatment period.
20
observed in patients in the impaired glucose
21
category at baseline; 5.7 percent of the
22
alirocumab-treated versus 3.8 percent of the
This pattern was also
A Matter of Record (301) 890-4188
162
1
placebo-treated patients met the criteria for
2
diabetes.
3
An anti-diabetic medication was initiated in
4
6 of the 49 alirocumab-treated patients, who
5
shifted to the diabetes category in the alirocumab
6
group versus none in the placebo group.
7
Conversely, 18 percent and 21 percent of patients
8
with impaired glucose control at baseline and the
9
placebo and alirocumab groups respectively shifted
10
to the normal glucose control category during the
11
treatment period.
12
This table displays the glycemic shifts
13
present in the ezetimibe-controlled pool.
14
slightly higher proportion of patients treated with
15
alirocumab shifted to a worse glycemic control
16
category at least once during the treatment period.
17
However, a similar proportion of patients in both
18
treatment groups shifted to the normal glucose
19
category from the impaired glucose control category
20
at baseline.
21 22
A
In summary, there were no meaningful changes in mean fasting plasma glucose or hemoglobin A1C
A Matter of Record (301) 890-4188
163
1
with alirocumab treatment.
There was a slightly
2
greater proportion of alirocumab-treated patients
3
that met the criteria for worsening glucose control
4
by adverse event or lab. However, many patients had values at
5 6
baseline that were closed to the category
7
thresholds, so small changes could lead to a
8
category change.
9
not account for changes in medications that could
10
In addition, this analysis does
influence the results. It is unknown at this time if the observed
11 12
shifts in glycemic control categories represent a
13
true risk with alirocumab treatment.
14
majority of patients, glucose control remained
15
stable and serious diabetes adverse events were
16
few.
17
are monitorable and treatable.
As for the
Furthermore, changes in glucose homeostasis
Alirocumab was administered subcutaneously
18 19
in the abdomen, thigh, or outer area of the upper
20
arm.
21
and recorded using specific case report forms or
22
high-level terms.
Local injection site reactions were monitored
A Matter of Record (301) 890-4188
164
1
In the global pool, higher incidences of
2
local injection site reactions were reported in
3
patients receiving alirocumab injection versus a
4
placebo injection.
5
were transient at a mild intensity and few patients
6
discontinued treatment due to an injection site
7
reaction.
8 9
Most injection site reactions
However, patients receiving the alirocumab injection reported a greater number of injection
10
site reactions with a longer average duration
11
compared to patients receiving injections of
12
placebo.
13
with treatment-emergent anti-drug antibodies
14
reported a higher incidence of local injection site
15
reactions, 10.2 percent compared to antibody-
16
negative patients, 5.9 percent.
In alirocumab-treated patients, those
17
Since alirocumab is a monoclonal antibody,
18
allergic events were evaluated as an adverse event
19
of special interest using a company MedDRA query
20
for hypersensitivity, which includes terms such as
21
rash, angioedema, drug hypersensitivity, and face
22
swelling but excluded terms associated with
A Matter of Record (301) 890-4188
165
1
possible local injection site reactions such as
2
injection site rash and injection urticaria.
3
A higher proportion of alirocumab-treated
4
patients reported an allergic event.
5
rash were the top two reported adverse events in
6
both safety pools and occurred in approximately
7
1 percent of alirocumab-treated patients.
8 9
Pruritus and
Serious adverse events were similar in proportion and none were fatal.
In the placebo
10
pool, the events were primarily related to asthma
11
exacerbation in both treatment groups, although
12
there were single alirocumab-treated patients with
13
serious events of hypersensitivity, allergic
14
dermatitis, and eczema.
15
There are also two cases reported as
16
anaphylactic reaction.
One case occurred in a
17
placebo-treated patient and one in an alirocumab-
18
treated patient reported within the 120-day safety
19
update and treated at a higher dose than what is
20
currently proposed.
21
history significant for allergy.
22
multiple doses of study drug prior to the event.
Both patients had a medical
A Matter of Record (301) 890-4188
Both had received
166
1 2
Neither patient exhibited anti-drug antibodies. In the placebo-treated patient, the event
3
occurred within 2 hours after the last injection
4
and, in the alirocumab-treated patient, 11 days
5
after the injection.
6
and the alirocumab-treated patient was actually
7
rechallenged with alirocumab without recurrence of
8
signs or symptoms of anaphylaxis.
9
of events occurring with higher incidence in the
Both patients were covered,
Other examples
10
alirocumab group included hypersensitivity,
11
vasculitis, and hypersensitivity.
12
The next adverse event of interest is
13
immunogenicity associated with alirocumab treatment
14
and any clinical impact the development of anti-
15
drug antibodies may have on treatment-emergent
16
adverse events.
17
Treatment-emergent anti-drug antibodies were
18
detected in 5 percent of patients treated with
19
alirocumab in the phase 3 studies and less than
20
1 percent in the control group.
21
a transient antibody response.
22
onset was 12 weeks.
Most patients had The median time to
Neutralizing antibodies were
A Matter of Record (301) 890-4188
167
1
observed in approximately 1 percent of alirocumab-
2
treated patients.
3
Adverse events evaluated by the presence or
4
absence of anti-drug antibodies were similar
5
between groups and did not suggest a specific
6
pattern of adverse events with the exception of the
7
previously mentioned local injection site
8
reactions.
9
The next adverse event of interested focuses
10
on neurologic events.
11
material surrounding neuronal axons, is unique to
12
other cellular membranes because of its high lipid
13
to protein ratio.
14
80 percent lipids, of which 20 to 30 percent is
15
cholesterol.
16
derived by de novo synthesis as the blood-brain
17
barrier limits peripheral cholesterol transfer.
18
Myelin, the insulating
Myelin is approximately 70 to
Within the brain, cholesterol is
Neurologic events related to myelin sheath
19
disorders or neuropathies were collected based on
20
theoretical concerns that low LDL levels may impair
21
myelination.
22
company MedDRA queries used in this evaluation of
There were three standardized or
A Matter of Record (301) 890-4188
168
1
neurologic events, defined as demyelination,
2
peripheral neuropathy, and Guillain-Barre syndrome.
3
Within the placebo pool, the incidence of
4
patients with a neurologic event of special
5
interest was similar.
6
higher incidence of alirocumab-treated patients
7
reporting a neurologic event compared to ezetimibe-
8
treated patients, there were no specific preferred
9
terms that showed a large imbalance.
While there was a slightly
10
Paresthesia was the only preferred term
11
reported with a higher incidence in alirocumab-
12
treated patients compared to placebo- and
13
ezetimibe-treated patients.
14
in the alirocumab group were not serious and did
15
not lead to treatment discontinuation.
16
These events occurring
The number of patients reporting a serious
17
neurologic event was small.
However, 4 alirocumab-
18
treated patients reported notable serious
19
treatment-emergent neuropathic conditions, which
20
are listed here:
21
review of the narrative was suspicious for multiple
22
sclerosis, occurring in a 57-year-old woman with
a case of demyelination, which on
A Matter of Record (301) 890-4188
169
1
symptoms noted on study day 64; a case reported as
2
Miller-Fisher syndrome, which is a variant of
3
Guillain-Barre syndrome, characterized by ataxia,
4
areflexia, and ophthalmoplegia occurring in a
5
47-year-old man preceded by GI symptoms of diarrhea
6
and nausea.
7
This patient had a transiently treatment-
8
emergent antibody response at day 29, several
9
months before the event, and was negative at all
10
other time points.
11
LDL of 1.5 milligrams per deciliter several weeks
12
prior to the event.
13
He also exhibited a very low
The third case was reported as optic
14
neuritis, which occurred in a 66-year-old man with
15
a history of vasculitis and hypothyroidism, with
16
preexisting anti-drug antibodies, which were
17
transiently neutralizing.
18
by a neuroophthalmologist and was determined to be
19
more consistent with an inflammatory condition that
20
is distinct from a demyelinating optic neuritis;
21
and last, a case of transverse myelitis occurring
22
in a 75-year-old woman with a history of
This case was evaluated
A Matter of Record (301) 890-4188
170
1
hypothyroidism, with symptoms occurring after two
2
months of treatment.
3
In summary, the incidence of neurologic
4
adverse events of special interest overall do not
5
suggest an association with alirocumab.
6
there were rare events of neuropathic conditions
7
occurring in this patient population.
However,
8
Therefore, while there appears to be no
9
unifying mechanism, these events are too few in
10
number to definitively rule out an association with
11
alirocumab treatment.
12
Neurocognitive events has been a topic of
13
interest with lipid-lowering drugs, particularly
14
statins, for several years.
15
safety labeling change regarding reports of memory
16
loss or confusion associated with statin use that
17
were generally non-serious and resolved with
18
discontinuation of statins.
19
In 2012, FDA issued a
Because of the concern that low LDL may
20
adversely affect neurologic function,
21
neurocognitive events were evaluated using a CMQ,
22
which included deliria, cognitive and attention
A Matter of Record (301) 890-4188
171
1
disorders, and disturbances, dementia, and amnestic
2
conditions, disturbances in thinking and
3
perception, and mental impairment disorders.
4
A second query using neurocognitive terms
5
requested by the FDA was also conducted.
As can be
6
seen in the table, the percentage of patients with
7
a treatment-emergent or serious adverse event was
8
small and similar among treatment groups in both
9
the placebo pool, in which all patients were on
10
background statin therapy, and ezetimibe pool.
11
alirocumab-treated patient discontinued to a
12
neurocognitive event.
13
were observed with the FDA-defined query of
14
neurocognitive events.
15
No
No significant differences
The preferred term "memory impairment"
16
occurred with high incidence in alirocumab versus
17
placebo- or ezetimibe-treated patients, with 8
18
alirocumab patients reporting an event versus 1
19
placebo patient.
20
serious.
21
stroke-related event, and one patient had a prior
22
history.
None of these events were
Two occurred in association with a
A Matter of Record (301) 890-4188
172
1
In total, there were 29 neurocognitive
2
events occurring in the alirocumab group, which is
3
too few to determine an association.
4
had potential alternative etiologies that could
5
also affect cognitive function.
6
Many events
Since the brain is not dependent on uptake
7
of peripheral cholesterol and alirocumab as a large
8
molecule is unlikely to cross the blood-brain
9
barrier, the concern for an alirocumab-induced
10
effect on neurocognition seems less plausible,
11
especially since the main safety pools did not
12
suggest a consistent safety signal.
13
Adverse hepatic events and hepatic-related
14
laboratories were collected primarily to evaluate
15
alirocumab's effect on the hepatobiliary system,
16
but also in part because all patients in the
17
placebo-controlled pool and a majority of patients
18
in the ezetimibe pool were on statins, which are
19
associated with elevations in liver transaminases,
20
but have a very low risk of serious liver injury.
21 22
Review of the adverse events reported within the hepatic disorder SMQ noted that most reported
A Matter of Record (301) 890-4188
173
1
as elevations in liver transaminases.
2
incidence of alirocumab-treated patients reported a
3
serious hepatic event.
4
alirocumab-treated patients versus 1 placebo-
5
treated patient reported a serious adverse event.
6
Further review of these cases listed here for
7
causality found alternative etiologies or possible
8
confounders.
9
A higher
In the placebo pool, 8
There were 3 cases, 2 in the placebo group
10
and one in the alirocumab group, that met the
11
laboratory values needed to satisfy Hy's law, which
12
is an indicator of drug-induced liver injury.
13
However, none of the cases met the final
14
requirement, which was the absence of an
15
alternative explanation for the laboratory
16
abnormalities.
17
etiologies, and therefore there were no Hy's law
18
cases observed in the clinical program.
19
All 3 cases had alternative
Elevations in liver enzymes were present and
20
occurred in a slightly higher proportion of
21
patients in the alirocumab group than control
22
groups in some of the categories, particularly in
A Matter of Record (301) 890-4188
174
1
the ezetimibe-controlled pool.
While review of the
2
cases for causality uncovered possible confounders,
3
such as obesity, alcohol use, and concomitant
4
medications, a relationship with alirocumab
5
treatment cannot be definitively excluded.
6
Musculoskeletal adverse events are
7
associated with statin use and were monitored and
8
collected in the alirocumab clinical program.
9
Since the ezetimibe pool includes the ALTERNATIVE
10
trial of patients considered statin intolerant due
11
to muscle-related symptoms and was discussed
12
previously by Dr. Golden, my presentation focuses
13
on the placebo-controlled pool of patients.
14
be seen, there were similar proportions of patients
15
reporting musculoskeletal-related adverse events
16
and laboratory values.
17
As can
There were 3 cases of alirocumab-treated
18
patients reporting serious adverse events,
19
rhabdomyolysis and myositis with elevated
20
creatinine kinase.
21
not suggest a convincing association with
22
alirocumab treatment and these events.
Review of the narratives did
A Matter of Record (301) 890-4188
175
1
The final group of adverse events of special
2
interests are cardiovascular events.
3
shows investigator-reported treatment-emergent and
4
serious adverse events in the original application
5
within the cardiac disorders system organ class.
6
Within the serious adverse events, the category
7
defined as ischemic coronary artery disorders had
8
the highest percentage of patients reporting an
9
event in the placebo- and ezetimibe-controlled
10 11
This table
safety pools. This slide provides more detail regarding
12
the investigator-reported serious adverse events
13
that populated the ischemic coronary artery
14
disorders group.
15
percentage of patients reporting an event within
16
this high-level term were similar.
17
ezetimibe pool, a slightly higher percentage of
18
alirocumab-treated patients reported an event.
19
In the placebo pool, the
In the
Review of the serious adverse events within
20
both pools demonstrated the greatest difference
21
between treatment groups was unstable angina.
22
the placebo pool, there were 25 patients or
A Matter of Record (301) 890-4188
In
176
1
1 percent of the alirocumab-treated group versus 9
2
patients or 0.7 percent of the placebo-treated
3
group reporting an event. In the ezetimibe pool, 1.4 percent of
4 5
alirocumab-treated patients versus 0.3 percent of
6
ezetimibe-treated patients reported unstable
7
angina.
8
hospitalization along with other suspected
9
cardiovascular events and deaths in the phase 3
Events of unstable angina requiring
10
trials that occurred from the time of randomization
11
until the follow-up visit were adjudicated.
12
This table displays the positively
13
adjudicated treatment-emergent cardiovascular
14
events in the phase 3 trials at the time of the BLA
15
submission.
16
control-treated patients versus 1.6 percent of
17
alirocumab-treated patients had a MACE event
18
defined as coronary heart disease, death, nonfatal
19
MI, fatal or nonfatal ischemic stroke, and unstable
20
angina requiring hospitalization.
21 22
Following adjudication, 1.8 percent of
Note, three serious events of unstable angina met the stringent criteria necessary for
A Matter of Record (301) 890-4188
177
1
inclusion in the MACE composite endpoint.
2
Importantly, the number of events accrued is
3
considered insufficient to determine alirocumab's
4
effect on cardiovascular morbidity or mortality. The last topic for discussion centers on the
5 6
safety of very low LDL levels.
In the alirocumab-
7
treated group, almost a quarter of patients had two
8
consecutive LDL values less than 25 milligrams per
9
deciliter, and roughly 9 percent had two
10
consecutive LDL levels less than 15 milligrams per
11
deciliter. To explore whether drug-induced low LDL
12 13
could be associated with adverse events, analyses
14
were performed by achieved LDL subgroup within the
15
alirocumab arm.
16
columns are included in the table for reference
17
only.
18
The control and total alirocumab
A comparison of adverse events in the
19
alirocumab low LDL group defined as achieving two
20
consecutive LDL values less than 25 versus the
21
alirocumab group that did not satisfy this criteria
22
is shown in the far two right columns of this
A Matter of Record (301) 890-4188
178
1 2
table. Only events that occurred after the first
3
LDL less than 25 were considered in the low LDL
4
group, since the objective was to evaluate the
5
effect of persistently very low LDL levels and
6
occurrence of adverse events.
7
fact that exposure time may be systematically less
8
for the low LDL group, we present incidence rates.
9
To account for the
Overall, there does not appear to be a
10
strong association with low LDL and incidence of
11
events, with the possible exception of diabetes and
12
cataracts.
13
rates in patients that achieved very low LDL at 2
14
consecutive visits compared to those that did not
15
has to be interpreted with caution, because this is
16
a within-group comparison and the two groups may
17
not be representative of each other.
18
be other factors that could possibly impact the
19
results.
20
However, the comparison of incidence
There could
At this time, no strong signal to suggest an
21
association with very low levels of LDL was
22
observed with alirocumab treatment and adverse
A Matter of Record (301) 890-4188
179
1
events.
2
levels of LDL with PCSK9 inhibition is unknown.
3
Dr. Julie Golden will now summarize the division's
4
benefit/risk assessment for alirocumab.
5 6
The effect of chronic exposure to very low
FDA Presentation – Julie Golden DR. GOLDEN:
Regarding efficacy, we agree
7
with the sponsor that there is substantial and
8
persistent LDL cholesterol lowering in all patient
9
populations studied in this program.
We consider
10
any conclusions regarding cardiovascular benefit to
11
be premature.
12
Regarding safety, given the potential for
13
widespread use, even small observed differences in
14
serious safety signals, if true, could have
15
significant public health implications.
16
Overall, based on the data so far, many of
17
the safety concerns raised appear monitorable
18
and/or manageable.
19
identified so far include a higher incidence of
20
shift to worse glycemic category, hypersensitivity
21
reactions, including allergic and injection site
22
reactions.
Safety issues that have been
Examples of the few serious events
A Matter of Record (301) 890-4188
180
1
observed included anaphylaxis and hypersensitivity
2
vasculitis.
3
liver enzyme abnormalities observed.
There is also a higher incidence of
Other very rare events, such as the serious
4 5
neurologic events discussed, were of unclear
6
causality, but cannot be dismissed.
7
strong safety signal for very low LDL was observed,
8
but the analyses have limitations.
Finally, no
In summary, despite treatment with maximal
9 10
standard-of-care therapy, many patients remain at
11
substantial cardiovascular risk.
12
20 years, FDA has accepted reduction in LDL
13
cholesterol as a surrogate for cardiovascular risk
14
reduction to support lipid-lowering drug approval.
For over
In the past, cardiovascular outcomes trials
15 16
were conducted post-approval and were not required
17
by FDA.
18
body of cardiovascular outcomes and safety data,
19
whereas recent cardiovascular outcomes trials for
20
non-statin lipid drugs have had mixed results.
21 22
As we have discussed, statins have a large
Importantly, alirocumab has the potential for widespread use.
So our question to you today
A Matter of Record (301) 890-4188
181
1
is, do you believe the applicant has sufficiently
2
established that the LDL cholesterol lowering
3
benefit of alirocumab exceeds its risks to support
4
approval in one or more patient populations? Finally, Mary and I would like to
5 6
acknowledge our colleagues, who provided
7
considerable support for this presentation.
8
you.
9 10
Thank
Clarifying Questions DR. SMITH:
Thank you.
So we now have some
11
time for clarifying questions.
What I'd like to
12
emphasize is, first of all, we'll take questions
13
related to the FDA presentation.
14
to sponsor questions a little later.
15
to stress that what we'd like to focus on now are
16
clarifying questions and not just discussion,
17
because we'll have ample time later to get to
18
issues of discussion.
We'll come back And I'd like
19
So focus on clarifying questions.
Again,
20
please state your name when you ask your question
21
for the sake of the record.
22
DR. BURMAN:
And Dr. Burman, yes?
Thank you.
A Matter of Record (301) 890-4188
Ken Burman.
I just
182
1
wanted to ask, where else is PCSK9 made besides the
2
liver?
3
DR. ELMORE:
Lee Elmore, pharm tox reviewer
4
for DMEP for FDA.
PCSK9 is expressed in high
5
levels in the liver, kidney, in the brain, and
6
intestine.
7
adrenal.
Those are the major areas, and also the
8
DR. SMITH:
Dr. Wilson?
9
DR. WILSON:
Thanks very much for the
10
hepatic data for newer drugs.
11
interested in the potential.
12
slide 66, it would be great to know something about
13
that in relationship to glycemic status.
14
recognize the percentages are very small, though.
15
We're very So for FDA's
Do you have any information?
I
Have you
16
considered looking at this in terms of whether the
17
ALT/AST changes differ according to diabetes,
18
impaired glucose tolerance?
19
DR. ROBERTS:
Sorry.
We haven't done that
20
analysis.
21
like that, if they want to share that.
22
I don't know if the sponsor has anything
DR. EDELBERG:
We have not done that
A Matter of Record (301) 890-4188
183
1
analysis there, of the relationship there.
2
DR. SMITH:
Dr. Blaha?
3
DR. BLAHA:
I'm following up on questions
4
before about CRP and inflammatory access.
5
known about the relationship between PCSK9
6
inhibition and then the inflammatory pathways in
7
particular?
8
inflammatory-related -- was there any analysis of
9
any inflammatory-related adverse events that might
10 11
What is
And also, what is known about any
be resulting from PCSK9 inhibition? DR. ROBERTS:
We did ask the sponsor to look
12
at the CRPs by shifts.
13
have a slide of that available.
14
the sponsor has one.
15
significant shift with patients that might have
16
been less than 1 milligrams per liter at baseline
17
shifting up to greater than 3 milligrams per liter
18
for the CRP.
19
difference.
20 21 22
And as I recall, I don't I don't know if
But I don't recall seeing a
But I did not see a significant
DR. SMITH:
Does the sponsor have a comment
on that? DR. EDELBERG:
Dr. Sasiela will present
A Matter of Record (301) 890-4188
184
1 2
these data. DR. SASIELA:
Yes.
So we did look at a
3
shift analysis as Dr. Roberts mentioned.
4
trying to see if we can get that slide pulled up.
5
Basically -- and I'm trying to see how much I
6
remember from memory, but I think, for the most part
7
we didn't see any real pattern in shifts in terms of
8
where patients start off at baseline.
9
And I'm
The one thing I can show you right now is
10
those who were above and below 2 milligrams per
11
deciliter at baseline, and you can see it in this
12
slide here.
13
data in three separate groups, with a cut of subjects
14
who had baseline CRP either greater than 2 milligrams
15
per liter or those who were less than 2 milligrams
16
per liter.
17
This is grouping the studies and the
You'll see the four groups of studies are
18
alirocumab, 150, versus placebo, 75 milligrams, the
19
75-milligram dose versus placebo, 75, 150 versus
20
ezetimibe with statins, and 75, 150 without statins.
21
And as you can see, there's no real difference
22
whether patients were above or below 2 milligrams per
A Matter of Record (301) 890-4188
185
1
liter in terms of the overall responses compared to
2
placebo.
3
ezetimibe.
4
with the alirocumab group.
There's a slightly greater shift with But otherwise, there's no real change
DR. SMITH:
5
So I think that's good
6
information.
7
we could come back to that later.
8
for another slide?
9
afternoon, even if that comes up.
10
If you do find additional data, maybe Are you looking
We could look at that this Just let us
know.
11
Dr. Budnitz?
12
CAPT BUDNITZ:
Dan Budnitz.
I have a
13
question about slides 71 and 72 of the FDA
14
presentation, about the cardiac disorders and
15
adverse events, and how there are over 40 reports
16
of unstable angina based on the SAEs, but after
17
adjudication, the number dropped considerably to 3.
18
I'm just wondering, is that unusual, unexpected, or
19
concerning?
20
Or if it takes adjudication, is that
21
specific to unstable angina or is this something we
22
have to worry about for other SAEs that were
A Matter of Record (301) 890-4188
186
1 2
reported? DR. ROBERTS:
So we saw that as well, and we
3
asked the applicant to explain why there were so
4
many in the SAE column and then only 3 in the
5
adjudicated group.
6
of the definition for an adjudicated unstable
7
angina requiring hospitalization also showing
8
evidence of progressive myocardial ischemia.
9
Part of the reason is because
Then the other reason would be that some of
10
the events for unstable angina were also
11
adjudicated actually as myocardial infarction.
12
then the third reason would be that the MACE
13
analysis would be maybe a time-to-first event.
14
if you had had unstable angina after the first
15
event, then it would have been censored from that
16
analysis.
17
CAPT BUDNITZ:
And
So
So I think that makes sense,
18
but it also makes me wonder about some of these
19
other adverse events like neurologic events and
20
neurocognitive events.
21
of these terms might substantially change the
22
number of events and how that might be interpreted.
How you slice and dice some
A Matter of Record (301) 890-4188
187
1
DR. ROBERTS:
So for these adverse events of
2
special interest, most of them were used by a
3
standardized MedDRA query, so that's kind of a
4
separate medical dictionary that has a grouping of
5
events that kind of screen for that particular
6
condition.
7
committee to make sure that these preferred terms
8
are trying to capture the event of notice.
9
not always perfect, and that's why looking at the
And they're usually vetted through a
It's
10
actual narratives of the cases is so important, to
11
kind of tease that out as well.
12
CAPT BUDNITZ:
Thank you.
And that's why
13
maybe, with such small events, how you slice and
14
dice to either sensitive or specific definitions
15
for some of these terms, particularly when there's
16
small numbers of events, I think can make it
17
challenging to make the final conclusions.
18
DR. SMITH:
19
DR. ORZA:
Dr. Orza? My first question is about the
20
adverse events.
And we keep talking about the
21
placebo-controlled group, but it's not really a
22
placebo group because all of those people were also
A Matter of Record (301) 890-4188
188
1
getting statins.
And so we're reaching conclusions
2
that there aren't any safety signals here, but not
3
in the way that we usually do.
4
seem to be there in a lot of cases.
5
they're not very much different from what we see
6
with the statins.
The safety signals It's just that
So I'm wondering, because of what Dr. Hiatt
7 8
was saying earlier, it's different when we're
9
looking for no differences in safety signals versus
10
when we're looking for no difference in efficacy.
11
And we've powered the study to be able to detect
12
those.
13
Can you say a little bit about what power do
14
we have, actually, to detect whether there's any
15
difference in the safety signals between statins
16
and statins plus or ezetimibe plus?
17
DR. ROBERTS:
For the placebo pool,
18
everybody is on a background, a maximally tolerated
19
background statin, so when you do the comparison
20
looking at the placebo group, you're comparing a
21
group that is on maximally tolerated background
22
statin versus a group that's on maximally tolerated
A Matter of Record (301) 890-4188
189
1 2
background statin plus alirocumab. Within that group, they're high risk.
The
3
study designs within that group were more
4
consistent.
5
the primary safety, just because of those
6
considerations.
7
ezetimibe pool, you're right.
8
more variable, and not everybody was on background
9
statin therapy.
10
So that's why we've said that's more
And then, when you look at the Those are where it's
So you could say maybe that -- I mean, one
11
way to do it would be to look at if it's consistent
12
between the placebo pool and the ezetimibe pool.
13
And that was one of the things that I tried to look
14
at, to see if the primary safety database was
15
seeing a signal; was it also showing up in the
16
ezetimibe, given the limitations for both of the
17
safety pools.
18
But I agree, it's sometimes difficult to
19
know what is actually a safety signal and a true
20
risk when you're looking at very small differences
21
and incidents.
22
DR. ORZA:
How large is the group of
A Matter of Record (301) 890-4188
190
1
patients who had no other background drug?
2
you look at it?
4 5 6 7
Were you able to look at it?
DR. ROBERTS:
3
And did
So I'm sorry.
Just repeat
one. DR. ORZA:
They were getting only
alirocumab? DR. ROBERTS:
It was 100 patients, so 50
8
patients on alirocumab versus patients that were
9
not on any background statin therapy.
10
DR. GOLDEN:
11
DR. ROBERTS:
12 13 14 15
So that's the mono Right.
That's the monotherapy
trial, total 100 patients. DR. ORZA:
That's a very small group, but
was there any interesting signals there? DR. ROBERTS:
Again, it's limited because of
16
the size of the group.
I did look to see if there
17
were things that would come out that I wasn't
18
seeing in other places, and there wasn't anything
19
that I saw different.
20
at was changes in liver transaminases.
21
didn't see any differences between the group that
22
was on alirocumab versus the group that wasn't on
One of the things I did look
A Matter of Record (301) 890-4188
And I
191
1
any background statin therapy.
2
patients that had elevations in ALT.
3
DR. SMITH:
4
DR. EVERETT:
There were no
Dr. Everett? Thank you.
This is a
5
clarifying question for Dr. Roberts with the FDA.
6
I'm looking at your slide 58 here, which lists
7
notable neurologic events that occurred.
8
noticed that all the patients who are listed, or
9
the cases that are listed, were patients who were
And I
10
on active therapy.
11
neurologic events in the placebo, or statin only,
12
or ezetimibe arms?
13
DR. ROBERTS:
Were there no notable
So I think there were -- and I
14
have to go back and look.
15
document.
16
adverse neurologic events.
17
in that they're considered more rare.
18
There were, I think, a total of 7 serious adverse
19
events, and it was divided between the alirocumab
20
group and the placebo group, but that's something
21
in the background.
22
I have it in my briefing
So these are not the only serious
DR. EVERETT:
These are more notable So no.
So it would be helpful to see
A Matter of Record (301) 890-4188
192
1
this because just reading this slide and
2
seeing -- I'm not a neurologist, but these events
3
seem serious and grave to me.
4
at least associated with the use of or being
5
randomized to alirocumab.
And all of them are
So it would be nice to see if there were
6 7
similar rates of demyelination, for example, or
8
other important neurologic events in the placebo,
9
or statin only, or ezetimibe arms.
I don't want to
10
be left with an impression, in other words, that
11
all these very serious and grave events were only
12
in one particular treatment arm if that wasn't the
13
case.
14
just want to be clear. DR. ROBERTS:
15 16
point.
No, I agree it's a valid
Just give me a second. DR. EVERETT:
17 18
If it is the case, then that's okay, too.
I'm happy to circle back to
that later if someone else wants -DR. SMITH:
19
I think that would be a good
20
idea.
Just we'll come back to that later on.
21
for balance here, I'd like to pick up an
22
opportunity for people who had some questions
A Matter of Record (301) 890-4188
So
I
193
1
earlier this morning.
So I'm going to shift over
2
to those, and then I'll continue down the list of
3
questions.
4
an urgent comment or just another --
And Dr. Geller, you look like you have
5
DR. GELLER:
I have a question that crosses
6
both presentations.
7
34A in the addendum to the briefing document, I
8
don't understand how I can reconcile that with the
9
FDA slide on cardiac disorders.
When I look at figures 34 and
So I see a highly significant difference in
10 11
the addendum to the briefing document, but nothing
12
quite so strong in the left-hand panel, which is
13
pooled data.
14
please?
15 16
Can somebody help me with that,
DR. EVERETT:
You're talking about the
cardiovascular event rates or something --
17
DR. GELLER:
18
DR. EVERETT:
19
DR. GELLER:
Yes.
Yes.
-- and a Kaplan-Meier curve? Yes, the final figures in the
20
briefing addendum or Kaplan-Meier plots of MACE
21
from the sponsor.
22
DR. SMITH:
Yes.
If we could find that
A Matter of Record (301) 890-4188
194
1
figure, maybe that's the way to work our way
2
through this.
3
DR. GELLER:
They're from the sponsor,
4
whereas the cardiac disorders overall and SAEs show
5
much less difference.
6
has the same end date.
7
combined.
8
DR. EDELBERG:
And I think the second one One is one trial, one is
Very good.
What we're seeing
9
here is the MACE, the adjudicated cardiovascular
10
events for the LT, the long-term 18-month trial,
11
which was half of the overall exposure in the
12
program.
13
LDL of 120 versus the treatment of 50.
14
Dr. Braunstein actually presented, the overall
15
event rate there, this is the data that was
16
published from the final results of the trial that
17
was published recently in the New England Journal,
18
what was earlier part of the interim analysis that
19
was done.
20
These were the patients who had a mean
DR. GELLER:
Yes.
And as
But below that is the
21
interim analysis, which actually shows a more
22
significant result.
A Matter of Record (301) 890-4188
195
DR. EDELBERG:
1 2
hazard ratio of .46 and the final results were .52. DR. GELLER:
3 4
The interim analysis showed a
Yes.
But I'm having trouble
reconciling that with what the FDA just presented. DR. EDELBERG:
5
So what the FDA was
6
presenting, I believe, was, we started with the
7
unadjudicated events.
8
adjudication from the global pool, which included
9
not just these, but data from many of the shorter
And then it was the
10
trials.
11
versus the active comparison, ezetimibe, there.
12
This is including the trials that were
DR. GELLER:
So this result is in this trial
13
only, but overall, it sort of evens out?
14
what happens?
15 16 17
DR. EDELBERG:
Is that
So actually, I'm going to ask
Dr. Braunstein to review the overall CV data. DR. BRAUNSTEIN:
Can I have the overall MACE
18
from either the Kaplan-Meier curve or the -- so we
19
could show this.
20
the overall pool, with a hazard ratio of 0.81.
21
this is from all of the studies combined, including
22
the ezetimibe.
These are the MACE events from
So this is a combination of
A Matter of Record (301) 890-4188
So
196
1
placebo-controlled data and ezetimibe-controlled
2
data.
3
from our global phase 3 pool.
4
So it's all of the data in the application
DR. GELLER:
Okay.
But the FDA did drug
5
versus control and said there was an insufficient
6
number of events to determine alirocumab's effect
7
on cardiovascular morbidity and mortality, whereas
8
long-term seems to show there is an effect.
9
when you put it all together, it seems that the
10 11
So
effect goes away. DR. BRAUNSTEIN:
Even in the long-term
12
study, we're seeing -- this was a safety analysis
13
that we did in the long-term study.
14
positive trend, which was useful, but we would
15
agree that the data that we have at present is not
16
sufficient to assess whether or not there actually
17
is a benefit.
18
And we see a
We are presenting our cardiovascular data
19
here really to demonstrate the safety of the
20
product because, overall, we only have 85 MACE
21
events in an outcomes study.
22
cardiovascular outcomes trial will have 1600 MACE
For example, our
A Matter of Record (301) 890-4188
197
1
events that we will have for analysis.
And that's
2
more typical of a cardiovascular outcomes study. So the data that we have, we feel are
3 4
encouraging.
5
types of patients that we envisioned would be
6
receiving this product if it were to be approved,
7
patients with high unmet need, underlying
8
cardiovascular disease, or FH, elevated LDLs with a
9
baseline of 120 despite maximally tolerated
10
The long-term study is evaluating the
statins. So that one study, we think, is a
11 12
particularly interesting study to look at, but
13
nonetheless, we are not here telling you that that
14
study proves cardiovascular benefit.
15
our intention here today.
16
DR. SMITH:
17 18
That is not
You need to turn your microphone
on. DR. GELLER:
On the outcomes trial, there
19
are HeFH patients included in the outcomes trial as
20
well as high-risk CV population?
21
DR. EDELBERG:
22
DR. GELLER:
The outcomes trial -Because all you said was ACS.
A Matter of Record (301) 890-4188
198
1
DR. EDELBERG:
Right.
The patients who will
2
be enrolled in the outcomes trial will be patients
3
who recently experienced an acute coronary
4
syndrome, which will include patients who are at
5
high risk, patients who have FH, and patients who
6
actually are statin intolerant.
7
DR. BRAUNSTEIN:
So it will include all of
8
the patients that -- all of the categories of
9
patients that we're talking today, all with recent
10 11 12 13 14 15
acute coronary syndrome. DR. GELLER:
Is there stratification on
those categories? DR. EDELBERG:
No.
We were enrolling all
patients who meet the enrollment criteria. DR. BRAUNSTEIN:
With a study of that size,
16
18,000 patients, we're relying on randomization to
17
take care of that.
18
DR. SMITH:
19 20
Dr. Sager, did you have a
question or comment on this same topic? DR. SAGER:
Yes, just a quick question.
On
21
the long-term results showing the MACE Kaplan-Meier
22
curve, is that the standard MACE criteria of
A Matter of Record (301) 890-4188
199
1
cardiovascular death, nonfatal MI, or stroke?
2
DR. EDELBERG:
3
DR. SAGER:
4
DR. EDELBERG:
Yes.
Nothing else included? Those are the MACE terms that
5
we've included.
6
included in the outcomes trial. DR. GELLER:
7 8 9
These are the same MACE that are
What happened to unstable
angina? DR. EDELBERG:
This includes unstable angina
10
requiring hospitalization.
11
discussed before, of cases of unstable angina are
12
adjudicated positively as myocardial infarction.
13
This is a recent trend that we have observed with
14
the use of high-sensitivity troponin, so it was
15
expected.
16
DR. SAGER:
A majority, as was
I'm still confused.
So the
17
long-term Kaplan-Meier you showed us does not
18
include unstable angina and it was not adjudicated
19
as myocardial infarction?
20
DR. EDELBERG:
The majority of the unstable
21
angina cases that were adjudicated positively were
22
adjudicated positively as nonfatal MI.
A Matter of Record (301) 890-4188
200
DR. SAGER:
1
So it's not a standard MACE.
2
You also have unstable angina as the required
3
hospitalization that were not adjudicated as
4
myocardial infarctions. DR. EDELBERG:
5
Is that correct?
So four categories there were
6
CHD death, nonfatal MI, ischemic stroke, unstable
7
angina requiring hospitalization. DR. SAGER:
8 9 10 11
Thank you.
That's our MACE.
No, I just wanted to
be clear on exactly what we were talking about. Thanks. DR. BRAUNSTEIN:
But in our entire program,
12
there were only 3 patients with unstable angina
13
that were adjudicated as unstable angina.
14
fact, in the LT study, we're talking about a single
15
patient with unstable angina actually in the
16
placebo group.
17
actually all the events but one, meet that strict
18
MACE criteria, so point taken.
19
that it's not a big factor in the data we've shown.
20 21 22
And in
So the vast majority of the events,
DR. SAGER:
Thank you.
But it turns out
That's very unusual,
but thank you. DR. SMITH:
Critical follow-up on that?
A Matter of Record (301) 890-4188
201
1
Okay. DR. GELLER:
2
One last clarification.
Your
3
definition of MACE is consistent throughout your
4
trials?
5
DR. BRAUNSTEIN:
6
DR. GELLER:
7
DR. SMITH:
Yes.
Thanks. Not to surprise you, Dr. Nason,
8
but you had a question a couple hours ago.
And
9
again, to reemphasize, these are questions that may
10
be -- I'm including the mix of questions for FDA
11
and questions for the sponsor.
12
DR. NASON:
Thank you.
And so Dr. Nason? Martha Nason.
13
Actually, most of my questions have been answered
14
in the time.
15
to, though, is my questions about the statin-
16
intolerant folks and the ALTERNATIVE trial, and a
17
little bit following up on the questions Dr. Orza
18
asked about that.
19
The one thing I would like to go back
The one thing I guess I would still like to
20
understand from the sponsor is, you recognize that
21
it was a limited section of the population that was
22
willing to try statins again.
A Matter of Record (301) 890-4188
And I was wondering
202
1
if there was discussion of having a trial that
2
didn't include a statin arm for those folks and why
3
you decided not to do that, to simply
4
randomize -- allow them to be randomized to the
5
antibody, or placebo, or a different drug, but not
6
a statin. DR. EDELBERG:
7
Actually, in the discussions
8
with the FDA, it was agreed the importance of being
9
able to include a possible randomization of
10
patients with a history of statin intolerance to a
11
rechallenge there, all recognizing that we would
12
not be able to then enroll patients who would not
13
consent to being possibly randomized to a
14
rechallenge.
15
forms of statin intolerance couldn't come into the
16
trial.
17
So the patients with the most severe
What was very impactful about the results,
18
while many patients were able to stay on a statin
19
rechallenge, is that the muscle events were
20
significantly greater with the atorvastatin
21
rechallenge than with the treatment with alirocumab
22
over the course of the six-month trial.
A Matter of Record (301) 890-4188
203
1
When all patients were allowed to go into
2
the open-label extension, we've seen excellent
3
tolerability of patients who had muscle AEs, both
4
on atorvastatin.
5
have done very well in open label.
6
Ezetimibe as well as alirocumab
DR. NASON:
So I understand why you designed
7
the ALTERNATIVE trial as you did, and I think it
8
makes sense.
9
an additional group for the people who were not
I was asking about the possibility of
10
willing to randomize into ALTERNATIVE, to have
11
either another arm or another trial that was simply
12
in those people.
13
DR. EDELBERG:
We didn't conduct a trial in
14
those patients, nor did we have a follow-up period
15
for them, no.
16 17 18
We haven't done that.
DR. SMITH:
Dr. Everett, you had an earlier
question. DR. EVERETT:
Yes.
Thank you.
I want to come back.
Brendan
19
Everett.
And this question
20
is for Dr. Braunstein, really, I think, your slide
21
number 75, again, to think about these
22
neurocognitive events.
If we could call the slide
A Matter of Record (301) 890-4188
204
1
up, that would be great.
Thank you.
In the left-hand side of this, we see the LT
2 3
study, which is alirocumab or placebo on top of a
4
statin.
5
treatment-emergent adverse events, a higher
6
proportion, 1.2 percent versus 0.5 and 1.1 versus
7
0.8, and a hazard ratio that while above 1, clearly
8
has very wide confidence intervals.
9
And here's where we see a higher number of
I guess my question is, I was going to ask
10
you earlier to drill down on what these actually
11
were, but I think we've seen some of that with the
12
FDA presentation.
13
LT study to the other placebo-controlled pools and
14
that effect estimate goes towards 1?
15
So what happens when you add the
My question is really based on the fact that
16
the majority of the patient-years of exposure are
17
actually in the LT study.
18
800 people were added to the active arm and about
19
500 to the placebo arm.
20
go away?
21
a proportion versus an incidence rate is the right
22
way to examine these data?
And it looks like about
But why do those effects
And do you think that looking at strictly
And do you have
A Matter of Record (301) 890-4188
205
1
incidence rate data? DR. BRAUNSTEIN:
2
Perhaps the best way to
3
answer your question -- so let's start out with
4
what we have on the slide.
5
to everybody in the room, the LT study is included
6
in the placebo-controlled pool.
7
DR. EVERETT:
8
DR. BRAUNSTEIN:
9
Just to make it clear
Yes. Right.
So what that is
telling us is that the other studies, the other
10
long-term placebo-controlled studies -- because in
11
our placebo-controlled pool, we have five studies,
12
four of which are 18-month studies like the LT
13
study, one of which is a 12-month study.
14
studies at the time of the BLA had at least
15
completed their 12-month visit.
16
completed their 12-month visit.
And all
Every patient had
So they were all reasonably comparable in
17 18
terms of length.
So what that's telling us is in
19
fact that there is some study-by-study variability.
20
And in fact, I can just show you that.
21
think, the easiest way just to demonstrate that
22
point.
This is, we
And it's just a forest plot to show you the
A Matter of Record (301) 890-4188
206
1
LT study is the top line.
2
other studies there, they are in fact more to the
3
other side.
4
And then you see in
The single dots are 1 versus zero kind of
5
situations, so we couldn't calculate, actually,
6
more than just that.
7
DR. EVERETT:
That's helpful.
It's not the
8
incidence rate, but it does give us a sense of the
9
totality of the data, so I appreciate that.
10
don't have the incidence rate?
11
DR. BRAUNSTEIN:
12 13
You
The incidence rate per
study or per -DR. EVERETT:
No, it's number of events per
14
person time of observation as opposed to number of
15
events per person.
16
DR. BRAUNSTEIN:
17
we have that?
18
We will get you that. DR. EVERETT:
20
DR. SMITH:
22
afternoon.
Karen, do
We do have that per patient-year.
19
21
You know what?
We have that. Fine, thanks.
Let's do that later this
That's great.
Dr. Shamburek, do you still have a question?
A Matter of Record (301) 890-4188
207
DR. SHAMBUREK:
1
Yes.
This is a question for
2
the sponsor or the FDA really targeting the
3
populations that this will be used on.
4
indication, it's for primary hypercholesterolemia,
5
non-familial and heterozygous familial.
6
we have or mixed dyslipidemia, including type 2
7
diabetes, to reduce LDL and the rest of the
8
markers.
And in the
And then
My question really is -- and we don't want
9 10
to get into a guideline thing where LDL is, but
11
there are a lot of dyslipidemic and mixed
12
hyperlipidemic in our clinics with cardiovascular
13
disease.
14
believe the HDLs are about 50 and triglycerides are
15
150.
16
patients with high LDL.
And I look at the baseline lipids, and I
And we certainly with the guidelines go after
The question is maybe more for the typical
17 18
patient with the low HDL and the higher
19
triglyceride, do we have enough data here, or is
20
that what we're calling mixed dyslipidemia in these
21
groups?
22
8 percent and the triglyceride is lowering.
Because the effect on HDL might be up to
A Matter of Record (301) 890-4188
But is
208
1
this our typical mixed dyslipidemia, or do we have
2
enough from the clinical trials that are shown
3
here?
4
DR. EDELBERG:
So we enrolled the patients
5
that we expect will be using alirocumab and will
6
get the greatest benefit from alirocumab.
7
going to ask Dr. Sasiela to review our experience
8
with the different patients there based on HDL and
9
other criteria.
10
DR. SASIELA:
Right.
I'm
We defined mixed
11
dyslipidemia in this case as patients who came into
12
the trial with baseline triglycerides above 150.
13
So, given that we had a population with a decent
14
amount of obesity, we actually had a fairly good
15
representation.
16
who would meet these criteria are numbers of
17
patients with mixed dyslipidemia that we used in
18
our program.
19
overall, over 2,000 would meet the definition of
20
mixed dyslipidemia.
21 22
I believe the numbers of patients
Out of the 5,000 plus patients
We've looked at the response of alirocumab in that subset of patients.
In regards to LDL,
A Matter of Record (301) 890-4188
209
1
their reductions are fairly similar to what we see
2
in the overall population.
3
at patients with lower levels of HDL and see
4
similar findings there as well.
5
DR. SHAMBUREK:
And we've also looked
Does the FDA have any
6
comments on mixed hyperlipidemia or dyslipidemia
7
definitions or indications?
8
DR. GOLDEN:
9
The definition, as said, was
triglycerides greater than 150.
And when you look
10
at subgroup analyses for LDL as well as
11
triglycerides and HDL in that subgroup, didn't look
12
different than the whole, similarly with HDL, less
13
than 50, I believe.
14
too many more comments about that specific patient
15
population because, again, you know, I presented
16
the triglyceride and HDL results overall, and it's
17
a little bit of a mix.
18
DR. J. SMITH:
But beyond that, I don't have
If I can add to that, I think
19
by calling out mixed dyslipidemia in the
20
indication, I'm not certain what the sponsor's
21
intention is.
22
We would still be looking at this as an
They would have to speak to that.
A Matter of Record (301) 890-4188
210
1
LDL-lowering drug.
2
to identify patients at high cardiovascular risk
3
for whom you believe that LDL lowering is a
4
clinical benefit, then we'd like to hear you
5
discuss that.
6
So if that population is a way
By calling out mixed dyslipidemia, I think
7
that we're not really looking to entertain
8
discussions about whether or not effects on HDL or
9
triglycerides in and of themselves would be the
10
piece that is leading the clinical benefit.
11
think we've had a lot of conversations about that
12
in other settings.
13
LDL-lowering effect of the drug here.
14
DR. SMITH:
15
DR. ROBERTS:
I
We're really focused on the
Yes, Dr. Roberts? Sorry.
I just wanted to go
16
back to the question about the neurologic serious
17
adverse events.
18
up, please?
19
If I could just have slide 57 back
So you can see on the patients with serious
20
adverse events that there were patients in the
21
placebo and ezetimibe groups that experienced a
22
serious adverse event.
So in the placebo-
A Matter of Record (301) 890-4188
211
1
controlled pool, that one placebo patient, that
2
was -- the preferred term I believe was "gait
3
disturbance."
4
treated in the ezetimibe-controlled pool.
5
preferred term was "paresthesia."
And then there was another patient And that
So there were a total of -- let's see.
6
There
7
were 7 alirocumab-treated patients that had a serious
8
adverse event, of which 4 I mentioned in the next
9
slide.
Then there were 2 control-treated, 1 placebo
10
and 1 ezetimibe.
And I have in the background
11
material for the FDA package on, I believe it is,
12
page -- it's table 95 -- we have the narratives of
13
all of those serious adverse events if you want to
14
take a look at those.
15
DR. SMITH:
So I think we're going to need
16
to take a break for lunch.
17
I would like to ask the panelists if there are any
18
questions they might have that would be requesting
19
a look at data that might take a little time for
20
either the FDA or the sponsor to pull together, if
21
they're able, for us to see later this afternoon.
22
So I'm seeing no hands for that.
But before we do that,
A Matter of Record (301) 890-4188
212
1
So we'll now break for lunch.
Again, we
2
will have some more clarifying question opportunity
3
before we get into the formal discussion questions
4
later today.
5
one hour from now, let's say, at 1:00 p.m.
6
take any personal belongings you may want with you
7
at this time.
8 9
We'll reconvene again in this room in Please
Committee members, please remember there should be no discussion of the meeting during
10
lunch, among yourselves, with the press, or with
11
any member of the audience.
12
panelists can go to Rook's Corner, which is located
13
in the lobby area.
14
area for panel members.
15
yourself as a panel member.
16
escorted into that seating area, which is I
17
understand where the lunch actually is.
18
don't have to get it and go there.
19 20 21 22
Regarding lunch, the
And there's a reserved seating You should identify And then you can be
So you
Okay?
So we'll see you all back here at 1:00 p.m. Thank you. (Whereupon, at 12:04 p.m., a lunch recess was taken.)
A Matter of Record (301) 890-4188
213
1
A F T E R N O O N
(1:02 p.m.)
2
Open Public Hearing
3 4
S E S S I O N
DR. SMITH:
Good afternoon.
We will next
5
have the open public hearing session of this
6
meeting today.
7
Administration, the FDA, and the public believe in
8
a transparent process for information-gathering and
9
decision-making.
Both the Food and Drug
To ensure such transparency at
10
the open public hearing session of the advisory
11
committee meeting, FDA believes it is important to
12
understand the context of an individual's
13
presentation.
14
For that reason, FDA encourages you, the
15
open public hearing speaker, at the beginning of
16
your written or oral statement, to advise the
17
committee of any financial relationship that you
18
may have with the sponsor, its product, and if
19
known, its direct competitors.
20
financial information may include the sponsor's
21
payment of your travel, lodging, or other expenses
22
in connection with your attendance at the meeting.
A Matter of Record (301) 890-4188
For example, this
214
1
Likewise, FDA encourages you, at the beginning of
2
your statement, to advise the committee if you do
3
not have any such financial relationships.
4
choose not to address this issue of financial
5
relationships at the beginning of your statement,
6
it will not preclude you from speaking.
If you
7
The FDA and this committee place great
8
importance in the open public hearing process.
9
insights and comments provided can help the agency
The
10
and this committee in their consideration of the
11
issues before them.
12
That said, in many instances and for many
13
topics, there will be a variety of opinions.
14
of our goals today is for this open public hearing
15
to be conducted in a fair and open way, where every
16
participant is listened to carefully, and treated
17
with dignity, courtesy, and respect.
18
please speak only when recognized by the
19
chairperson.
20
One
Therefore,
Thank you for your consideration.
So at this point, will speaker number 1 step
21
up to the podium and introduce yourself?
22
state your name and any organization you're
A Matter of Record (301) 890-4188
Please
215
1
representing for the record.
2
MS. O'CONNOR:
Good afternoon.
My name is
3
Peggy O'Connor.
4
flew down from Boston this morning, and I am
5
rushing back to Spain to fly out on my family
6
vacation this evening.
7
the personal expense and the effort.
I
Being here today is worth
Frankly, I wouldn't be able to do it without
8 9
Thank you for letting me speak.
being on alirocumab.
It has given me my life back
10
and gotten my cholesterol down.
My history with
11
high cholesterol has been difficult. I eventually started on statins several
12 13
years ago.
14
below 100, but I had muscle spasms as a side
15
effect.
16
types of adjustments.
17
months, and then she prescribed a second statin.
18
On the second dose, I ended up in the emergency
19
room with intense muscle spasms that were only
20
relieved by IV Ativan and IV valium given together.
21
Three months later, I had an MI.
22
I was able to get my LDL down to just
They were pretty severe; we need various Went off it for a few
Finally, frankly, the muscle spasms from the
A Matter of Record (301) 890-4188
216
1
statins were far worse than the symptoms of my MI.
2
So there I was, a mother of two high schoolers who
3
were active.
4
were active, and I was struggling, focused on
5
everything I ate, and the need to exercise, and
6
then of course problems with the meds.
And I was their chauffeur, and we
My doctor immediately tried me on two
7 8
cholesterol-lowering meds in addition to the stent
9
that was placed.
10
We adjusted the dose.
The second
dose, more spasms. So then we tried Niaspan.
11
I woke up the
12
first night at 2:00 a.m. thinking the house was on
13
fire.
14
for four nights, thinking that I have to do this.
15
So I kept trying.
16
And I exercised five days a week and ate right, but
17
I couldn't get my LDL below 111.
18
It was not.
It was me.
I kept taking it
I stretched out cardiac rehab.
Then I got into this trial.
I was terrified
19
at the prospect that I might be one of the patients
20
on statins, but I decided to take the chance.
21
I finished the blind trial, and I have been in the
22
open trial for a year and a half, and my LDL is at
A Matter of Record (301) 890-4188
And
217
1 2
50 or below. I have a normal life again.
I would not
3
have been able to take a two-week family vacation
4
such as what I'm about to leave on without being on
5
this drug.
6
what I was going to eat, exercising, and whether or
7
not I would get adequate healthcare overseas.
8 9
I would be too concerned about eating,
For some, statins work, but for others like me, just can't tolerate them.
Alirocumab is a
10
medication that enables people to get their LDL
11
goals, and I want to thank you very much for your
12
attention.
13
DR. SMITH:
Thank you.
Will the second open
14
public hearing speaker please step up to the
15
microphone, identify yourself, and any organization
16
you may be representing?
17
MS. HORSCH:
Good afternoon.
My name is
18
Terry Lim, and I've come here from Houston, Texas
19
entirely at my own expense to speak with you today
20
about my strong belief in the need for this drug
21
and in its benefit to patients like me.
22
a time when I didn't feel I would live to be 40.
A Matter of Record (301) 890-4188
There was
218
1
After failing to heed warnings of my FH
2
dating back to the mid-80s, I had a heart attack in
3
1994 at the age of 34 and a triple bypass.
4
then on, my doctors and I got very aggressive about
5
treating my high cholesterol with everything
6
available, but we could never get my numbers
7
anywhere close to the recommendations.
8 9
From
In 1999, I had a heart valve replacement and another bypass.
I felt like a ticking time bomb.
10
Thankfully, I was able to get into this trial and
11
am now in the open-label extension trial as well.
12
My total cholesterol is down from 400 to 129, and
13
my LDL is 57, down from over 300.
14
taking six huge fish oil pills to one each day and
15
have discontinued Zetia and 3750 milligrams of
16
Welchol.
17
the first time since I began taking statins over
18
20 years ago.
19
I have gone from
My liver function tests are normal for
I'm an accountant by profession, but I've
20
had no problems giving myself these injections.
21
During the blinded phase of the trial, I
22
experienced some hives at the injection site, but
A Matter of Record (301) 890-4188
219
1
they stopped when we moved the injection site. My great lab results put me a lot more at
2 3
ease.
I think I just may see old age with my
4
husband, who is here with me today.
5
rest of my family is here in spirit as well.
6
of my parents had elevated cholesterol and died of
7
heart disease.
8
her two sons have been diagnosed and are being
9
treated for FH.
And I know the Both
One of my two sisters and one of
I know what it's like to wait for
10
a medication to become available, and when it
11
finally comes out, it brings hope. My FH has been very difficult to treat.
12 13
I've swallowed as many as 20 pills a day, but my
14
numbers still weren't where they were supposed to
15
be.
16
hope, and the specter of FH does not haunt me quite
17
so much.
18
my story with you.
19
questions you may have.
20
I felt helpless.
Alirocumab gives me new
Thank you for this opportunity to share
DR. SMITH:
I'll be happy to answer any
Thank you.
Will speaker
21
number 3 please step up to the podium and introduce
22
yourself?
Please state your name and any
A Matter of Record (301) 890-4188
220
1
organization you are representing for the record. MR. KARAS:
2
Hi.
My name is Jonathan Karas.
3
I'd like to thank the FDA for this opportunity to
4
talk.
5
so that I could attend today.
6
relationship with any drug or healthcare
7
organizations.
8
here today.
The FH Foundation helped cover my expenses I have no financial
I have volunteered my time to be
I am a patient with FH and a heart attack
9 10
survivor.
11
our brave son, Maxwell, has HoFH.
12
with high cholesterol around 1987 at age 13.
13
this time, I never heard the term "familial
14
hypercholesterolemia."
15
prescribe a statin.
16
only.
17
My beautiful wife, Erin [ph], has FH and I was diagnosed At
My pediatrician did not
He suggested diet and exercise
I went on through my teens and 20s with LDL
18
numbers in the 300s.
I was in great shape, though,
19
and a fearless 20-something.
20
informed that my high cholesterol was likely
21
genetic and that diet and exercise alone would not
22
be enough, until Friday, January 3, 2003.
I had never been
A Matter of Record (301) 890-4188
221
At age 28, I woke up nauseous and with chest
1 2
pain.
3
much attention to that.
4
I eventually went to a local hospital, where they
5
determined I was having a heart attack.
6
rushed to a major hospital in Boston and had three
7
stents put in. I was very lucky.
8 9
My left arm was numb also, but I didn't pay I tried to go back to bed.
I was
My life changed from
never taking meds to taking many, a statin, Plavix,
10
ACE inhibitor, beta blocker, aspirin.
I also
11
started cardiac rehab, which led to running three
12
miles a day, but my LDL was still above 100,
13
somewhere between 125 and 150.
14
after my heart attack, we added Zetia, which had
15
only recently been approved by the FDA.
16
that new drug did the trick.
17
since my heart attack, my LDL is consistently below
18
100, and I feel great.
Then about a year
Adding
It's been 12 years
I'm not sure what would have happened if
19 20
Zetia was not FDA approved.
Having another
21
treatment for my FH has made a big difference for
22
me.
A heart attack at age 28 is not normal, but my
A Matter of Record (301) 890-4188
222
1
story seems fairly common for FH patients.
On
2
FH-related forums, like the FH Foundation's
3
Facebook group, there are posts daily about
4
struggles with keeping LDL numbers at suggested
5
goal levels. Some cannot tolerate statins or ezetimibe.
6 7
Some people are concerned about safety and side
8
effects.
9
the levels their doctors recommend even with the
And some with FH and HoFH cannot reach
10
highest doses and combinations of statin,
11
ezetimibe, and bile acid sequestrants. I'm here to tell my story and hopefully
12 13
influence this agency on the importance of options
14
for the FH community.
15
different, resulting in a wide variety of possible
16
treatment needs, even changing over the course of a
17
lifetime.
18
Facebook group is that having options equals hope.
Each FH patient is
A common message on the FH Foundation
There is no silver bullet for the FH
19 20
patient, therefore, we need to continue to promote
21
research for safe new drugs and therapies.
22
you.
A Matter of Record (301) 890-4188
Thank
223
1
DR. SMITH:
Thank you.
Will speaker
2
number 4 now please step to the microphone?
3
identify yourself and any organization you may be
4
representing.
5
MR. WOLFE:
Sid Wolfe, Public Citizen Health
6
Research Group, and I have no conflicts of
7
interest.
8 9
Please
I want to spend a minute or so of my three minutes on a really well-written editorial in the
10
New England Journal that accompanied the
11
publication of the alirocumab and evolocumab
12
studies by two of the people who were on the HA ACC
13
task force of this issue, Stone and Lloyd-Jones.
14
And they start off by saying, because these
15
inhibitors achieve a much lower LDL level, a close
16
look at safety is of paramount consideration.
17
The thing that struck me the most was their
18
next sentence, which I'll read, "It would be
19
premature to endorse these drugs for widespread use
20
before the ongoing randomized trial is
21
appropriately powered for primary endpoint analysis
22
and safety assessments are available.
A Matter of Record (301) 890-4188
224
1
"Reports from several lipid treatment trials
2
provide important object lessons in this regard."
3
And torcetrapib was mentioned this morning; and
4
there are a couple others.
5
They also pointed out specific assessment of
6
neurocognitive function is needed.
And they went
7
on to quote something in the guidelines, the ACC
8
AHA guidelines, which they, and Dr. Eckel, and
9
others have participated in.
In terms of possible
10
use of non-statins, which is what this is obviously
11
in the category -- with a strong preference for use
12
of non-statins that have been determined to be safe
13
and effective in randomized controlled trials.
14
I think that's really the issue here, is
15
that the large randomized controlled trial, which
16
seems well-designed -- the outcomes study is not
17
complete yet.
18
Just a couple comments from the FDA's
19
package, which were not mentioned this morning
20
directly.
21 22
"The unexpected and disappointing results from CV outcomes trials, again, for other non-
A Matter of Record (301) 890-4188
225
1
statins, fenofibrate, cholesteryl ester transfer
2
protein, CTP, torcetrapib, and niacin should at
3
least give us pause as we consider the use of lipid
4
biomarkers and the assessment of benefit/risk." They also pointed out, as was discussed this
5 6
morning, the 3.2 instance of diabetes diagnosed in
7
the alirocumab group and 2.2 percent in the placebo
8
group.
9
way from the first patient, "We have concerns that
And then something that you just heard in a
10
many patients who have symptoms that may be
11
entirely unrelated to statins could prematurely
12
discontinue their statins and turn instead to a
13
PCSK9 inhibitor, which will lack long-term safety
14
data."
15
I'll skip over the next slide, which is a
16
review of the 18,000-person Odyssey study not
17
completed yet, and go to the conclusions.
18
"There may well be people who might benefit
19
from the additional LDL cholesterol lowering of the
20
PCSK9 inhibitors," certainly a possibility, "but at
21
this time, the inadequate evaluation of a large
22
enough number of patients over a long-enough time
A Matter of Record (301) 890-4188
226
1
with prespecified evaluation of cardiovascular
2
outcomes and other safety concerns does not provide
3
information for the necessary valuable benefit/risk
4
balance.
5
"In the absence of such critical
6
information, any approval at this time would
7
essentially be a premature endorsement for
8
widespread use, which was the concern in the
9
editorial."
10
I'd just like to add one thing in about five
11
seconds, which was mentioned very briefly this
12
morning that, which is that if this drug gets
13
approved, there's no question that will get into
14
widespread use.
15
people currently in the trial, the outcomes trial.
16
I would bet, if I were a betting person, large
17
amounts of money that the outcomes trial will never
18
be able to be finished if this drug is approved.
19
Thank you.
20
And amongst those will be the
DR. SMITH:
Thank you.
Will speaker
21
number 5 now please step up to the podium?
22
identify yourself and any organization you may be
A Matter of Record (301) 890-4188
Please
227
1 2
representing. MS. WILEMON:
God afternoon.
My name is
3
Katherine Wilemon.
4
Foundation, a not-for-profit organization that
5
supports research and advocacy for familial
6
hypercholesterolemia or FH.
7
I am the president of the FH
I am here today to speak to the pressing
8
need for additional therapies for those who are
9
born with and often die from heart disease as a
10
result of this inherited genetic disorder.
11
you for the opportunity to speak to you on behalf
12
of the more than 1.5 million people in the U.S. and
13
30 million people worldwide who have FH.
14
Thank
I came to this work after I had a heart
15
attack at age 39, and I have FH, as does my oldest
16
daughter.
17
window into the relationship between LDL
18
cholesterol and heart disease.
19
cannot recycle LDL effectively, or at all depending
20
on the mutation, we have extremely elevated
21
cholesterol from birth.
22
Individuals with FH give us a unique
Because our livers
Our LDL cholesterol levels range from
A Matter of Record (301) 890-4188
228
1
approximately 200 to well over 500.
It is this
2
lifelong exposure to very high cholesterol that
3
results in aggressive premature coronary artery
4
disease.
5
heart attacks in our 20s, 30s, 40s, and 50s are
6
common.
7
advances in interventional cardiology, 50 percent
8
of people die from their first heart attack.
9
with high-dose statins, bile sequestrants,
For individuals with heterozygous FH,
And as we know, even with the incredible
Even
10
ezetimibe, rigid, low-fat diets, and sometimes
11
apheresis, individuals with FH rarely reach the
12
optimal range of LDL.
13
FH is an autosomal-dominant condition.
Each
14
child born to an affected parent has a 50 percent
15
chance of also inheriting this deadly condition.
16
FH families are haunted by stories of loved ones
17
having heart attacks and often dying in the prime
18
of their lives.
19
tell you that we live in fear of not living long
20
enough to see our own children grow up.
And as FH parents ourselves, I can
21
Yet, longitudinal studies of people born
22
with FH show that this condition does respond to
A Matter of Record (301) 890-4188
229
1
early and intense treatment.
2
cholesterol is lowered and the sooner it is
3
lowered, the longer we are living.
4
The more our LDL
This is a pivotal time for FH and for
5
greater understanding of heart disease within all
6
populations.
7
unmet need of those with FH in mind as you weigh
8
the advantages as well as the unknowns about this
9
new class of PCSK9 inhibitors.
10
We urge you to keep the tremendous
DR. SMITH:
Thank you.
Thank you. Will speaker
11
number 6 now please step up to the microphone?
12
Please identify yourself and any organizations you
13
may be representing.
14
MR. DOLLARHYDE:
Good afternoon.
My name is
15
Daniel Dollarhyde, and I am president of the
16
National Capitol Area Chapter of Mended Hearts.
17
have currently congestive heart failure.
18
been battling heart disease for about 30 years.
19
And most of those 30 years, I have been on statins.
20
Statins, along with exercise and the other drugs
21
that I have taken, have enabled me to live this
22
long and see nine grandchildren, five of which are
A Matter of Record (301) 890-4188
I
I have
230
1
now in college or graduating.
2
doctor's help in that the statins have brought me
3
this far.
4
own expense from Rockville.
5
So it is with a
By the way, I have traveled here on my
Mended Hearts is a national organization.
6
It's a voluntary group, nonprofit.
7
largest cardiovascular patient peer-to-peer support
8
organization in the nation.
9
nationwide who conduct 215,000 patient visits a
10 11
And it is the
We have 20,000 members
year in 460 hospitals across the nation. Our focus is on education, peer support, and
12
empowering patients to take charge of their own
13
health, and actively participate in medical
14
decisions.
15
to provide education, advocacy, and support for
16
64 years.
17
Mended Hearts has been helping patients
As this panel is addressing a new FDA
18
regulated prescription drug to treat cholesterol,
19
particularly for patients who are statin-resistant
20
or who have a genetic disorder characterized by
21
high cholesterol levels, specifically high levels
22
of LDL -- and the reason I say it all that way is
A Matter of Record (301) 890-4188
231
1
because I have difficulty saying
2
hypercholesterolemia, and that's FH, thank
3
you -- we believe it is important to note the
4
following.
5
Heart disease remains America's number one
6
killer of men and women, resulting in more deaths
7
each year than the top five cancers combined.
8
Heart disease is getting younger.
9
more U.S. citizens are being diagnosed with high
By that, I mean
10
cholesterol at younger ages, persistent, including
11
many of our children.
12
High cholesterol, particularly LDL-C
13
cholesterol, is one of the most persistent building
14
blocks in heart disease.
15
healthcare providers have as many tools to manage
16
cholesterol as possible, especially for young
17
patients and those whom currently treatment
18
protocols find ineffective.
19
It is important that
We are here to support innovation and
20
treatments of heart disease and to express our
21
support for companies who are developing this new
22
option for patients.
We hope further that you will
A Matter of Record (301) 890-4188
232
1
consider issues of accessibility, patients on fixed
2
income and with limited access to advanced medical
3
care facilities.
4
consideration. DR. SMITH:
5
Thank you for your time and
Thank you.
Will speaker
6
number 7 now please come to the microphone?
7
identify yourself and any organization you may be
8
representing. MS. DAMM:
9
Please
Hello, my name is Marie Damm.
10
I'm here from Doylestown, Pennsylvania.
11
relationship with any drug company or healthcare
12
organization.
13
allowing me to speak today regarding something so
14
important to me and my family.
I want to thank the committee for
I'm 53 years old, and I have FH.
15
I have no
Both of my
16
grown children have FH, and the majority of my
17
mother's family does as well.
18
six of her nine siblings to heart disease due to
19
FH.
20
and, thankfully, I have not yet had one.
21
the past year, two of my cousins have buried their
22
sons due to a heart attack.
My mother has lost
My mother had a heart attack at the age of 56 Within
They were 32 and 40
A Matter of Record (301) 890-4188
233
1
years old.
2
bypass surgery at the age of 30, while yet another
3
one was told by his cardiologist that he needed a
4
chisel to get through his aorta when doing a
5
catheterization.
6
Another one of my cousins had triple
Clearly, FH runs rampant in my family.
I
7
have been treating my cholesterol since I was in my
8
early 30s.
9
Zetia, and others, all in conjunction with statins,
I've used Welchol, Lescol, niacin,
10
but have never been able to achieve an acceptable
11
cholesterol level.
12
My exposure to high cholesterol my whole
13
life has taken a toll.
14
demonstrating a significant amount of plaque in my
15
coronary arteries, which calculates into me having
16
more plaque than 95 percent of women my age.
17
Knowing that I am such a high risk of a heart
18
attack, I worry daily about every little bit of
19
discomfort I feel in my chest, as I want to live
20
long enough to see my grandsons grow up.
21 22
My doctors have run tests
Today, I have hope.
I have been on the
PCSK9 inhibitor for almost two years since I am
A Matter of Record (301) 890-4188
234
1
statin intolerant.
2
total cholesterol was 475.
3
went from 382 to 166.
4
these results on statins because I could never take
5
them for any length of time before the unbearable
6
side effects kicked in.
7
has allowed me to reduce my cholesterol to its
8
lowest level ever.
9
benefit could be achieved by other members of my
10
extended family, who also live with this serious
11
medical condition.
12
When I began treatment, my It is now 246.
My LDL
I was never able to achieve
Being on this PCSK9 trial
And I hope and expect the same
Most importantly, I'm hoping additional
13
treatment options will become available for my
14
children to use.
15
cholesterol level of 476, and my 24-year-old son
16
has a level of 375.
17
treatment on a statin, so it remains to be seen if
18
his body will tolerate it or not.
19
begin her treatment as soon as she is finished
20
bearing my grandchildren.
21
be invaluable to my son and daughter as well as my
22
grandchildren.
My 27-year-old daughter has a
My son has just begun
My daughter will
A PCSK9 inhibitor would
A Matter of Record (301) 890-4188
235
Lastly, I have tolerated the PCSK9 very
1 2
well.
3
my doctors.
4
part of this clinical trial at the University of
5
Pennsylvania.
6
children from the pain and suffering that I've
7
watched my mother endure for years.
8
listening.
9
I am still trying to reach the goal set by And I'm so thankful to have been a
I'm hoping to save myself and my
DR. SMITH:
Thank you.
Thank you for
Will speaker
10
number 8 now please step to the microphone?
11
identify yourself and any organization you may be
12
representing.
13
DR. MCKENNEY:
Good afternoon.
Please
I am Dr. Jim
14
McKenney from National Clinical Research in
15
Richmond, Virginia.
16
alirocumab and several other drugs in this category
17
and have worked for a research company that
18
receives grants for this work.
I have conducted research with
19
I wish to briefly review with you this
20
afternoon the setting into which alirocumab is
21
being considered.
22
era of cholesterol, of cardiovascular risk
In my view, we are in the modern
A Matter of Record (301) 890-4188
236
1
reduction.
And it began in the 1980s, when
2
research strongly pointed to a direct relationship
3
between blood cholesterol levels and coronary
4
disease.
5
In 1984, the Lipid Research Clinics
6
published for the first time a randomized clinical
7
trial showing that lowering blood cholesterol
8
levels lowered coronary events, and this
9
breakthrough study led directly to the formation of
10
the National Cholesterol Education program by
11
NHLBI.
12
around cholesterol into practice.
13
And that helped translate the science
In 1987, the first statin was released.
In
14
1988, the first cholesterol treatment guidelines
15
were released.
16
seen major advances, including many outcomes
17
studies supporting updated treatment guidelines, an
18
emphasis on lifestyle changes, interventions for
19
other risk factors, and ways of treating patients
20
with urgent cardiovascular symptoms.
21 22
Over these 30 years since, we have
Most notably, there's been in that period of time a 65 percent reduction in deaths due to
A Matter of Record (301) 890-4188
237
1
coronary disease.
But -- and there is a but -- we
2
also know that 50 percent of patients who receive
3
all that we can provide them today with current
4
therapy will still experience a cardiovascular
5
event.
6
still the new number one killer.
7
ways of lowering risk, and anti-PCSK9 therapies may
8
just be that therapy, that way.
And it has been stated, heart disease is So we need new
You've heard today about FH patients.
9
There
10
are many other patients in the clinic that need
11
this help, including patients with high-risk
12
cardiovascular disease, recurring events, and less-
13
than-desirable cholesterol levels, as well as those
14
patients intolerant to statin therapy. The final thought that I'd like to leave
15 16
with you is this.
17
cholesterol levels by pulling cholesterol particles
18
out of the bloodstream by LDL receptors, the same
19
mechanism, by the way, that bile acid sequestrants
20
use.
21
monoclonal antibodies also lower LDL cholesterol.
22
Statin therapy reduces
And it is the same mechanism that PCSK9
I believe that this gives us comfort that we
A Matter of Record (301) 890-4188
238
1
are extending known and proven mechanisms of risk
2
reduction with these new drugs.
3
have to wait for the outcomes studies to prove
4
this.
5
the reasons stated.
6
Of course, we'll
I encourage you to approve this new drug for
DR. SMITH:
Thank you very much. Thank you.
Will speaker
7
number 9 now please come to the microphone?
8
identify yourself and any organization you may be
9
representing.
10
MR. PICKHARDT:
Good afternoon.
Please
My name is
11
Dave Pickhardt, and I am a person with FH.
12
you for allowing me to be here today.
13
disclosure related to my appearance here today, I
14
wanted to let you know that I am a past employee of
15
Sanofi Aventis, retired in 2005.
16
board member and chairman of the FH Foundation.
17
did pay 100 percent of my own travel costs to come
18
here from the Kansas City area, and I have received
19
no compensation to appear here today.
20
Thank
As full
I'm also a past I
I discovered I had high cholesterol during a
21
complete physical 37 years ago, and despite being
22
healthy in every other way, my cholesterol was
A Matter of Record (301) 890-4188
239
1
extremely high, 450 total in my case.
I knew my
2
cholesterol would lead to cardiovascular disease,
3
and, thus, I began my journey.
4
addressing my cholesterol as high cholesterol only
5
through diet, exercise, and even some periods of
6
denial and a lot of anxiety, I was introduced to a
7
lipidologist.
8
understanding that I had FH, a genetic disease, and
9
helping me to fight my high cholesterol.
After years of
And this changed my life in terms of
I'm very
10
grateful for this knowledgeable clinician and his
11
sponsor.
12
However, the fact is that there were two
13
things true 37 years ago, when I was first
14
discovered to have high cholesterol, that are I
15
think still true today.
16
layperson, but the first thing I believe to be true
17
is that there's not enough awareness of FH.
18
know for a fact that only a very small percentage
19
of patients with FH in the United States have
20
actually been diagnosed.
21
that the available therapies for FH are often
22
inadequate at their highest tolerable doses.
I'm saying this as a
We
And the second thing is
A Matter of Record (301) 890-4188
240
1
For example, despite my highest levels of
2
dosage on currently available therapies at the
3
highest levels of approved dosage, my LDL
4
cholesterol usually remained at more than three
5
times what would be considered ideal levels.
6
recently, I participated in a trial for a PCSK9
7
drug, this compound you're evaluating.
8
didn't know during the trial whether I was on
9
placebo or the active drug, I did get my
And
Although I
10
cholesterol checked on my own immediately after the
11
study was completed.
12
The results for me were miraculous, total
13
cholesterol, 150, LDL cholesterol, 50.
14
only dream previously, despite a regimen of diet,
15
exercise, and available drug therapy, of achieving
16
these levels.
17
but together, these two drugs helped me to say
18
mission accomplished.
19
I could
I love my statin for what it does,
My interest in your deliberations goes
20
deeper than my own situation, however.
21
learned that each child or person with FH has a
22
50 percent chance of also having FH.
A Matter of Record (301) 890-4188
I've
All three of
241
1
my children have FH.
2
of course.
3
all the FH patients in the U.S. who want a shot at
4
also conquering this disorder, I'm simply asking
5
you to give this compound every consideration so
6
that it might be available to them, and they can
7
somehow, someday soon, also say that their mission
8
with their FH is accomplished.
9
much.
10
And they were born with it,
And for their sake and for the sake of
DR. SMITH:
Thank you.
Thank you very
Will speaker
11
number 10 now please step to the microphone?
12
Please identify yourself and any organizations you
13
may be representing.
14
MS. CLOSE:
Good afternoon.
I'm Kelly
15
Close, the founder of the diaTribe Foundation,
16
which works on improving life for people with
17
diabetes and pre-diabetes.
18
30 years.
19
potential of PCSK9 inhibitors for people with
20
diabetes, especially type 2 diabetes.
21 22
I've had diabetes for
Today, I want to discuss the huge
The diaTribe Foundation is funded primarily by the Helmsley Charitable Trust as well as about
A Matter of Record (301) 890-4188
242
1
100 other organizations profitable and non-
2
profitable, including Sanofi.
3
disease is the leading cause of people with
4
diabetes, and it has become increasing clear that
5
effective prevention requires a holistic approach
6
that addresses a bunch of risk factors.
7
of them.
8 9
Cardiovascular
LDL is one
We know for people with type 2 diabetes that lipid management and LDL lowering in particular
10
must be better addressed.
11
of leaders in diabetes and cardiology that managing
12
cholesterol is "the" most important thing that
13
people with diabetes can do for their
14
cardiovascular health.
15
We've heard from plenty
And this happening.
So we see a bunch of curves about how we're
16
doing on cardiovascular health in the U.S., and
17
it's nothing to write home about anymore,
18
especially as type 2 diabetes expands in such
19
crisis proportions and as cardiovascular disease in
20
people with diabetes is so rampant.
21
affects the broad population's CV curves so
22
negatively.
A Matter of Record (301) 890-4188
And that
243
1
Despite some significant progress over time,
2
many people with diabetes have trouble meeting
3
their LDL goals, and there's a significant need for
4
new therapies out there that work well for those
5
that don't have other options or for those for whom
6
the side effect profile is so challenging that it
7
brings up major adherence problems.
8 9
According to NHANES, only 56 percent of people with diabetes with an LDL target of 100 and
10
only 28 percent of people with diabetes with a
11
target of 70 achieve their goals.
12
As type 2 diabetes and cardiovascular
13
disease both hit lower ends of the socioeconomic
14
spectrum, please consider how you could help make
15
public health better by making better CV therapies
16
available, particularly for those at especially
17
high risk.
18
Twenty percent of people with diabetes can't
19
even take statins, and 20 percent might not sound
20
like that high of a percentage, but that is
21
6 million people with type 2 diabetes alone.
22
you could really enable an intervention on a large
A Matter of Record (301) 890-4188
So
244
1
scale that would have very far-reaching
2
implications.
3
To sum up, the 29 people [sic] with diabetes
4
in the U.S. need more options to manage CV risk.
5
PCSK9 inhibitors offer a compelling risk/benefit
6
profile for these.
7
to seeing data from the ongoing CV outcome trials,
8
we believe that there is already a strong body of
9
evidence to support approval for those at high risk
10 11
with diabetes.
While of course we look forward
Thank you.
DR. SMITH:
Thank you.
Will speaker
12
number 11 now please step to the microphone?
13
Please identify yourself and any organizations you
14
may be representing.
15
Okay.
Now, would speaker number 12 then
16
please step to the microphone?
17
yourself and any organizations you may be
18
representing.
19
MR. PEARSAL:
Please identify
Good afternoon, distinguished
20
members of the advisory committee.
21
the opportunity to briefly comment.
22
introduction, my name is Rob Pearsal.
A Matter of Record (301) 890-4188
Thank you for By way of I'm an
245
1
attorney from Richmond, Virginia, and I have
2
familial hypercholesterolemia.
3
expense and have no financial interest in the
4
outcome of this hearing.
5
I'm here at my own
I inherited hypercholesterolemia from my
6
mother's side of the family, and she and my
7
siblings are also affected.
8
fortune to be diagnosed decades ago and started
9
receiving interventional care early on.
My mother had the good
10
Unfortunately, she is one of the many that are
11
statin intolerant, that is unable to receive the
12
aggressive statin treatment that's medically
13
indicated.
14
Despite having no other risk factors, my
15
mother had to undergo bypass surgery and
16
installation of a pacemaker this past fall.
17
Fortunately, due to outstanding care, she's
18
recovering with a good prognosis.
19
know for sure the availability of a drug like
20
alirocumab could possibly have prevented or at
21
least slowed the progression of coronary artery
22
disease for my mother and many other statin-
A Matter of Record (301) 890-4188
And we'll never
246
1
intolerant patients. I was diagnosed with hypercholesterolemia in
2 3
1989.
4
number was over 510.
5
intolerant, and at the recommendation of my primary
6
care physician, Dr. Steven Richard, I started a
7
statin regimen and then participated in numerous
8
clinical trials for second- and third-generation
9
statins as well as niacin.
10
And not to brag, but my total cholesterol Luckily, I am not statin
Despite the fact that I received excellent
11
interventional medical care and take a maximum dose
12
of statin daily, my lipid profile still indicates
13
that I am at very high risk for heart disease.
14
Novel lipid-lowering agents such as alirocumab
15
offer patients such as me the hope that heart
16
disease and vascular disease will not be an
17
inevitable part of our future.
18
I respectfully request the committee to give
19
consideration to the current need for new
20
treatments for patients such as me, my mother, and
21
many others.
22
attention.
Thank you for your time and
A Matter of Record (301) 890-4188
247
DR. SMITH:
1
Thank you.
Will speaker
2
number 13 now please step up to the microphone?
3
Please identify yourself and any organization you
4
may be representing. DR. ZANGANEH:
5
Good afternoon.
I'm Farhad
6
Zanganeh.
I'm a clinical endocrinologist.
I'm
7
representing the American Association of Clinical
8
Endocrinologist, AACE.
9
We put out very innovative guidelines for diabetes,
You are familiar with us.
10
lipids.
11
and I'm heading back to the office to see patients.
12
I thought I'd start with a quote from Shakespeare
13
to kind of change the mood of the afternoon.
14
for a muse of fire that would ascend the brightest
15
heaven of invention!"
16
And I actually flew from Boston, the ADA,
"O
So the point is, it's not a rite of passage.
17
You've seen the science, and the science is
18
beautiful.
19
and the medicine looks great, it's not a rite of
20
passage.
21
really placed my number 13 really nicely.
22
you guys heard about the patients, because it might
So just because the studies are done
But I think the serial position effect
A Matter of Record (301) 890-4188
I think
248
1
really change my presentation because, really, it's
2
about the patient.
3
patient, third, the patient, only then comes the
4
science.
First, the patient, second, the
So I think the point is, there's a statement
5 6
of need out there.
7
are hurting.
8
back in time.
9
when atorva and rosuva were coming.
10 11
them?
People are not at goal.
People are nowhere near goal.
People Go
Similar conversations took place Do we need
Yes, we do. So I think the point is, this is exactly
12
where we are, so I think this is how science
13
advances.
14
I was planning to bring some patients over -- and
15
the stories are here.
16
this is exactly where we are and that you will hear
17
more of these stories as you go through.
18
I think clinically these stories -- and
So I think the point is,
So just to sum up with the story here, it
19
would be that we are looking at a safety study that
20
is a safety of this class of drugs that is similar
21
to ezetimibe and efficacy, which is far greater
22
than statins, and so far a good start as far as the
A Matter of Record (301) 890-4188
249
1
trends that they're looking at. I think for the FH patients, for the non-FH,
2 3
for those that are not at goal, those that are
4
hurting, that cannot tolerate, and we have done
5
complete rearranging of this statin, that statin,
6
taking it intermittently, all kinds of innovative
7
plans.
So I think the point is, we definitely need
8 9
People are not at goal.
additional therapies to get the patients to go.
10
And one of my patients asked me last week, "Why are
11
we rearranging my schedule?"
12
going to the hearing."
13
help my cholesterol?"
14
guys are pretty smart.
15
thing."
16
I said, "Well, I'm
And he said, "Would that And I'm like, "Yeah, these They always do the right
So I think the point is now, people are
17
really depending on these things.
18
science is there.
19
So please don't break their heart.
20
DR. SMITH:
So I think the
You heard about the patients.
Thank you.
Will speaker
21
number 14 now please step up to the microphone?
22
Please identify yourself and any organizations you
A Matter of Record (301) 890-4188
250
1
may be representing. DR. JURY:
2
Thank you for the opportunity to
3
testify today.
My name is Dr. Nicholas Jury, and I
4
work at the National Center for Health Research.
5
And our research center scrutinizes scientific and
6
medical data and provides objective health
7
information to patients, providers, and
8
policymakers.
9
bring today.
And those are the perspectives I We do not accept funding from
10
pharmaceutical companies, and, therefore, I have no
11
conflicts of interest. The center strongly supports research to
12 13
advance the understanding of and solutions of
14
cardiovascular diseases.
15
search for safe and effective alternatives to
16
statins as well.
17
lowered LDL cholesterol compared to placebo, but
18
this effect is more likely to be significant in
19
white men, and the studies seem to indicate that
20
there might be a little bit of a difference in
21
women.
22
We also support the
We're glad to see that alirocumab
In addition, the lack of efficacy in the
A Matter of Record (301) 890-4188
251
1
women, there were no useful data for people of
2
color.
3
were Caucasian.
4
minorities to allow a subgroup analysis would be
5
very useful and effective in other racial and
6
ethnic populations.
7
higher rates of heart disease than whites.
8
course, the fact that the drug lowered LDL does not
9
necessarily improve health outcomes for any group,
10 11
Approximately 95 percent of the patients Adding additional patients of
And yet, minorities have And of
as you all know. In summary, the lack of a subgroup analysis
12
makes it inappropriate to approve this drug for
13
people of color.
14
cholesterol in white men, there was no conclusive
15
evidence that it lowers the risk of heart disease.
16
And although the drug lowered LDL
So in conclusion, we are asking you to
17
reject approval of this drug for all populations
18
until more studies are done to indicate that it is
19
safe for all adults.
20
DR. SMITH:
Thank you.
Thank you.
Will speaker
21
number 15 now please come to the microphone?
22
Please identify yourself and any organizations you
A Matter of Record (301) 890-4188
252
1
may be representing. MS. KEITH:
2
Good afternoon.
My name is
3
Catherine S. Keith.
I am the wife of speaker
4
number 12.
5
interest.
6
Richmond, Virginia.
7
Ms. Grace Pantano from Connecticut this
8
morning -- excuse me, this afternoon -- and her
9
statement is as follows.
I have no financial conflict of And I am here at my own expense from I am speaking on behalf of
"My name is Grace Pantano.
10
I have high
11
cholesterol, like all of the other patients you're
12
hearing from today, I'd imagine.
13
me, I cannot take statins, which is how I came to
14
be in this clinical trial beginning in June of
15
2013. "Alirocumab works for me.
16
Unfortunately for
My total
17
cholesterol went from 300 to 145 on this drug, and
18
my LDL from 220 to 71 the last time my blood was
19
tested in May.
20
low.
21
my experience on the trial and with the drug has
22
been what I had hoped.
My cholesterol has never been this
My cholesterol is now where it should be, and
I never had any side
A Matter of Record (301) 890-4188
253
1
effects and the injections were easy to do once I
2
got past the idea of giving myself a shot.
3
"This is a far cry from what my battle with
4
high cholesterol had been like.
5
successfully on a statin medication for 20 years,
6
and then I started having problems with my knees
7
and with leg cramps during the night.
8
went off the statin, my symptoms disappeared.
9
I had been
As soon as I
"That began a period of trial and error.
My
10
doctor experimented with putting me on a series of
11
statins, and I exhausted them all.
12
combinations of drugs in different doses.
13
started off slowly on one and added to the dose
14
gradually, but I could not tolerate the side
15
effects.
16
She tried I
"The side effects included a variety of
17
things such as hair loss, stomach upset, leg
18
cramps, and migraine types of headaches.
19
headaches were especially worrisome for me since I
20
suffered a stroke at age 37, which was caused by a
21
reaction to a prescription medication.
22
off all medications, and my cholesterol was over
A Matter of Record (301) 890-4188
The
So I went
254
1
300 and LDL was at 220.
I dieted and lost weight,
2
got into an exercise program, but my cholesterol
3
remained high. "Alirocumab has been a life-saver.
4
I am at
5
high risk due to my family history, previous
6
stroke, and inability to use statins.
7
has gotten my cholesterol down to where it should
8
be.
9
only can I continue to use it, but so that other
10 11
This drug
I want the drug to be approved so that not
people will benefit from it, too. "It is also my hope that this drug will be
12
provided to the public at an affordable price.
13
Thank you for this opportunity."
14
DR. SMITH:
Thank you.
Will speaker
15
number 16 now please come to the microphone?
16
Please identify yourself and any organizations you
17
may be representing.
18
MR. CLYMER:
Good afternoon, and thank you
19
for the opportunity to address you today.
20
John Clymer, executive director of the National
21
Forum for Heart Disease and Stroke Prevention, a
22
nonprofit coalition of organizations dedicated to
A Matter of Record (301) 890-4188
I am
255
1
preventing heart attacks and strokes and reducing
2
disparities.
3
several other HHS agencies are among the 85 members
4
of the National Forum drawn from the public,
5
private, and nonprofit sectors.
6
financial interests in the sponsors.
7
Regeneron, Sanofi, the FDA, and
I have no personal
As the Department of Health and Human
8
Services' Million Hearts Initiative has
9
highlighted, high LDL-C is a major risk factor for
10
heart attacks and strokes.
11
underscored many times over today by several of the
12
speakers who have preceded me.
13
We've had that
Million Hearts has drawn attention to the
14
large deficit of LDL-C control in the general
15
population and significant disparities in specific
16
populations.
17
gaps, one of which is that for some people, the
18
array of medical and nonmedical therapies available
19
today are insufficient or cannot be tolerated.
20
Thus, the National Forum applauds the development
21
of new therapies that provide more options to
22
reduce people's risk for heart disease and stroke.
There are multiple reasons for these
A Matter of Record (301) 890-4188
256
1
The National Forum sees value in the PCSK9
2
inhibitor class.
3
data from recent trials.
4
available has the potential to reduce the incidence
5
of cardiac episodes.
6
We are encouraged by the efficacy Making new options
We are optimistic that the risk of heart
7
disease in the U.S. can be reduced through safe and
8
effective new treatment options such as PCSK9
9
inhibitors in combination, of course with
10
behavioral, educational, and other important
11
initiatives and efforts, and that these therapies
12
will help bring us closer to achieving the Million
13
Hearts goal of preventing heart attacks and
14
strokes.
15
DR. SMITH:
Thank you.
Will speaker
16
number 17 now please step to the microphone?
17
Please identify yourself and any organizations you
18
may be representing.
19
DR. KNOWLES:
Hi.
My name is Josh Knowles.
20
I'm a physician scientist at Stanford.
21
adult cardiologist there in the Center for
22
Inherited Cardiovascular Disease, I take care of
A Matter of Record (301) 890-4188
And as an
257
1
about 100 FH patients with my colleagues.
2
chief medical advisor for the FH Foundation, a
3
patient-led charitable organization whose mission
4
is to increase the awareness of treatment of FH.
5
I'm also
The FH Foundation is a 501(c)(3), nonprofit.
6
All board members are either FH patients or FH
7
healthcare providers.
8
I don't have a financial relationship with either
9
of the companies that are presenting here today.
My position there is unpaid.
I
10
flew in on the red-eye to help represent patients
11
and dedicated healthcare providers.
12
here tomorrow, but I want to stress that my remarks
13
would pertain to tomorrow as well.
14
I can't be
We've all heard it, autosomal-dominant
15
condition, characterized by profound lifelong
16
elevations in LDL cholesterol that lead to a
17
20-fold lifetime increased risk of coronary heart
18
disease versus the general population.
19
history studies in the pre-statin era showed that
20
men with FH have a 50 percent chance of having a
21
heart attack by age 50 if untreated, and women have
22
a 30 percent chance by 60 if untreated.
A Matter of Record (301) 890-4188
Natural
258
Recent genetic studies from the National
1 2
Heart, Lung, and Blood Institute XM Sequencing
3
project importantly have shown that FH is more
4
common than we once thought.
5
as 1 in 200.
6
earlier work by Brown and Goldstein, demonstrating
7
that FH accounts for approximately 2 percent of
8
early-onset myocardial infarction in the United
9
States.
10
It may be as common
And they also confirmed results from
We know that only a very small fraction of
11
FH patients have been identified in the U.S., and
12
the failure to identify such individuals can have
13
dire consequences, as we've heard, due to delays in
14
implementation of guideline base therapeutic
15
approaches to lower LDL cholesterol.
16
multiple national and international guidelines,
17
including the recent ACC AHA guidelines on
18
cholesterol treatment that placed special emphasis
19
on this population, which is appropriate, and
20
underscored that aggressive cholesterol-lowering
21
strategies are needed.
22
There are
The rationale for this is that these studies
A Matter of Record (301) 890-4188
259
1
have shown treatment aggressively lowers morbidity
2
and mortality to that of the general population if
3
LDL goals can be achieved.
4
the natural progression of the disease in children.
It also seems to alter
Unfortunately, the currently available
5 6
therapy, it's only clear that a small minority can
7
achieve healthy LDL levels.
8
where FH is a national priority, only 20 percent of
9
patients are able to achieve an LDL of less than
In the Netherlands,
10
100, and only 40 percent achieve an LDL less than
11
130.
12
We have similar data emerging from the
13
United States through the FH Foundation's national
14
patient registry, called CAscade.
15
established in 2013 and is active at more than 10
16
lipid centers.
17
Mean pre-treatment LDL levels were 256.
18
Cardiovascular disease was present in 37 percent of
19
patients.
20
percent of patients, the mean on-treatment LDL was
21
only 156.
22
and only 41 percent achieved an LDL of 50 percent
This was
The mean age of the adults was 54.
Despite combination therapy in 85
Only 25 percent got LDLs less than 100,
A Matter of Record (301) 890-4188
260
1
reduction. This speaks to the urgent need for new
2 3
therapies for this population, and I urge you to
4
consider the incredible stories we've heard from FH
5
patients today during your deliberations.
6
you.
7 8 9
Thank
Clarifying Questions (continued) DR. SMITH:
Thank you.
And I would like to
thank all of the open public hearing speakers for
10
your contributions, which are very important for
11
this meeting today and for the process.
12
public hearing portion of this meeting now is
13
concluded, and we'll no longer take comments from
14
the audience.
The open
15
Before proceeding with the questions and
16
discussion issues raised by the FDA, I'd like to
17
first turn to the sponsor and ask if you can walk
18
us through some additional data presentations, as
19
briefly as possible, in response to some of the
20
questions this morning.
21 22
DR. EDELBERG: outstanding questions.
We have answered some of the I'm going to ask
A Matter of Record (301) 890-4188
261
1
Dr. Sasiela to review some of the CRP data that was
2
requested. DR. SASIELA:
3 4
think, by Dr. Everett about shifts in CRP.
5 6
DR. EVERETT:
I think it was Dr. Blaha.
DR. SASIELA:
Oh.
Yes.
7 8
There had been a request, I
It was Dr. Blaha.
My
apologies. So we did get the data table that you were
9 10
asking about, and I'll put that up here.
11
apologize for a somewhat busy table, but this is a
12
fairly traditional shift table.
13
population out into three subgroups, those who had
14
baseline CRP less than 1 milligram per liter, those
15
who are 1 to 3 at baseline, and then the bottom
16
rows of those greater than 3 at baseline, and then
17
the four columns are the same two groups, the
18
placebo-controlled pool, the ezetimibe-controlled
19
pool.
20
So I
We've broken the
I won't walk through all this merely to say
21
that, when you compare to the placebo-controlled
22
pool, the shift table is fairly identical between
A Matter of Record (301) 890-4188
262
1
alirocumab and placebo.
When you look at the
2
ezetimibe-controlled pool, there is a slight
3
difference, particularly in patients on statins.
4
When you add ezetimibe, you can see incremental
5
reductions in CRP, and I think that's reflected in
6
the shift table here.
7
DR. EDELBERG:
Dr. Sasiela will also address
8
an outstanding question on clarifying the
9
population in the ALTERNATIVE study.
10
DR. SASIELA:
Right.
One of the, I think,
11
questions that came up was about patients, a little
12
bit more about the patient population than what
13
we've seen.
14
what we've seen in our open-label extension trial
15
may be helpful.
16
from the earlier core presentation to remind you of
17
the design of the trial.
18
And I thought that an examination of
So I'll just put this slide up
If you remember, a total of 314 patients
19
were randomized into the trial.
20
patients had the opportunity, whether they
21
discontinued off the drug during the 24-week
22
treatment period or not, to enter into the open-
A Matter of Record (301) 890-4188
And all these
263
1
label extension trial. What I can show you now in the next slide is
2 3
data from the open-label extension trial.
4
the column on the left is the total patient numbers
5
and then the break-out by whether they were
6
randomized during the double-blind treatment period
7
to either atorvastatin, ezetimibe, or alirocumab.
8
You can see the treatment discontinuation
9
You see
rates overall in the patient populations during the
10
24-week treatment period.
And then you can see the
11
overall number of the patients who have
12
participated in the open-label extension, which has
13
been 281 out of the 313 patients, around
14
90 percent, and how many of them are ongoing, and
15
how many of them have undergone treatment
16
discontinuation during the open-label extension
17
period.
18
patients who now all in this open-label extension
19
period are receiving alirocumab over the course of
20
the trial.
And you can see it's very few numbers of
21
This includes obviously patients who
22
discontinued due to skeletal muscle adverse events,
A Matter of Record (301) 890-4188
264
1
obviously, during the double-blind treatment
2
period.
3
tolerance in the open-label extension on
4
alirocumab.
5
They're doing very well in terms of
DR. EDELBERG:
The third question we aimed
6
to address was around insulin, and I'll ask
7
Dr. Eckel to actually comment on the HOMA substudy
8
that was done in the program.
9
DR. ECKEL:
OPTIONS I and II really got some
10
insulin levels, but I'm going to show you more than
11
insulin.
12
index, but repeatedly used to assess insulin
13
sensitivity.
14
goes up, that's not good.
15
drops, that is favorable.
16
The HOMA-IR is really a pretty crude
So the idea here is, if the HOMA-IR And if the HOMA-IR
So what I'm showing you here is a slide that
17
looks at the HOMA-IR in option 1, changed from
18
baseline.
19
middle panel of week 12, the median value, and
20
week 24, the median value.
21
is, the HOMA-IR is 0.23 on the double-dose statin
22
arm, 0.37 on the ezetimibe arm, and minus 0.47 in
And I want you to concentrate on the
And what you see here
A Matter of Record (301) 890-4188
265
1 2
the alirocumab arm. We go down to 24 weeks.
Those numbers are
3
0.24, -0.35, and -0.39.
4
24 weeks, the HOMA-IR is improved compared to the
5
other groups.
6
Next slide.
In both intervals, 12 and
Now we're looking at the group
7
who are more likely to develop type 2 diabetes if
8
in fact they do have an effect of any intervention,
9
statin or alirocumab, on diabetes risk.
So again,
10
same columns, week 12, week 24 median values, 1.5,
11
1.5, 0.1, 1.1, 1.4, -0.2.
12
So a small sample, but the conclusion I make
13
from this HOMA-IR data is favorable, not
14
unfavorable, in terms of the change in insulin
15
sensitivity.
16 17 18
DR. SMITH:
Thank you.
Dr. Thomas, did you
have a question in regard to this? DR. THOMAS:
Yes, just in relation to this.
19
You mentioned about the HOMA-IR, but actually, it
20
depends on the mechanism of the diabetes.
21
this is a decrease in insulin secretion, actually,
22
this is not going to be the right test to use, so
A Matter of Record (301) 890-4188
So if
266
1
the insulin levels themselves actually would be
2
helpful. You have the same data.
3
If you can
4
calculate the HOMA-IR, you can do the whole insulin
5
sensitivity measurement as well. DR. ECKEL:
6
We have the HOMA-beta 2.
7
that slide available, perhaps?
8
correct?
10
Is that
Okay.
DR. EDELBERG:
9
It's not.
Is
We wouldn't be able to
present that today. The last question that we wanted to address
11 12
was the incidence rate of neurocognition that was
13
asked.
14
DR. BRAUNSTEIN:
So you were interested in
15
the incident rates of neurocognitive events.
16
those are shown here, and I think there are a
17
couple points I want to make.
18
when you look at this -- I'm sorry.
19
being shown to me, but it wasn't being shown to
20
you, but here it is.
21 22
And
First is that, even The slide is
You can see that, when you look at the event rater, the crude event rates, now by rates for
A Matter of Record (301) 890-4188
267
1
100 patient-years, we're seeing very comparable
2
rates on alirocumab or control, both in the
3
placebo- and ezetimibe-controlled pools.
4
differences we're seeing -- there's not an effect
5
of different times that would be accounting for any
6
differences.
7
ezetimibe pool is probably due to the fact that
8
those are shorter studies.
9
So any
The little higher rates in the
But there are two points that I want to make
10
about these data.
11
events are transient and not serious.
12
long-term study, we saw a 2 and a half percent rate
13
of heart endpoint MACE events.
14
considering MACE events and the potential for
15
benefit, one needs to consider that, with
16
alirocumab, we have very strong genetic data that
17
loss of PCSK9 function leads to decreased
18
cardiovascular risks in the ERIC study and in
19
genetic and epidemiologic studies, 88 percent over
20
a 15-year period.
21 22
The first is that neurocognitive And in the
And when
The fact that we had profound LDL lowering that we could get with alirocumab, these all
A Matter of Record (301) 890-4188
268
1
suggest that there's unique aspects of the
2
mechanism of action of alirocumab that the
3
committee should keep in mind. The other point I want to make is that to
4 5
the extent that there's any concern about the
6
safety effects related to low LDL, I want to remind
7
the committee that the dose regimen that we have
8
proposed is actually designed to minimize the
9
chances of having low LDL levels. DR. SMITH:
10
So thank you.
Thank you to the sponsor.
I
11
would next like to focus just on -- again, before
12
going to discussion questions, any more clarifying
13
points? Dr. Blaha, you were passed over earlier
14 15
today.
Did that get resolved for you, a question
16
you had?
17
it may have been addressed.
18
opportunity.
A long time ago, you had a question and I'm giving you an
19
DR. BLAHA:
[Inaudible – microphone off.]
20
DR. SMITH:
Dr. Hiatt?
21
DR. HIATT:
I have a question -- this is
22
William Hiatt -- to Dr. Golden.
A Matter of Record (301) 890-4188
At the end of your
269
1
presentation on net clinical benefit, you made the
2
comment that, based on the data, many of the safety
3
concerns raised appeared to be monitorable or
4
manageable, and certainly maybe the very low LDL
5
levels are monitorable.
6
included things like glycemic shifts,
7
hypersensitivity, and liver enzyme abnormalities.
8 9
But the things of interest
So there's a little over 5,000 patient-years of exposure, and I'm wondering if this committee
10
were to make a recommendation to you to approve
11
this drug and you follow that recommendation, do
12
you have enough information to write a label?
13
I'm looking at table 108 in your background
14
document.
15
It just talks about treatment-emergent diabetes,
16
which has a hazard ratio in the placebo-controlled
17
trials of 1.07, upper bound of 1.50.
18
I don't think you need to pull that up.
That's not overwhelming to me.
I'm just
19
wondering if you have enough information to take
20
these concerns you've listed here and actually put
21
that in the label, or do you feel that it's
22
inadequate to even write a label that we would
A Matter of Record (301) 890-4188
270
1
appoint to these concerns? I realize this is turning the question a
2 3
little bit back on you, but I'm just wondering what
4
your comfort level is in terms of the safety
5
database we have on some of these key adverse
6
events of interest, in terms of whether you think
7
there's enough there to write a label. DR. GOLDEN:
8
In essence, this is the issue
9
that we always have to address when we have a new
10
drug that's facing us, where there's uncertainty.
11
And so, we certainly do the best we can with the
12
data that are in front of us.
13
information comes in, we make adjustments as
14
needed.
15
And then as new
But this is not unique to this drug.
I
16
mean, certainly, this is what we deal with, but
17
that's part of the question.
18
the question.
19
The risk/benefit is
So I'm going to turn it back to you.
DR. J. SMITH:
This is Jim Smith.
I can
20
just add to that a little bit.
21
programs, even larger than this one, cannot
22
reasonably be expected to identify all safety risks
A Matter of Record (301) 890-4188
I mean, even large
271
1
that may eventually turn up in the postmarketing
2
setting.
3
rely on continued pharmacovigilance of any drugs
4
that we approve.
5
turn that question back to you, as you discuss,
6
with regard to whether or not you think that the
7
safety of this product has been adequately
8
characterized, that you can come to a risk/benefit
9
decision.
10 11 12
And that's in part why we have to also
So I think that we are going to
DR. SMITH:
Dr. Wilson, did you have a
clarifying question? DR. WILSON:
I was interested in all these
13
people who started on the study drug at very low
14
cholesterols.
15
used in the IMPROVE-IT trial.
16
instance, the background information, page 10,
17
table 1, sponsor document, it gives the mean or
18
median baseline LDL cholesterol, and several of
19
these are very close to 100.
That means 50 percent
20
of the people are below 100.
And my guess is
21
probably a third are even below 70, yet they were
22
randomized to a trial.
And it's even lower than what were So if you go to, for
A Matter of Record (301) 890-4188
272
Then the other document is also the sponsor
1 2
document, which is CO-95, which certainly could be
3
projected.
4
is the people who get at 74, slide 74.
5
LDL cholesterol is below 25.
6
fit together.
7
how -- there are a lot of people with LDL
8
cholesterol less than 25.
9
there twice.
10 11
twice.
And then the third one, the third part, All this
All three of these
I am trying to figure out exactly
And some of them were
About 8 or 10 percent were there
Many more were there just once. Moving forward, is that one of the targets?
12
Is there a lower level at which a person's LDL
13
cholesterol -- say, even if they're at higher risk,
14
were not likely or we do not plan to have to
15
prescribe this medication to patients?
16
Most of the testimonials and most care, up
17
until the recent guidelines, has been
18
target-driven, but we're still hearing about
19
targets.
20
target-driven.
21
what Chris Cannon was reporting, we have a
22
tremendous number of people with an LDL cholesterol
The current guidelines are not so And most of all the trials such as
A Matter of Record (301) 890-4188
273
1
of 70 to 100.
2
this medication and a lot of them go below 25.
3
Yet, you've tried those folks on
So I guess that's the simplest question, is
4
moving forward, how are they handled now?
5
the planned trials, how is that going to be
6
handled?
7
DR. EDELBERG:
And in
Could you please give me the
8
demographic slide from the introduction to discuss
9
the baseline levels of the LDL in the target
10
population, the patients who we think will be
11
treated and get the most benefit from alirocumab
12
would be?
13
If I can get that up?
The patients we
14
studied in the clinical program, those with
15
familial hypercholesterolemia on maximally
16
tolerated doses of statin had a mean LDL of 150,
17
the maximally tolerated dose of statin, most of
18
them on additional lipid-lowering agents.
19
The high-risk population, the majority of
20
whom already had disease, half of whom had
21
cardiovascular events, those patients on
22
max-tolerated statins had an LDL of 110 and the
A Matter of Record (301) 890-4188
274
1
statin-intolerant population with LDLs approaching
2
200.
3
We intend that 75 milligrams be the usual
4
starting dose to help patients to achieve their
5
individual goals, recognizing that for many
6
patients, that goal may be, if you're very high
7
risk, to get below 70.
8
advocating for getting patients below 25.
9
our dosing regimen starting patients out at 75
We're certainly not In fact,
10
milligrams is aimed at being able to help patients
11
prevent those very low levels of LDL.
12
We reserved the 150 milligrams only for the
13
patients who need that extra efficacy because they
14
don't get to their individual goal on the 75, or
15
have such high baseline levels of LDL that they
16
start - we recognize that our data set with
17
patients below 25 milligrams of LDL is just
18
slightly over a year.
19
concern that we found, we certainly recognize that
20
that's a limited data set, and we've got a dosing
21
regimen to avoid those levels.
22
DR. BRAUNSTEIN:
And while there's no safety
All the patients in our
A Matter of Record (301) 890-4188
275
1
study were above 70.
2
High-risk patients had to be above 100.
3
risk patients had to be above 70.
4
study, we said, well, anybody above 70 can come in
5
as long as they're high or very high risk.
6
treated everybody with the highest dose because
7
that was our largest study where we were seeking
8
the greatest amount of additional safety data.
9
we wanted to explore in a controlled setting what
10 11
We had two cutoffs for entry. Very high-
But in that LT
And we
And
would happen with low LDLs. This is something that we don't intend in
12
actual clinical use of the drug.
13
recommending that the initial starting dose
14
actually be 75 for the vast majority of patients,
15
except for those who need the greatest LDL
16
reduction.
17
We're
But this was done intentionally, and we had
18
additional monitoring in place.
19
as part of our data monitoring committee who was
20
specifically assigned to monitor patients who
21
develop low LDL and to alert sites if there was in
22
fact a problem, if there was a concern to do
A Matter of Record (301) 890-4188
We had a physician
276
1
additional testing. So all of this was built into the program
2 3
because we felt that it was important to explore
4
this issue in a controlled environment of a
5
clinical study as opposed to just let it be worked
6
out after the drug might be approved. DR. WILSON:
7
Just partially to respond to
8
that, the number needed to treat, even for patients
9
on statin, adding ezetimibe, in the recent reported
10
trial, was a 2 percent delta, for the NNT, is 50,
11
for a person already on statin generally, with the
12
paper that came out last week.
13
clarification on that.
And we might --any
In addition, we've had other trials with
14 15
lipid-lowering agents with the dalcetrapib trial,
16
which was null, similar low LDL cholesterols and no
17
proven efficacy.
18
Dr. Cannon would be appropriate to clarify what's
19
going on.
So perhaps either you or perhaps
DR. BRAUNSTEIN:
20
The one point I want to
21
make, I just want to emphasize, is the genetics
22
here.
It was very different than what we had for
A Matter of Record (301) 890-4188
277
1
CETP inhibitors.
2
inhibitors is there's very strong genetic data,
3
that loss of PCSK9 function is actually associated
4
with lower risks of cardiovascular events over long
5
periods of time, 88 percent lower risk of
6
cardiovascular events over 15-year observation
7
periods.
And that's a very unique aspect.
I think we agree with the FDA.
8 9
What's unique about the PCSK9
One of the
points that they made was that when considering
10
whether or not LDL is a good surrogate, perhaps one
11
should also take into account what the mechanism of
12
action is.
13
where we have this strong genetic data, that
14
actually supports that this mechanism of action
15
will ultimately result in the kind of outcome that
16
we hope to see in the cardiovascular outcomes
17
study.
18
additional he'd like to add to that.
19
And for a product like alirocumab,
And maybe Dr. Cannon has something
DR. CANNON:
The design of IMPROVE-IT was
20
actually making sure that patients were at goal in
21
the control arm.
22
whether going even lower still, so getting around
And then we were trying to see
A Matter of Record (301) 890-4188
278
1
70 to 55, whether that would be beneficial on the
2
one hand, and then also examining ezetimibe. This program is built on everybody else who
3 4
is not at goal on a stable, maximally tolerated
5
dose.
6
many people were statin naive.
7
on lower doses, et cetera, so they weren't stable.
8
But we picked actually just the low to medium
9
cholesterol levels in order to make sure they were
So in the ACS setting that we looked at, Other people were
10
at goal so we could explore an even lower goal, and
11
then to see the benefit.
12
DR. SMITH:
Are these comments directly on
13
this point?
14
Dr. Gellar, in a couple minutes.
15
So Dr. Sager?
DR. SAGER:
We'll come back to you,
I have just a question for the
16
sponsor.
17
for physicians or patients who get to very low
18
LDL-C levels, 10, 15, 20?
19
anything like that in terms of your label?
20
Are you imagining any kind of guidance
DR. EDELBERG:
Are you imagining having
So as the sponsor, we have
21
not seen any safety signal, and we've developed a
22
dosing regimen that can allow patients to avoid
A Matter of Record (301) 890-4188
279
1
those levels.
2
guidelines that we have today actually provide some
3
important clarification there.
4
lucky that we can actually ask Dr. Eckel to
5
actually comment on that.
6
I think that, actually, the
DR. BRAUNSTEIN:
And I think we're
Before we do that, though,
7
can we show slide EL-50?
Because you mentioned two
8
levels that we can actually just avoid, because,
9
with the low dose, we see that effectively this is
10
patients getting to 15 or below.
With the low
11
dose, we don't get there, basically, as long as the
12
patients have an LDL greater than 80 to start with.
13
Only with a high dose would that be a concern.
14
So with using the starting dose of 75 for
15
the vast majority of patients, we're not going to
16
get down to 15.
17
patient, but it's going to be a very rare thing.
18
But perhaps Dr. Eckel could --
19
Of course, it could be a rare
DR. SAGER:
There probably will be people
20
who get started on this who start at 60 or
21
70 LDL-Cs, so we'll get that.
22
practice, that will happen.
A Matter of Record (301) 890-4188
In real clinical
280
1
DR. EDELBERG:
Right.
So this really comes
2
down to clinical guidelines, and I think that
3
Dr. Eckel would be optimal to discuss.
4
DR. ECKEL:
Since I was a member of the 2013
5
ACC AHA report, which shows no evidence for target
6
setting, in practice, we all set targets.
7
though it's not evidence based doesn't mean the
8
evidence says no.
9
Even
So let's take the arbitrary situation,
10
Dr. Sager, where in fact I'm going to start a PCSK9
11
inhibitor, specifically alirocumab, for someone
12
whose LDL cholesterol is 82 on maximum statin
13
therapy.
14
indicated every two weeks.
15
And I start the 75-milligram dose as
If the LDL cholesterol goes to 37, then I'm
16
going to be okay there despite the fact that the
17
guideline says levels under 40 should get the
18
attention of the physician, the prescribing
19
physician.
20
from 82 to 37 is going to be okay.
21 22
So my best judgment would say going
Now, if that level gets to 10 or 12, I'm going to back off.
So if the patient is on 75, I
A Matter of Record (301) 890-4188
281
1
will stop the drug and continue maximum statin
2
therapy, and maybe revisit it months later to see
3
if that very low nadir is again ultimately
4
achieved.
5
and my opinion only, but that's the way I would
6
deal with this in a clinical situation.
7
So I think that's a personal judgment
Now, if the patient were on 150 and at 20 or
8
25, then I would go back to the 75 dose, expecting
9
a rise of 15 percent of that LDL cholesterol.
10
DR. SAGER:
Let me just follow up, though.
11
But in real practice, would you expect some
12
physicians to instead back off on their statin?
13 14 15
DR. ECKEL:
No, that's not recommended, nor
would that be anything that I could ever recommend. DR. SMITH:
Dr. Orza, you had earlier some
16
clarifying questions.
17
clarifying question or two you want to ask?
18
DR. ORZA:
Do you still have a
One of them relates to a question
19
that one of the public speakers raised about the
20
demographics of the trials.
21
in what FDA made of the apparent differences in
22
efficacy in women, and also about the predominance
And I was interested
A Matter of Record (301) 890-4188
282
1
of Caucasians when we know that there are lots of
2
other -- in African-Americans, and Hispanics, and
3
other groups -- very high rates of being high risk
4
for heart disease.
5
I was wondering whether you thought that the
6
populations in the studies were representative of
7
and therefore generalizable to, one, the whole
8
population that potentially would need this drug,
9
the population that's at high risk for heart
10
disease, and, two, the very specific population of
11
people with FH.
12
DR. GOLDEN:
So just to talk about the issue
13
of women, it was noted that there was a small
14
difference in treatment effect between men and
15
women, where women were about -- I'm looking at the
16
long-term trial here because that was the largest
17
trial -- 9 percentage points difference so that
18
they were getting a lesser effect.
19
This small treatment effect difference was
20
seen consistently throughout the trials.
In some
21
trials, the interaction was significant.
In some,
22
it wasn't.
And it's hard to know what to make of
A Matter of Record (301) 890-4188
283
1
it, although, still over all, the treatment benefit
2
of LDL lowering appears consistent across both
3
sexes.
4
So the treatment effect difference exists.
5
It appears small.
And it's not quite clear why
6
it's seen at this point.
7
might be some room for some exploration there.
So I think that there
8
In terms of -- sorry.
9
DR. ORZA:
And the numbers were probably too
10
small, but on the safety side, were there any
11
differences between men and women?
12 13 14
DR. ROBERTS:
No, we didn't appreciate any
significant differences by gender. DR. GOLDEN:
In terms of the right racial
15
and ethnic demographics in the program, I agree.
16
mean, it's a limitation certainly that a majority
17
of patients were white in this program.
18
was a trial of about 300 patients, was done
19
exclusively in the United States, and the treatment
20
effect was similar in COMBO I as to the other
21
placebo-controlled trials.
22
I
COMBO I
So I think that it confirms the efficacy in
A Matter of Record (301) 890-4188
284
1
a more diverse population that's more
2
representative.
3
head, I think that there were 13 percent African-
4
American.
5
Hispanic, so a little bit more representative.
6
yes, I agree.
7
So thinking off the top of my
I think there were about 10 percent
It is a limitation.
DR. ORZA:
Is it as much of a limitation for
8
the FH population?
9
from the population as a whole or are they
10 11
But
Are they appreciably different
similarly diverse? DR. GOLDEN:
I don't actually know the
12
answer to that question.
Maybe the sponsor does.
13
But the efficacy was certainly similar in the FH
14
population.
15
trial, which I mentioned in my presentation, the
16
efficacy was similar in FH and non-FH patients.
with the exception of the high FH
17
DR. SMITH:
Dr. Thomas?
18
DR. THOMAS:
I just had a question,
19
actually.
In the planned outcomes trial, one of
20
the issues that we were looking at is
21
neurocognitive issues.
22
neurocognitive testing?
Is there any formal And what type of testing
A Matter of Record (301) 890-4188
285
1
is done if there is, and what age group?
2
think you're probably more likely, if you are
3
testing, to find deficits from any medication, if
4
there really is an effect in an older group versus
5
a younger group.
6
treated for many different age ranges.
7
I would
And this drug actually could be
DR. EDELBERG:
Dr. Braunstein's going to
8
review our neurocognitive testing assessment plan
9
in the ongoing outcomes trial.
10
DR. BRAUNSTEIN:
11
neurocognitive testing.
12
adverse events, and we're doing it in an enhanced
13
manner because we have neurocognitive events as an
14
adverse event of special interest, and we have a
15
dedicated case report form page around
16
neurocognitive so we can have enhanced collection
17
around events.
18
We're not doing formal Instead, we're collecting
But importantly, we've also put together a
19
panel of experts in neurocognition who are going to
20
evaluate the data as they come in, will ask us
21
about additional information that we should be
22
requesting about these events.
A Matter of Record (301) 890-4188
We'll be providing
286
1
periodic reports to the DSM V.
We will have
2
ability to look at the data on an unblinded basis.
3
And at the end of the study, we'll be able to
4
provide their analysis of the ultimate unblinded
5
data. So that's how we're proposing to look at
6 7
this.
And if we have a 50,000 patient-year study
8
and a 1 percent event rate, which would mirror what
9
we saw in our clinical program, we'd be able to
10
exclude a 30 percent difference between treatment
11
groups.
12
DR. THOMAS:
Yes.
I mean, from what was
13
presented, I thought that was probably what was
14
going to be done.
15
lifelong drug for many people, and I'm not sure if
16
that's going to address some of the concerns,
17
especially with the low LDL cholesterol, of
18
declines that may happen as people go over time.
19
You have to pick up the event.
20
recognized and then adjudicated, formal testing, as
21
is being done in other diabetes trials.
22
done in ACCORD, and it's being done in ASPREE,
But this is going to be a
A Matter of Record (301) 890-4188
It has to be
It was
287
1
which is an aspirin trial.
2
especially if you're using a rich population of
3
older patients.
4 5 6
It might be helpful,
So just something I'm curious if that's even being done in a subgroup or smaller subset. DR. EDELBERG:
In viewing the different
7
options, we focused on this at the recommendation
8
of experts.
9
on this approach as compared to dedicated
10 11
I'd actually ask Dr. Harvey to comment
neurocognitive testing. DR. HARVEY:
Certainly.
I'm Phil Harvey
12
from the University of Miami, Miller School of
13
Medicine.
14
methodology question.
15
up unless there was some concern that we are
16
missing neurocognitive events.
17
in the study to date is a very low number of
18
neurocognitive adverse events.
19
And this is clearly an assessment The question wouldn't come
What we're seeing
What we know is that if you add
20
neuropsychological testing, it's unlikely to
21
increase the sensitivity.
22
in mind the psychometric properties of
It's important to keep
A Matter of Record (301) 890-4188
288
1
neuropsychological tests. The most commonly used memory test, the
2 3
California Verbal Learning Test, second edition,
4
actually can only detect worsening in memory
5
performance at a level of 50 percent.
6
less than that is viewed as unreliable.
7
argument would be that a worsening in memory
8
performance of 50 percent is equivalent to the
9
change you see when you develop mild cognitive
10
Any change So my
impairment. Now, in mild cognitive impairment, the
11 12
person is brought in by their relatives, who have
13
noticed a substantial cognitive change, and then it
14
is evaluated.
15
test cannot even detect unreliable changes of less
16
than 10 percent in performance because the smallest
17
change an individual can have at delayed recall on
18
the California Verbal Learning Test is about 1 out
19
of 10 items because you can't learn less than one
20
word.
21
detect is 10 percent.
22
What happens is, the neurocognitive
So the smallest unreliable change you can
Group differences are not the same.
A Matter of Record (301) 890-4188
We are
289
1
looking at events in individuals, which may average
2
to some kind of group difference.
3
detecting memory change events, the instrumentation
4
requires a very substantial event that would be
5
detected by virtue of conversation with the
6
individual for the most part.
7
But as far as
So if we're talking about event rates that
8
may be half of a percent relative to placebo or
9
less, then we are talking about testing 999
10
patients with no adverse events to detect 1 AE.
11
it's a lot of testing.
12
required suggests that we're probably not missing
13
that many tremendously large changes in cognitive
14
performance in the existing database.
15
DR. SMITH:
16
DR. GELLER:
So
And the large change that's
Dr. Geller? I can't resist adding to what
17
you've just said, that the best way not to find
18
something is to not look.
19
(Laughter.)
20
DR. GELLER:
My question is probably first
21
for the FDA.
Could we recommend approval for FH
22
and not for the other subgroups?
A Matter of Record (301) 890-4188
And if we could
290
1
do that, what would be the implications for
2
outcomes and for the other subgroups?
3
DR. SMITH:
I'm wondering if we should defer
4
that because we're going to come back to that
5
question in a specific discussion in a voting
6
question.
7
Other clarifying questions?
8
DR. BLAHA:
9
Yes, Dr. Blaha?
I think CRP is a relevant thing
to discuss because, of course, statins lower CRP.
10
And it's been proposed in the JUPITER trial, for
11
example, the CRP-lowering effect of statins may be
12
relevant.
13
IMPROVE-IT study, if I read it right, there is a
14
mild reaction in CRP with an addition of ezetimibe
15
on top of statins, I think, if I'm clear, if that's
16
causal in any way.
All that is hotly debated.
And in the
17
I know others are investigating and
18
targeting the inflammatory pathway specifically as
19
a way of lowering events.
20
can maybe have Dr. Cannon comment just on the
21
possibility or what we've learned from IMPROVE-IT,
22
if CRP lowering correlated with the event
So I'm wondering if we
A Matter of Record (301) 890-4188
291
1
reduction, et al., if there are implications for
2
the specific LDL lowering with PCSK9s. DR. CANNON:
3
As you've noted, in IMPROVE-IT,
4
there was a .3- to .5-milligram-per-deciliter,
5
depending on the time difference, of ezetimibe
6
versus placebo on a background of statin.
7
analyses to come forth that event rates track with
8
your achieved CRP and LDL levels, as have been seen
9
in multiple trials.
We have
But we aren't able to attribute benefit to
10 11
any particular change, of any of the lipid changes.
12
So the changes in lipid lowering, I suppose, with
13
the different agents are the best we can do in a
14
randomized comparison. DR. BLAHA:
15
I'll follow up with one
16
question.
17
similar, at least with the up-regulation of the LDL
18
receptors, thoughts on why there may be
19
differential effects on CRP between statins and
20
PCSK9s?
21 22
I'm curious.
DR. EDELBERG:
Since the mechanisms are
I would remind you that, in
our program, the vast majority of the patients were
A Matter of Record (301) 890-4188
292
1
already on maximally tolerated doses of statin.
2
while I can't specifically comment on the
3
differences related to CRP regulation -- I think
4
that's a field of ongoing investigation -- I would
5
note that in addition to LDL lowering, PCSK9
6
inhibition actually reduces Lp(a) by 30 percent.
7
And we know that that is associated as a
8
cardiovascular risk factor.
9
be ongoing work to be done there.
10 11 12
DR. SMITH:
So
So I think there would
Dr. Orza, did you have another
clarifying question? DR. ORZA:
You'll have to bear with me
13
because I'm having difficulty even formulating this
14
question.
15
reading is that, in layman's terms, the mechanism
16
by which statins work is to interfere with the
17
body's ability to produce cholesterol, and that one
18
of the things we're seeing with statins is that the
19
body is compensating for that by doing other things
20
like increasing its ability to pull cholesterol
21
from the bloodstream, from what you're eating.
22
But my understanding from the background
So I'm wondering, if that's kind of been the
A Matter of Record (301) 890-4188
293
1
way that the body is adapting to statins, are there
2
any signals for how it would adapt to this other
3
mechanism, which this drug seems to act by, which
4
is to interfere with the body's ability to pull it
5
from the bloodstream?
6
and it would actually -- the body would compensate
7
by producing more cholesterol de novo?
8 9
So would it go the other way
Are there any kind of signals or is there anything in the data that would allow us to look at
10
that kind of compensatory --
11
DR. EDELBERG:
Statins increase the density
12
of LDL receptors, which helps the body to clear LDL
13
out of the bloodstream.
14
mechanism by which PCSK9 inhibition is able to
15
work, by increasing the recycling of the LDL
16
receptors back to the cell surface.
17
That's the same basic
I think one of the important parts to your
18
question is addressed by the genetics.
19
seen that patients with a loss of function in PCSK9
20
are able to get lower levels of LDL due to the
21
increased clearance of the LDL of the body, and
22
that's what we've seen leads to the reduction or
A Matter of Record (301) 890-4188
And we've
294
1
the lower level of cardiovascular events over the
2
course of time.
3
So I don't think that we are seeing a
4
compensatory mechanism, as you were concerned
5
about.
6
DR. SMITH:
Dr. Stanley, you had a question?
7
DR. STANLEY:
In the public commentary
8
session, one of the comments was that if these
9
drugs get approved too broadly, that it's going to
10
actually interfere with doing the studies to
11
demonstrate cardiovascular effect in this.
12
wonder if we get comments from both the sponsor and
13
perhaps also the FDA on that concern.
14
DR. EDELBERG:
I
We're working with all of our
15
outcomes sites in 50 countries around the world,
16
and all of the investigators, and the education to
17
the investigators, and to the patients themselves
18
about the importance of staying in the clinical
19
trial to help to answer this very important
20
question.
21
working towards already in the United States and
22
around the world.
So it's something we've already been
A Matter of Record (301) 890-4188
295
I would think that Dr. Cannon actually would
1 2
be an excellent person to comment on it because
3
he's just concluded a large outcome trial in which
4
the drugs were available. DR. CANNON:
5
Certainly, we do want more
6
evidence.
I think these types of things have
7
happened with other classes of drugs, positive or,
8
in the case of ezetimibe, a lot of negative things.
9
But within a trial, as investigators, we always
10
work with the patients and local investigators to
11
try to finish out the experiment that we've done,
12
to gather the information that everyone's set out
13
for us.
14
So it's certainly always a concern of the
15
changing landscape, but something that trial groups
16
are well-poised to try and respond to.
17
immediate access will differ at all the different
18
countries, as there's about 20, 25 percent in the
19
U.S., so that would leave the other 75 percent of
20
people around the world certainly still able to do
21
the trial.
22
big trials and do, I think, a reasonable job in
I think
But we face this all the time in doing
A Matter of Record (301) 890-4188
296
1
keeping them going in general. DR. STANLEY:
2
I guess the question is not
3
only from the problem of retention, but also of
4
recruiting new subjects because you're not -DR. EDELBERG:
5
Our recruitment rates remain
6
actually ahead of target here, so we're looking
7
forward to full enrollment fairly soon, so we don't
8
anticipate that to be a problem. DR. SMITH:
9 10 11
Would the FDA like to comment on
that? DR. J. SMITH:
I think it's something that I
12
raised in my introductory comments as a
13
possibility.
14
suspect that the sponsor won't do everything in
15
their power to encourage investigators and patients
16
to remain in the trial and to uphold trial
17
integrity, it's a risk.
18
large cardiovascular outcomes trials that have
19
utilized drugs that are on the market, that have
20
not had great success because of a crossing with
21
active therapy into the placebo group, narrowing
22
the difference between groups.
And although we have no reason to
And there have been other
A Matter of Record (301) 890-4188
297
So that can happen.
1
And I think that it's
2
important to recognize that is a possibility.
And
3
it's one of the reasons that we believe it's
4
incredibly important for this committee to
5
understand that if an approval is based on LDL
6
cholesterol, that is essentially a substitute for a
7
clinical outcome.
8
trial, we can't control that once the drug is on
9
the market.
And whatever happens to that
10
So I'm sure they'll do everything they can,
11
and we fully support that, but we can't predict the
12
future. DR. SMITH:
13
There are a couple more hands
14
from the panel, but if it's to discuss that point,
15
I'd rather defer that to the discussion that we're
16
about to undertake.
17
things before we enter discussion, I'm raring to
18
go.
19
If it's directed to clarifying
So yes, Dr. Sager? DR. SAGER:
Just a quick question.
Is the
20
outcomes study being run through an unrestricted
21
grant to an educational institution, academic
22
institution?
A Matter of Record (301) 890-4188
298
DR. EDELBERG:
1
Duke is the academic
2
organization leading it.
3
LDL reduction is very important, but proving
4
cardiovascular outcomes is why we've developed
5
this.
6
outcomes study is delivered and delivered on time.
7
Questions to the Committee and Discussion
8 9
We will deliver that.
And we're looking to make sure that that
DR. SMITH:
At this point, if acceptable to
the panel, I would like to proceed with the
10
questions, and the discussions, and the voting as
11
posed by the FDA.
12
objections.
13
questions to the committee and the panel
14
discussions.
15
And I don't see any major
So we'll now proceed with the
I'd like to remind public observers that,
16
while this meeting is open for public observation,
17
public attendees may not participate except at the
18
specific request of the panel.
19
So at this point, if we could project the
20
first of the discussion points, it's a discussion
21
question.
22
I will read it.
So this is number 1.
Discuss the safety of alirocumab as observed
A Matter of Record (301) 890-4188
299
1
in the clinical development program, and in your
2
discussion, comment on the following.
3
A.
Discuss your interpretation of the
4
safety data with respect to any adverse effects
5
related to diabetes, liver-related safety, muscle,
6
neurological, neurocognitive events,
7
hypersensitivity, immunogenicity as well as any
8
other concerns you may identify.
9
B.
Discuss the adequacy of the current
10
clinical database to characterize the safety of
11
alirocumab.
12
i.e., number of patients and duration of exposure,
13
the strengths, limitations of the study designs
14
themselves, and the generalizability of the trial
15
populations to the target population or populations
16
if approved.
17
C.
Consider the extent of drug exposure,
Discuss your level of concern regarding
18
the safety of achieving very low levels of LDL-C
19
induced by alirocumab.
20
So first of all, any questions from panel
21
members about the wording of these discussion
22
questions?
A Matter of Record (301) 890-4188
300
1
(No response.)
2
DR. SMITH:
So if not, if my proposal is
3
acceptable to the FDA, if we actually kind of
4
couple A and B.
5
interpretation of safety data, and as relevant to
6
that and separately, the adequacy of the current
7
clinical database because I think they go a little
8
hand in hand.
9
discussed A and B, because I think it's a little
And that is discussing
And I'd like to defer C until we've
10
separable and would make it a little clearer.
11
seeing nods.
12
I'm
So I'd like to open this for discussion and
13
comments from panel members about points A and B.
14
Dr. Hiatt?
15
DR. HIATT:
The question did come back to
16
us, didn't it?
William Hiatt.
Yes.
I think the
17
safety database is as one would expect from the
18
design of the trials that target LDL as a primary
19
endpoint.
20
exposure is relatively modest.
21
this drug, I think, certainly would be to approach
22
a wide population of patients who I do believe have
And therefore, the patient-years
A Matter of Record (301) 890-4188
And the intent of
301
1 2
an unmet need. Therefore, I would assume this is not a
3
niche market like approval of the drugs we talked
4
about a few years ago for homozygous familial
5
hypercholesterolemia, which is a very small market.
6
So I do think there are signals in the
7
database.
There may be a diabetes signal.
8
may be a liver signal.
There may be a
9
neurocognitive signal.
And I think that the
10
hypersensitivity and immunogenicities are
11
monitorable things.
12
There
The signals that are being highlighted today
13
obviously don't have hard, strong signals.
14
think that the approach to safety in any
15
development program is usually not hypothesis
16
directed.
17
requires acquiring a number of events.
It's usually experiential.
But I
It usually
18
So as the physician who spoke about the
19
neuropsychological tests talked about it, and I
20
think adverse event reporting may be a good way to
21
detect that, you just have to have a number of
22
events to know if there's a problem.
A Matter of Record (301) 890-4188
And
302
1
ultimately, it's the ratio of those events between
2
drug and placebo or drug and active control.
3
then we fall on the confidence intervals around
4
that to interpret whether we have a problem or not.
5
The history of statins is illustrative here
And
6
because, as was discussed earlier with
7
rosuvastatin, I think that advisory committee
8
meeting, really for the first time, highlighted the
9
diabetes risk.
It might have been there, adverse
10
event databases from other statin drugs, and then
11
that became a labeling issue, and you dealt with
12
it.
13
So I think now physicians and patients are
14
aware that that is a concern.
15
gets played out in terms of the overall benefit of
16
lowering LDL versus the risk of reducing diabetes
17
has been well-discussed in the literature.
18
And I think how that
So therefore, by definition, the larger the
19
database, certainly 50,000 patient-years will be a
20
fantastic database to make assessments of these
21
individual signals, which today may or may not be a
22
threshold to require a label to address them, that
A Matter of Record (301) 890-4188
303
1
may become so.
And I think you have a good ability
2
to address them if they were to arise. So these signals, to my interpretation of
3 4
the data, would not be negating the issue right now
5
other than I think we need more data to really
6
adequately address them.
7
about whether these signals are real or not.
8
would certainly want to see more data to sort it
9
out.
So I remain uncertain
10
DR. SMITH:
Dr. Sager?
11
DR. SAGER:
I'm overall of a similar
I
12
opinion.
13
or liver signals, but I think, as the FDA said
14
earlier, these can be handled through both
15
pharmacovigilance as well as the fact that there's
16
this large outcomes study that's currently being
17
performed, which will have a lot of randomized data
18
and a much larger data set.
19
There may be weak diabetes, or cognitive,
I think it's nice to see if there was long-
20
term data and long-term studies done, and the
21
data set is a big data set.
22
this overall reassuring.
So I actually found
A Matter of Record (301) 890-4188
304
1
DR. SMITH:
Dr. Everett?
2
DR. EVERETT:
Thanks.
Brendan Everett.
So
3
I think the core issue for me here is that there's
4
not enough exposure time to really be certain about
5
the possibility of adverse effects.
6
with the context of this being a very important
7
drug for a lot of people in the near future in
8
terms of reducing cardiovascular events.
9
it's likely to have enormously wide use.
10
And I say that
And so
The reason I say that is because I think
11
that we have certain hypotheses.
12
sponsor had certain hypotheses going into the trial
13
about adverse events that were important to monitor
14
for, and I think they've done a good job of doing
15
that with one caveat.
16
The FDA and the
That said, many adverse events are
17
unanticipated.
They happen, and they only happen
18
when there's a lot of patient-years of exposure.
19
And I think the example of diabetes being detected
20
in JUPITER in an 18,000-patient study with
21
rosuvastatin, which we thought was entirely safe,
22
is a good example of that.
A Matter of Record (301) 890-4188
305
So my caveat is that I think the things that
1 2
came to my attention are the same things that came
3
to Dr. Hiatt's attention, that maybe there's a
4
little whiff of diabetes, there might be a little
5
bit of concern about neurocognitive events, and
6
then the liver enzymes. But I would say that -- I'm not a
7 8
neurologist.
I'm not a geriatrician.
But memory
9
is just one domain of cognitive function.
There's
10
executive function.
11
which you could potentially, in a very reasonable
12
cost efficient way in a subset of patients, assess
13
cognitive function at baseline and at one or two
14
years of follow-up, nested within your outcomes
15
trial.
16
There's a lot of other ways in
Actually, my nickel would be on, you
17
actually improve cognitive outcomes, I mean, if I
18
were to have to bet.
19
worthwhile doing for that reason.
20
So it may actually be
I worry that just relying on adverse events
21
where people are unlikely to report to you, "You
22
know what?
My memory is shot.
A Matter of Record (301) 890-4188
I don't have any
306
1
short-term memory," they're going to hide that from
2
you.
3
relatives.
4
themselves if they can.
5
cognitive -- it could just be a questionnaire that
6
takes 30 minutes with a research assistant in a
7
subset of the population might be a reasonable way
8
to go.
9
They're going to hide it from their They're going to hide it from And some kind of formal
I've sat on this panel when we've talked
10
about meta-analyses of hundreds of thousands of
11
patients followed for many years, so there's never
12
an adequate data set to fully assess safety.
13
even after the outcomes trial, I worry that that is
14
going to be a persistent concern of this advisory
15
committee.
16
followed for a total of 50,000 person-years is
17
probably much better than about a tenth of that,
18
which is what we have now.
19
close my comments.
20
DR. SMITH:
21
DR. BURMAN:
22
So
Nonetheless, I think 18,000 patients
And with that, I'll
Dr. Burman? Thank you.
I generally believe
that the drug appears safe at the present time.
A Matter of Record (301) 890-4188
I
307
1
echo the sentiment that I'm concerned about the
2
lack of benefit on CRP, the fact that they mainly
3
studied Caucasians and less so in other ethnic
4
groups.
5
We haven't talked about special
6
circumstances.
7
medication and it lasts in their bloodstream for a
8
month or two and they become pregnant, or they have
9
surgery, or they're in the intensive care unit, or
10
they have some infection -- and I know the booklet
11
discussed things on hepatitis infections as a
12
theoretical possibility -- those are completely
13
unknown, what would happen under the influence of
14
this medication.
15
For example, if a patient is on the
I agree that there are potential
16
neurocognitive abnormalities, and I agree with more
17
specific analysis.
18
elevated glucose, and the unstable angina, and also
19
that the short-term studies are focused on a well-
20
maintained and monitored group.
21
approval, we use in a larger population, and the
22
effects are largely unknown.
I'm concerned about the
And after
There also seems to
A Matter of Record (301) 890-4188
308
1
be a lack of studies solely looking at the
2
medication itself against the placebo.
3
I don't think I heard anything about the
4
effect of the medication on LDL particle size
5
today, although there may be something in the
6
booklet that I missed.
7
of the studies, obviously we need longer-term
8
studies, and the question is, should that be in a
9
postmarketing group or not.
Then in terms of adequacy
To date, the studies
10
are promising, but appear less than 18 to
11
24 months.
12
know the long-term safety of medication when it's
13
approved, but it seems probable that the drug will
14
be safe over time.
15
DR. SMITH:
16
DR. ORZA:
And as everyone knows, we never really
Thank you.
Dr. Orza?
So my reaction is not so much to
17
any one of the safety issues or the adverse events,
18
but the sheer number of them.
19
about, especially since one of the concerns we have
20
is in how broad a population this drug might
21
ultimately come to be used, whether it serves the
22
big public health equation.
And I'm thinking
While we're busy
A Matter of Record (301) 890-4188
309
1
focused on lowering LDL cholesterol, what other
2
things are happening on a population basis?
3
increasing diabetes?
4
neurocognitive effects?
5
musculoskeletal disorders and the consequences of
6
those?
Are we
Are we increasing Are we increasing
I think this is where Dr. Thomas and
7 8
Dr. Geller were headed also.
When do some of these
9
or some combination of these become something that
10
we should not just monitor for as adverse events,
11
but we should actually actively study them?
12
beyond that, we should actually elevate them to
13
endpoints that we power the studies for.
14
would be my sort of overarching sense of the safety
15
signals. DR. SMITH:
16
do?
You raised a question.
18
about the answer to that question? DR. ORZA:
20
about the answer.
21 22
DR. SMITH:
That
So what do you think we should
17
19
And
Do you have thoughts
I don't feel capable of thinking
I'm going to just make a comment
that I want to insert now -- there's more comments
A Matter of Record (301) 890-4188
310
1
to come -- so that others can react or object.
But
2
I think it's important to note I agree with what
3
I've heard said so far.
4
important, from my perspective, that there have not
5
been observed any alarming adverse event signals.
6
And what I mean by that is either the nature of the
7
event, the frequency of the event, or the severity
8
of the event.
I also think it's
That is not disagreeing with anything that
9 10
was said about lingering concerns.
But very often,
11
in studies on a new drug, without having the answer
12
as to the ultimate significance of an adverse
13
event, there's very often one or more alarming
14
signals.
15
know what they are.
We call them a signal because we don't They're really not.
So again, I'm not disputing any of the
16 17
levels of concerns that perhaps vary with
18
neurocognitive versus some other potential adverse
19
events.
20
perspective, or worth noting, that there is no
21
standing-up specific adverse event that represents
22
a really major focus of concern.
But I think it's notable from my
A Matter of Record (301) 890-4188
And I think it's
311
1
worth noting that because that doesn't always
2
happen with new drugs.
3
DR. SMITH:
Dr. Budnitz?
4
CAPT BUDNITZ:
Dan Budnitz.
I agree with
5
your comment, that Dr. Smith just made about it is
6
reassuring, the data we've seen, about the lack of
7
signals for between 6 and 18 months of therapy.
8
think that's the key point, is that these are drugs
9
intended for years, maybe decades, of therapy, but
10
we really only have experience for maybe up to 18
11
months presented today.
12
I
So I think the attention for adverse events
13
might focus on those that we expect would require
14
years to accrue.
15
asked about adverse events when, for efficacy,
16
we're assuming that it takes years in time to
17
accrue these outcomes, and so we're looking for a
18
surrogate endpoint, and we need to guess that
19
surrogate for adverse effects, which we don't have.
20
Again, we're looking for signals, but again, the
21
kind of things to look for are those that will take
22
years and years to accrue.
I think it's kind of ironic to be
A Matter of Record (301) 890-4188
312
1
So the kind of things I have in mind are
2
like some of these neurocognitive slowly-developing
3
changes, maybe things like fat-soluble vitamin
4
deficiencies over time that might not become
5
evident in a six-month trial.
6
DR. SMITH:
7
DR. THOMAS:
And that's it.
Dr. Thomas? I agree with what everyone else
8
has said.
I think, for me, in general, from what's
9
been presented, the data looks relatively safe with
10
no alarming signals.
11
We just need more people to be studied.
12
specifically, to highlight some of the areas, in
13
diabetes, it looks like it could be real based on
14
the point estimates.
15
it's a small number.
16
However, there are concerns. And
It's not significant, but
The mechanism is unknown, so to have some
17
understanding of what the mechanism is.
18
think it's not related to insulin resistance.
19
I think what's important is we look at the JUPITER
20
data, and people developed diabetes.
21
subgroup analysis, those people who developed
22
diabetes actually died less than the people who
A Matter of Record (301) 890-4188
I would But
But in the
313
1
didn't because they were on rosuvastatin.
That
2
seems more a thing that we need to know in a larger
3
trial. The neurocognitive issue, the agents, the
4 5
same thing happened.
As part of ASPREE, we did
6
neurocognitive testing.
7
up very well.
8
have memory deficits.
9
or 80 patients as a first-year follow-up 3 people
And Dr. Everett brought it
People don't want to admit that they We picked up out of our 70
10
who clearly had a significant or cognitive decline
11
that made them actually not very functional.
12
Neither they admitted to it and neither did their
13
family recognize it. So sometimes, there, cognitive adverse event
14 15
reporting may not pick up on this.
The age of the
16
patients in the study are somewhat younger.
17
people with heterozygous familial
18
hypercholesterolemia -- I looked it up -- the
19
median age was 52.
20
was 60.
21
patients where you could pick it up by testing, but
22
my guess is, if you approve this, the useful ones
The
Overall, the rest of the study
You may not see these deficits in these
A Matter of Record (301) 890-4188
314
1
span to different populations, including much older
2
populations, where something, if it does show up,
3
will be more of an impact.
4
I do think there's something to be said for
5
doing formal testing.
I spent my time before I
6
went to medical school actually doing this in my
7
summers, so I was trying to do neurocognitive
8
testing.
9
younger people unless they have very severe
And it really doesn't pick it up in
10
deficits, but we probably need a way to mature this
11
and expand into older populations because they're
12
an impact for a long-term agent.
13
Overall, though, I think most of the data
14
seems really comfortable, but you need larger
15
numbers to really see what the safety risks are.
16
But I think what's important in the end is, there
17
are going to be some safety risks, and you have to
18
make a trade-off for balance.
19
If you don't know what the benefit is, other
20
than lowering LDL cholesterol, it's really hard to
21
make that decision.
22
terms of the current therapies, we know more about
And I think what we have in
A Matter of Record (301) 890-4188
315
1
statins, what the benefit is, and we can apply the
2
risks and explain that to a patient.
3
In isolated groups, it's familial
4
heterozygous/homozygous hypercholesterolemia, we
5
know what the risks are.
6
prove the benefits because of the small numbers in
7
some of these populations, I think you make an
8
assumption based on what is available.
And even if we can't
9
I think that in the end, that's really what's
10
been happening to -- we have to know -- in the end of
11
the outcomes study, we might have that information.
12
It's large enough to be powered for these outcomes.
13
But I think there are some additional things that we
14
could do to make this more comfortable of long-term
15
use with older populations.
16
DR. SMITH:
Dr. Nason?
17
DR. NASON:
Martha Nason.
I was going to
18
make two points, one of which was basically the
19
same as Dr. Budnitz made already about the time it
20
might take some of these safety signals to develop
21
if they really were real and that person-years
22
isn't always person-years.
I mean, there's one
A Matter of Record (301) 890-4188
316
1
thing, the number of people, but another thing is
2
the amount of follow-up.
3
where, when you've screened people to be in your
4
trial, they need to be relatively healthy and
5
functional, and you wouldn't necessarily expect to
6
see any safety signals that need time to
7
accumulate.
8
need longer time.
9
And if it is something
That first amount of time, you would
The other thing I wanted to bring up, which
10
is not something that has been discussed much in
11
the last few minutes, is the group of statin-
12
intolerant patients.
13
puzzled about the feeling that this is one of the
14
target groups that the sponsor would like to
15
address.
16
subset of that population that has actually been
17
tested in the one trial.
18
I'm still, I guess, a little
And yet, there seems to be a very small
The sponsor acknowledges that it was only
19
really people who were willing to try a statin.
20
And I understand why that was.
21
the people who had more severe reactions to statins
22
or problems with the statins -- this has never been
A Matter of Record (301) 890-4188
But that means that
317
1
looked at in that group, even though it sounds like
2
that would be a large part of who would be targeted
3
with the therapy, were it to be licensed.
4
So I think that the current clinical
5
database of safety is insufficient in those people,
6
in the people who are statin intolerant.
7
more data on that.
8
echoing what everyone else has said about really
9
needing more time, not just more person-years, but
10
We need
For everybody else, I am
more time.
11
DR. SMITH:
Thank you.
Dr. Blaha?
12
DR. BLAHA:
I agree with what everyone said.
13
It's been a great discussion.
14
want to agree with what you said, Dr. Smith.
15
actually impressed.
16
for such a potent therapy.
17
all these things could be hooked up, particularly
18
neurocognitive, and I think that for the reasons
19
mentioned, that you do need perhaps a longer
20
follow-up time.
21 22
But I specifically I'm
There is no major signal here I'm impressed.
I think
Diabetes doesn't concern me as much because we know how to talk to our patients about diabetes
A Matter of Record (301) 890-4188
318
1
as far as statins, and I think we put it as a net
2
benefit.
3
everyone, as we already know, that these medicines
4
will be probably lifetime therapies for these
5
patients, absolutely.
6
needed.
But I do want to, of course, remind
The long-term follow-up is
But I just will make a general comment in
7 8
agreeing with you.
9
because there hasn't been a major signal to this
10
I've actually been impressed
point with these potent therapies.
11
DR. SMITH:
Dr. Shamburek?
12
DR. SHAMBUREK:
I would concur with the
13
others who have said that there are little signals
14
with diabetes, liver, maybe that, and no major
15
ones.
16
The ones probably going into it initially was the
17
immunogenicity with 4.8 percent of patients having
18
the antibody or the anti-drug antibodies, which may
19
or may not be clinically important.
20
However, I think there's some unknown ones.
But I think, as clinicians -- I mean,
21
another concern is, which I think there is some
22
data and there's some other use of monoclonal
A Matter of Record (301) 890-4188
319
1
antibodies.
2
to be 20, 30, 40 years, will there be anything from
3
patients going on them for six months, and then
4
going off, and then deciding they need to come back
5
on.
6
set up a trial.
7
that would become an issue.
8 9
But as you were saying, if we're going
Issues like that, we just probably can't even We're just going to have to see if
The other issue, which in fact may or may not be as big of an issue for me, is the
10
neurocognitive events.
11
say, on statins, I see them occurring when their
12
LDLs are 300.
13
the drop.
14
And with my patients, I'd
I see them -- perhaps it's due to
Perhaps it's due to the statin itself.
We've heard how this all got going with the
15
gain of function mutations in PCSK9, and we've
16
heard about heterozygous FH.
17
look at another rare genetic disease like
18
abetalipoproteinemia, and those patients have no
19
LDL whatsoever.
20
Well, maybe we can
I follow a dozen or more of those patients
21
up into their 60s.
And the one side effect they do
22
not get is neurocognitive side effects.
A Matter of Record (301) 890-4188
And that
320
1
is probably the one.
2
other things.
3
other neuropathies if they're not treated, but
4
they're not getting them in a couple years and
5
that.
6
You can't say potentially
The fat-soluble vitamins give them
So we might not see such effects for
7
decades.
But I don't know.
The neurocognitive
8
could be the acute drop in someone who normally is
9
sitting with an LDL of 200 rather than having it
10
their whole life.
But I don't know if we can learn
11
any lesson from patients with no LDL, not ending up
12
with major neurocognitives.
13
DR. SMITH:
Dr. Cooke?
14
DR. COOKE:
David Cooke.
I just wanted to
15
make a brief comment about the diabetes question,
16
specifically, in that the small signal raising the
17
question about diabetes risk, I think, in my mind,
18
is at least balanced a little bit by the metabolic
19
labs that they did report out in that, at least
20
from a simple mechanistic standpoint, if the drug
21
were to increase insulin resistance through, for
22
instance, causing a fatty liver, you would expect
A Matter of Record (301) 890-4188
321
1
the HOMA-IR to go up, which it didn't.
And if it
2
was to cause decreased insulin secretion due to
3
fatty absorption into the beta cells, you would
4
have expected the fasting glucose level to go up,
5
which it didn't. So those are also relatively short time
6 7
frames that we're looking at, so I think clearly,
8
we still need to see what happens over longer time
9
frames.
10
But that in my mind does balance that
concern.
11
DR. SMITH:
Dr. Geller?
12
DR. GELLER:
I think one of the reasons
13
we're having such a discussion about the safety
14
issues is that, in some other drugs that haven't
15
worked out real well, very little signal of safety
16
issues was seen.
17
point -- lowest cholesterol, you know.
18
women, it did increase cardiovascular risk.
19
also torcetrapib, if you look very carefully, there
20
was a slight increase in blood pressure in the
21
early trials, but nothing to write home about.
22
I mean, estrogen is a case in But in But
So here we have these slight increases in
A Matter of Record (301) 890-4188
322
1
certain characteristics.
2
about a year, and that makes me nervous.
3
think it makes other people nervous, too, although
4
they may not have to put it quite the same way.
5
I think we need that longer-term follow-up to be
6
sure something awful isn't going to happen. DR. SMITH:
7
We have data for only And I
So
So I will endeavor to summarize
8
this part of the discussion.
And from the various
9
members of the advisory committee, what we've heard
10
is that, on the one hand, noting that there has not
11
been observed in the available data a marked signal
12
for a highly concerning adverse effect of the
13
drug -- and that's considered notable -- but multiple
14
members of the advisory panel have expressed the
15
opinion that the available data are not adequate to
16
really resolve potential concerns about adverse
17
events. That really comes from a second
18 19
concern.
It comes from several directions.
And
20
some of it is focused on specific potential adverse
21
effects of the drug that emerged from the data
22
that's available.
And that's not to say there are
A Matter of Record (301) 890-4188
323
1
strong signals, but they emerge from the data
2
that's available.
3
a concern about diabetes, about neurocognition,
4
about immunogenicity, and about -- sort of separate
5
from that but related, just potential problems
6
related to very long-term exposure to a monoclonal
7
antibody, for example.
And that includes, for example,
8
The point has also been made that,
9
often -- or at least there is a risk that adverse
10
effects of a drug like this may really not be
11
anticipated mechanistically, and they may not be
12
evident from relatively small databases.
13
only become evident when there's both a lot of
14
patient experience and longer-term patient
15
experience.
16
And they
So both of those circumstances need panel
17
members to express strong opinions that we need
18
additional safety data.
19
ongoing clinical trial is likely to have
20
substantial input on those points, and that's a
21
major reason for continuing and completing that
22
trial certainly.
It's noted that the
A Matter of Record (301) 890-4188
324
1
It's also noted that that trial itself is
2
really unlikely to resolve all of these concerns.
3
And part of that, even though it's a large number
4
of patients, potentially a larger number could
5
reveal more information.
6
made that patient-years in terms of lots of
7
patients with a relatively short number of years of
8
exposure is different from a very long number of
9
years of exposure, which is what's anticipated for
But also the point was
10
a drug of this type, that patients may be on it for
11
many years.
12
So that really leads again to an expression
13
of the view that it's going to be important to have
14
ongoing monitoring programs in some way that may
15
enable one to discover, if they occur, adverse
16
effects of drugs that occur only after many years
17
of exposure.
18
In regard to the ongoing clinical trial, the
19
point was made that when we observe adverse effects
20
of the drug, inevitably, judgments on the use of
21
that drug are also influenced by the benefits and
22
the efficacy, and that a trial that has the
A Matter of Record (301) 890-4188
325
1
potential to demonstrate efficacy in terms of
2
cardiovascular outcomes instead of just LDL
3
cholesterol lowering would be extremely helpful in
4
terms of navigating that situation, if there are
5
emerging increased concerns about adverse effects
6
with more experience.
7
it's so important to complete that trial and
8
perhaps other trials.
9
So it's another reason why
In regard to that trial, and others, the
10
point was made that there are certain patient
11
groups that at this point are much less represented
12
and very much inadequately represented in the
13
available database, and that includes various
14
different racial groups.
15
who may be statin failures, rather than just having
16
an underlying disorder that brings them into the
17
treatment category.
18
those patients if they come up from the sponsor
19
would be again invaluable, important, and helpful
20
to patients.
21 22
It also includes patients
So opportunities to study
It also was noted that there is very limited knowledge about certain patient groups or clinical
A Matter of Record (301) 890-4188
326
1
situations, and that it's important to remember
2
that there are things that are not known from the
3
available data, and that there could be
4
undiscovered very important findings in that
5
context.
6
pregnancy in people who might be taking the drug,
7
but also intercurrent illnesses that may be
8
unrelated to the reason for taking the drug.
9
they're simply not needed to inform us about what
And that includes the occurrence of
But
10
would happen with perhaps a severe illness of
11
another cause altogether.
12
So I come to the end of my summary.
13
anyone like to make some corrections, or additions,
14
or subtractions from this?
15
(No response.)
16
DR. SMITH:
Would
So I hope that's helpful to the
17
FDA for the record.
So what we're going to do now
18
is we need to take a break on schedule.
19
the middle of this discussion question.
We're in
20
So we're going to take a 15-minute break,
21
but that'll give you 15 minutes to think about C,
22
recalling that there should be no discussion of any
A Matter of Record (301) 890-4188
327
1
of the matters before this advisory committee
2
outside of the formal advisory committee meeting,
3
i.e., during your break. So please, everyone, back here in
4 5
15 minutes.
That's 3:28 because I'm 15 minutes
6
ahead here.
Thank you.
(Whereupon, at 3:13 p.m., a recess was
7 8 9
taken.) DR. SMITH:
Again, welcome back, everyone.
10
We're going to pick up where we left off.
11
reminder, we're on discussion question 1, and we're
12
on C, which as a reminder I'll just briefly reread
13
that one, which is, discuss your level of concern
14
regarding the safety of achieving very low levels
15
of LDL-C induced by alirocumab.
16
you were going to comment on that?
17
DR. WILSON:
As a
And Dr. Wilson,
Thanks very much.
First, I
18
wanted to preface my comments by the idea that it's
19
pretty well accepted in laboratory science that low
20
levels of LDL cholesterol are difficult to measure
21
in many hospital laboratories.
22
concern of the confidence of the measurements.
A Matter of Record (301) 890-4188
There's some And
328
1
research projects and trials are using higher
2
levels, often ultracentrifugation, for measurement
3
of LDLs.
4
difficult to tell an LDL of 40 versus 60 in some
5
cases, for instance.
So that's number one.
So it may be
The second is, just, I don't think we have
6 7
enough safety data here.
And I'm not sure
8
measuring just the lipids and/or vitamins that are
9
fat-soluble vitamins, that are carried with the LDL
10
molecules -- and these have been mentioned a few
11
times.
12
Vitamins A, D, E, and K, those four, are
13
carried as fat-soluble vitamins on the apo B
14
molecule.
15
clinical outcomes related to deficiencies of these.
16
And a short-term study, as Dr. Geller and others
17
have remarked, is that really going to pick it up?
18
And we should reflect on what are the
So vitamin A can be associated with color
19
vision differences, retinitis, pigmentosa.
20
Vitamin D deficiency, we measure 25-hydroxy D, but
21
there's an increased risk perhaps for osteoporosis
22
with borderline D deficiency and especially in
A Matter of Record (301) 890-4188
329
1
people with darker pigmentation. Then finally, vitamin K -- there's less for
2 3
E -- would be concerned with interaction with
4
anticoagulants that work on this mechanism,
5
especially warfarin.
That's all I wanted to say.
6
DR. SMITH:
Dr. Hiatt?
7
DR. HIATT:
The approach to this question, I
8
think, would be what's the benefit and what's the
9
risk of very low levels.
In addition to the risk
10
Dr. Wilson described, there may be other unknown
11
risks.
12
And we've certainly changed our threshold for
13
lowering LDL cholesterol from 130 to 100 to 70, and
14
now IMPROVE-IT, by taking it down to 50.
15
But I guess the question is, why go there?
But we'll have sparse data at 25.
And so as
16
far as data on risk and benefit, it seems to me
17
appropriate to make an arbitrary cutoff of some
18
number, unless there's a compelling reason to think
19
that 15 is going to be 25 and 25 is going to be 50,
20
in which case of course you may want to go as low
21
as possible and assess that risk/benefit.
22
the absence of that information, if the phase 3 is
A Matter of Record (301) 890-4188
But in
330
1
not designed to test that, if the dosing strategy
2
is not designed to go that low, then it seems that,
3
at this point in time, kind of an arbitrary lower
4
bound would be a reasonable thing.
5
DR. SMITH:
Dr. Burman?
6
DR. BURMAN:
7
DR. SMITH:
Thank you. Excuse me, Dr. Burman.
Before
8
that, Dr. Hiatt, do you have any thoughts about
9
what that arbitrary lower bound should be?
10
DR. HIATT:
Not really, but the number 25
11
has been projected.
12
arbitrary cut point, I guess I'd stick with it,
13
since that seems to be a number that's been
14
incorporated in the experience to date.
15
DR. SMITH:
16
DR. BURMAN:
And seeing as it's an
Sorry, Dr. Burman. No problem.
Thank you.
There
17
is little or no compelling evidence that a lower
18
LDL concentration is detrimental as compared to a
19
"normal" LDL level.
20
patients with an MI have a "normal" range LDL,
21
either indicating that there are other mechanisms
22
involved or we should focus not on an LDL "normal"
Approximately 50 percent of
A Matter of Record (301) 890-4188
331
1 2
range, but on an optimal range. Babies, for example, are born with an LDL of
3
approximately 40 to 70 milligrams per deciliter,
4
and apparently mainly environmental influences
5
raise the level over time.
6
was mentioned, showed that a lower-than-normal LDL
7
level was beneficial with regard to CV events.
8
Preliminary evidence with regard to PCSK9
9
inhibitors also suggest the cardiovascular benefit
10
The JUPITER study, as
of a low LDL.
11
Evidence suggesting that a lower LDL is
12
associated with fewer cardiovascular events and
13
that LDL is generally a reliable surrogate for CVD
14
include the following:
15
different cultures; clinical trial data such as the
16
JUPITER study, which decreased LDL from
17
approximately 100 to 40 milligrams per deciliter;
18
genetic disorders with deficiency of PCSK9, as
19
noted, don't have demonstrable CVD; and also,
20
statins and PCSK9 inhibitors have a similar
21
mechanism of action, it appears.
22
epidemiologic data across
Lipoprotein(a) also decreases with PCSK9
A Matter of Record (301) 890-4188
332
1
inhibitors.
2
intensity should depend on the patient's risk for
3
arteriosclerotic vascular disease, and it should be
4
remembered that the duration of decreases in serial
5
LDL concentration will likely play a role in the
6
extent of effect on decreasing cardiovascular
7
disease. I do believe that longer-term studies
8 9
However, the degree of therapeutic
assessing adjudicated CV events need to be
10
performed, and I'm concerned that CRP apparently
11
does not decrease with inhibition of PCSK9.
12
torcetrapib and niacin stories give us pause, but I
13
believe there are alternative explanations for
14
these results perhaps relating to alternative
15
events.
16
The
I am also concerned about the clinical data
17
from patients with abetalipoproteinemia.
However,
18
I'd like to quote Goldstein and Brown, who received
19
a Nobel prize for their work, when they were
20
discussing PCSK9 inhibitors.
21
lesson of PCSK9 is clear.
22
88 percent reduction in incidence of coronary heart
And they said the
"If we are to attain an
A Matter of Record (301) 890-4188
333
1
disease, we must lower LDL levels well before
2
atherosclerosis has become advanced." Lastly, Steinberg and Witztum have suggested
3 4
that the goal LDL level, based on their evidence,
5
and the historical evidence should be 50 milligrams
6
per deciliter.
Thank you.
7
DR. SMITH:
Dr. Thomas?
8
DR. THOMAS:
I think what I would say is,
9
the question reads, discuss your level of concern
10
regarding the safety of achieving very low levels
11
of LDL cholesterol.
12
that's not in a way that, really, what we should be
13
thinking about.
14
achieving very low levels of LDL cholesterol.
15
there's no point in achieving those levels in
16
safety if there's no benefit to it.
I'm actually going to say that
We don't know there's a benefit of And
I don't think you can actually do a trial
17 18
because you're not really going to try and get
19
people down to below 25 at this point.
20
that'll happen in the future.
21
on what we have, that's really now where we're
22
going.
A Matter of Record (301) 890-4188
Maybe
But currently, based
334
I think the real issue is, how do we monitor
1 2
people, so how frequently so they don't achieve
3
very low LDL cholesterol levels.
4
dose-adjust or how do we make decisions on this?
5
Now, in the sponsor slide that they showed during
6
the clarification this afternoon, they showed a
7
slide of less than 15 achieved twice.
8
most part, people on the lower dose did not achieve
9
that.
And how do we
And for the
But if you looked all the way to the left,
10
in the group that had cholesterol between 70 and
11
less than 80, a little over 10 percent on the graph
12
did have LDL cholesterol less than 15. That's not an insignificant amount.
13
And
14
though Dr. Eckel said we're not supposed to use
15
levels, which is true.
16
that.
17
event with LDL cholesterol of 75, I would probably
18
use another agent to try to get them lower.
19
10 or 11 percent, there is a monitoring issue of
20
how often you have to do this.
21 22
We're all still guided by
And if I had someone who had a repeated
And at
So it's really not achieving low levels because you have an outcome that's your benefit.
A Matter of Record (301) 890-4188
335
1
It's really what's the safety concern or prevention
2
of reaching those levels.
3
DR. SMITH:
So I might make a comment at
4
this point also, which is that I agree with much of
5
what I've heard from other panel members.
6
that an important point, which has come up earlier
7
in the discussions today about a concern of
8
achieving very low levels of LDL cholesterol, is
9
what response that will then generate in the actual
10 11
I think
care of patients. So a concern is that decisions might be made
12
about how to adjust medications perhaps to avoid
13
such a low or to raise slightly the LDL cholesterol
14
level, which have the potential to have adverse
15
consequences for patients.
16
concern that one would like to try to find a way to
17
address.
18
And so I think that's a
A big part of what I'm referring to is the
19
potential of decreasing statins in order to
20
accommodate a continuation of this medication when
21
the LDL cholesterol is of concern, whether there's
22
data to back up that concern or not.
A Matter of Record (301) 890-4188
And so again,
336
1
this is hypothetical.
2
would happen and how it would happen, to me that's
3
a concern that potentially could be addressed in a
4
way, in an advisory way, because there's not really
5
data to back it up.
6
respond to very low LDL cholesterol levels, perhaps
7
expert advice in a sense.
8
Dr. Thomas?
9
DR. THOMAS:
But in terms of whether it
And that is advice on how to
This is Abraham Thomas.
Let me
10
just add to that.
11
that.
12
lipidologists, and endocrinologists, and
13
cardiologists, people that take care of these
14
patients wouldn't do that.
15
practices that aren't going to follow those type of
16
guidelines or recommendations.
17
Right?
So Dr. Eckel said he wouldn't do
And I think experienced
However, we see
I still remember very clearly a transplant
18
nephrologist when I was at the Brigham's saying, "I
19
can reduce the steroid doses so they don't get
20
diabetes," and all of us diabetologists were aghast
21
and said, "No, no, we can treat diabetes.
22
have someone reject their kidney because you don't
A Matter of Record (301) 890-4188
Don't
337
1
want them to have diabetes.
2
priorities wrong."
You're putting the
3
So someone might make a decision, without
4
the expertise of some of the people in this room,
5
of cutting the statin, and that may not be the
6
right choice. So just to highlight, just because we think
7 8
it shouldn't happen, we know when it gets out into
9
practice, it may happen much more often than we'd
10 11
like. DR. SMITH:
That's right.
And that's where
12
there really might be an opportunity for kinds of
13
guidance.
14
because I understand that that's not going to be
15
based in data, because we don't have those data.
16
Dr. Blaha?
17
And I almost hate to have to say that
DR. BLAHA:
Yes, lots of great points.
You
18
certainly can get a measurement there at these low
19
LDL levels.
20
measurement error is an underestimate, usually, by
21
direct ultracentrifugation of the lipid at a higher
22
LDL at very low LDL evolves compared to a
And it should be noted, usually a
A Matter of Record (301) 890-4188
338
1
calculated LDL.
That actually might be a little
2
bit reassuring.
Well, it might not be quite as low
3
as the clinicians might think, although maybe it
4
might cause alarm.
5
With regard to this exact question -- and I
6
think it's worth reading it again, discuss your
7
level of concern regarding the safety of achieving
8
very low levels -- to me, it's entirely dependent
9
on benefit/risk ratio, as was said.
It depends on
10
the risk of the patient.
I'm not as concerned
11
about low LDL levels in the very high-risk
12
patients, but I would be very concern about the
13
patient who's not so high risk, statin intolerant,
14
for example.
15
DR. SMITH:
Dr. Sager?
16
DR. SAGER:
Yes.
I'm also concerned about
17
the possibility that very low LDL-C levels will
18
result in people just trying to get them up a
19
little bit by reducing the statins.
20
have real data that the statins really can result
21
in reductions in cardiovascular morbidity and
22
mortality.
A Matter of Record (301) 890-4188
And then we do
339
1
So I think that's something that would have
2
to be really carefully addressed, if the drug is
3
approved, in the label and with some kind of very
4
strong wording, because I think until there's
5
outcome data, I don't think we can make the
6
assumption that -- as much as we are really excited
7
about these LDL-C reductions -- that they do result
8
in reductions of cardiovascular mortality until we
9
have the data.
But this is one place where if
10
someone has a low LDL-C level, it could result in
11
getting a statin reduced or stopped, and that could
12
be deleterious.
13
DR. SMITH:
Other comments on this point?
14
And again, now, just to the FDA, as we're
15
discussing these points, I know you'll do this, but
16
if you feel that we're not addressing some issues
17
that you would like us to comment on, please don't
18
hesitate to ask us to do that.
19
Yes, Dr. Shamburek?
20
DR. SHAMBUREK:
I was just going to
21
reiterate what they were saying.
22
us are comfortable and, actually in the past,
A Matter of Record (301) 890-4188
I think many of
340
1
dreaming of getting to an LDL of 50.
2
below that is an unknown.
3
are not at the utmost risk may not be, until we
4
have outcome trials, something that's warranted.
5
And it's not that it's not going to be beneficial.
6
But anything
And putting people who
The one other comment I was just going to
7
make is for a typical patient who does have an LDL
8
of 25, if you measure things like vitamin E -- it
9
is as we heard the only carrier -- in many or most
10
cases, you're going to get low-level vitamin E and
11
think many patients are vitamin deficient, much
12
like, if you go with low albumin, and get a low
13
level of calcium and want to give more.
14
So I think it's going to be a difficult
15
issue to look at, but not that we don't want to,
16
and not that it's not going to be an important
17
issue when you're on it for three years, five
18
years, or something like that.
19
DR. SMITH:
So I can try to summarize what
20
we've heard in the discussion, which is that it's
21
acknowledged that there is little or no evidence
22
that low LDL levels, low LDL cholesterol levels,
A Matter of Record (301) 890-4188
341
1
are harmful.
2
members in general comfort that there might not
3
indeed be adverse consequences of very low LDL
4
levels.
5
But that does not give the panel
As a potentially somewhat modifying factor
6
but also a complicating factor, it's been noted
7
that lower LDL cholesterol levels are difficult to
8
measure.
9
interpreting numbers.
And so it's a reason for caution in On an optimistic side, one
10
might recognize that may tend towards being an
11
underestimate rather than an overestimate of
12
levels, so at least some slight comfort in terms of
13
the inaccuracies of measurement strategies.
14
At the same time that there is little or no
15
evidence that they're harmful, there is concern
16
about monitoring for in fact potential
17
consequences.
18
other potential adverse effects of the drug, these
19
represent concerns not only about acute effects of
20
very low levels, but very long-term effects of very
21
low levels.
22
And as we've heard in regard to
So that needs to be an open consideration in
A Matter of Record (301) 890-4188
342
1
that there could be unanticipated adverse effects,
2
as well as things that might be closer to being
3
anticipated.
4
that one might link to low levels of vitamins A, D,
5
E, and K, with things like color vision changes, or
6
changes in bone mineral density, or changes that
7
may affect anticoagulant activities and
8
clotting-system properties.
And among that group would be things
9
The point was made that in an ideal world,
10
we should be looking at the question of benefit of
11
very low LDL cholesterol levels as well as the
12
potential adverse effects.
13
really have data at this point that convincingly
14
enable us to evaluate that question.
15
And again, we don't
The concern was raised by a number of panel
16
members about the potential for response to very
17
low LDL levels, which will occur.
18
project in patients in clinical practice based on
19
the data that are available, and concern about the
20
potential that there might be a substantial number
21
of patients or physicians who might respond to that
22
by lowering levels of statins, which have of course
A Matter of Record (301) 890-4188
We should
343
1
a very much stronger database for clinical outcomes
2
benefit.
3
this drug, that's of course that one would not
4
recommend.
5
So in the absence of similar data for
So without being very specific about that,
6
the point was made that careful consideration
7
should be given to mechanisms of trying to inform
8
users about how to respond to and manage a
9
circumstance of very low LDL levels and to avoid,
10
until further information might be available,
11
lowering levels of statins.
12
That can potentially be addressed
13
significantly through what's done with labeling.
14
That's really an FDA issue about what they can and
15
should do there.
16
could be addressed potentially outside the FDA,
17
perhaps encouraged by the sponsor in terms of
18
information that might incorporate expert opinion
19
and distributed in various ways to the user
20
community.
But it's also something that
21
Would anyone like to make additions, or
22
modifications, or subtractions to that summary?
A Matter of Record (301) 890-4188
344
1
Yes, Dr. Albrecht? DR. ALBRECHT:
2
Helmut Albrecht.
I would
3
like to actually make a statement just reflecting
4
on the three points that we've just discussed.
5
before I do, though, I'd like to make another
6
observation.
7
or these committees are not always of the same kind
8
of quality.
9
I am very impressed with the quality of the data
But
The data that comes to this committee
And I wanted to actually mention that
10
that were presented here by the sponsor.
11
a very impressive program.
It's been
But the reflection on those three points is
12 13
this.
I am worried that we might sit here one or
14
two years from now, when we get the large study
15
results, and we may still have the same questions.
16
Fifty thousand patient-years, that's just 10 times
17
as much we have now.
18
compare it to market experience in many, many new
19
and more patients.
It's not that much if you
20
So I think we should not lose sight of the
21
fact that the true safety picture may not come out
22
until we've exposed many more patients to this drug
A Matter of Record (301) 890-4188
345
1 2
than just 50,000 patient-years. The other aspect is, signals don't just come
3
out sometimes until millions have been exposed.
4
But no drug that has a meaningful effect is
5
entirely safe.
6
safety concerns we legitimately have here against
7
the benefit, both in terms of what the data have
8
shown and what we've heard from patients.
9
And I think we need to weigh the
So I think the benefit/risk ratio in the end
10
needs to make the decision.
11
number of tools available to manage risk later on,
12
one being of course the labor put into the product,
13
the other being postmarketing surveillance.
14
again, let's just consider those tools as well in
15
making a decision on this product.
16
DR. SMITH:
And the FDA has a
So
I think those are very good
17
points and we may be able to come back to the issue
18
of postmarketing later.
19
fits in terms of how we're going to discuss these
20
questions.
21 22
We'll see if that actually
One thing I forgot to mention in that summary, that I think is important, is that -- it
A Matter of Record (301) 890-4188
346
1
may seem obvious, but it's worth stating that since
2
we don't know whether there are or what the
3
consequences of very low LDL levels are, that it is
4
then very difficult to specify a level that is a
5
level of concern as compared to one that isn't.
6
I pressed Dr. Hiatt on that a little,
7
knowing it was going to be hard to answer that
8
question.
9
to approach problems like this, I think this
But again, as the FDA thinks about how
10
committee, which includes certainly lipid experts,
11
has not been able to move from available data and
12
make a really helpful comment on what level of LDL
13
is of concern and what level isn't.
14
a perspective as one then tries to approach the
15
challenge of how to manage this.
16 17 18 19 20
So that's just
So I think we should move on to question 2. Yes, Dr. Thomas? DR. THOMAS:
Actually, this is more of a
clarification from the FDA about this. Dr. Smith, in your presentation this
21
morning -- unless I misheard you -- you said you
22
couldn't ask the sponsor to do an outcomes trial,
A Matter of Record (301) 890-4188
347
1
and you were glad that they were going to do it.
2
thought you now have the ability to have outcomes
3
trials and to make the sponsor do them
4
post-approval, as is being done with some agents,
5
as opposed to before, where you could ask it, but
6
they didn't have to follow your directions.
7
DR. J. SMITH:
So we could ask for a
8
cardiovascular outcomes trial to confirm efficacy,
9
that is, to confirm a cardiovascular benefit if we
10
are working under the accelerated approval
11
paradigm, which we are not discussing today.
12
think what you're referring to is our ability to
13
ask for postmarketing cardiovascular outcomes
14
trials for safety, for example.
15
I
I said this morning -- but I'll
16
reiterate -- a postmarketing cardiovascular
17
outcomes trial that's being designed for safety is
18
because one has a concern that the drug could
19
increase the cardiovascular risk.
20
concern, that this drug could actually increase
21
cardiovascular risk, I'd have a hard time
22
recommending approval based on the surrogate when
A Matter of Record (301) 890-4188
If we have that
I
348
1
the intent of the drug is to reduce cardiovascular
2
risk.
3
For other safety signals for which the
4
cardiovascular outcomes trial is obviously a
5
placebo-controlled trial of a lot of patients and
6
will accrue a lot of patient-years, that trial may
7
provide an acceptable platform to evaluate some of
8
those concerns, if we determine that those concerns
9
meet the statutory requirements for our requiring
10
of a postmarketing clinical trial.
11
entirely possible.
12
very carefully, and we'll have to obviously discuss
13
that with the sponsor.
14
And that's
But we'll have to navigate that
So it's possible that safety concerns that
15
have either been identified today or that we've
16
identified may meet that statutory standard for
17
which we could require a postmarketing clinical
18
trial of some sort.
19
to use that ongoing trial as a platform to answer
20
those questions.
21
going to have to navigate.
22
And the sponsor might be able
But that's something that we're
What I wanted to be very clear about is that
A Matter of Record (301) 890-4188
349
1
establishing cardiovascular benefit with a
2
cardiovascular outcomes trial, it would really not
3
be the intent of those studies.
4
would look at benefit/risk at the end, and we're
5
always evaluating benefit/risk.
6
distinction that I was trying to make.
7
DR. THOMAS:
Yes.
Obviously, we
But that's the
I was really getting to
8
the other side effects, because, clearly, if there
9
was a cardiovascular risk signal, we wouldn't have
10
the meeting today because then there's no point at
11
all for giving the drug.
12
else shows up, then you could have the ability,
13
because I think that will be important for the
14
discussion about it.
15
DR. J. SMITH:
But whether something
And that has certainly
16
happened with other drugs that we've even approved
17
recently, where there is an ongoing large trial.
18
And during the preapproval review, the advisory
19
committee or we identified other safety concerns.
20
And we're able to build that in as a prespecified
21
look, for example, into an ongoing blinded trial.
22
But the purpose, though, still, would be to
A Matter of Record (301) 890-4188
350
1
evaluate that safety concern in this ongoing trial.
2
It wouldn't be tied to cardiovascular benefit. DR. SMITH:
3
Dr. Hiatt, you wanted to comment
4
also on the same point or question on the same
5
point?
6
DR. HIATT:
I do, and I was going to bring
7
this up with number 2, but let me just make a few
8
comments.
9
I think we're being asked to make somewhat
10
of an unfair choice.
And the reason is that it
11
sounds like accelerated approval is off the table.
12
But the only issue here is whether you can force
13
the sponsor to follow through with a cardiovascular
14
outcomes trial or not.
15
So it's a dichotomous outcome here.
So if
16
we say, yeah, approve it on the surrogate, because
17
that's what you've always done and it worked out
18
pretty well for statins, then I understand your
19
concern that we're left with a lot of residual risk
20
about safety.
21
you're stuck with the drug on the market with
22
unclear risk and benefit.
If the outcomes trial derails,
A Matter of Record (301) 890-4188
351
If we say no, then they'll come back in two
1 2
years.
And what essentially no means in my mind
3
would be that patients should have a two-, two and
4
a half-year window of missed opportunity.
5
that's assuming the cardiovascular outcomes trial
6
is positive.
7
negative, we were really smart.
8
and the trial was positive, I mean, there are a lot
9
of patients that just talked to us a couple hours
Now,
If we say no, and the trial is But if we say no,
10
ago that have an unmet need that we're denying
11
access to another agent. So I'm a little stuck on this point, quite
12 13
frankly.
If you could just -- the criteria for
14
accelerated approval, it fits beautifully with this
15
drug.
16
if you were to say just do it, and we want you to
17
agree to hold to that, it would sort of resolve a
18
lot of difficulties I'm having with the ultimate
19
choice here.
And they've already made the commitment.
So
Now, I don't want to hammer this any further
20 21
if you choose not to because you said it's off the
22
table.
I respect that.
But I'm just commenting
A Matter of Record (301) 890-4188
352
1
that I think it puts the panel in a difficult
2
position about a dichotomous choice that has
3
downsides on both, but we've got an outcomes trial
4
that's already enrolled.
5
comments. DR. SMITH:
6
So those are my only
I may make a quick comment on
7
that, and then we'll come back to the FDA.
8
that we are being asked some difficult questions.
9
And they're questions that, in an ideal world, we
10
might not want to be asked.
11
some reality to the situation.
I think
But I believe there's
So I think what I see from my
12 13
perspective -- and the FDA can comment on this as
14
well.
15
group of people who appreciate the difficulty of
16
being asked this question.
17
FDA is looking for is in that difficult situation,
18
not only how will you vote, which is important and
19
is going to come up in a voting question, but what
20
decision will you personally support, and why?
21
how do you expand on the dilemma?
22
communicate about that, that then will be perhaps
What I see is that the FDA has assembled a
And I think what the
A Matter of Record (301) 890-4188
Or
And what can you
353
1
helpful to the people who really have to make this
2
decision?
3
Because remember, we're an advisory
4
committee, and our decision may or may not be
5
followed.
6
best we can do at expert opinion.
7 8
We are expressing our expert, or the
Would the FDA like to add anything to that effort to clarify the situation?
9
DR. J. SMITH:
10
let me add one thing.
11
obviously, for years based on LDL cholesterol as a
12
surrogate.
13
helps you a bit.
14
hypothetically, we don't want to get bogged down in
15
all the regulatory considerations that we would
16
have to consider for accelerated approval.
17
we were to shift to that sort of a paradigm, that
18
would mean that we have downgraded LDL cholesterol
19
from a position as a surrogate suitable for full
20
approval to something less than that, yet still at
21
or above reasonably likely.
22
I think that was great, but We have approved drugs,
So let me pose it to you this way if it If we were to shift -- just
But if
So we don't really want you to have to get
A Matter of Record (301) 890-4188
354
1
into that so much.
What would be helpful for us as
2
you discuss this is if there are data that you
3
believe have accrued over the years, that have
4
knocked LDL cholesterol down from that position
5
where it has been used as a regulatory precedent
6
for drug approval, we'd like to hear you talk about
7
that.
8
If you have some uncertainty, where you are
9
now saying I really want to see the results of this
10
cardiovascular outcomes trial to confirm benefit,
11
when we've said that LDL cholesterol historically
12
has been a substitute for clinical outcomes, why
13
are you uncomfortable with that?
14
I'm asking that as a rhetorical question to
15
help you perhaps have a discussion that will be
16
helpful to us.
17
if we hear that this panel of experts is telling us
18
that things have changed and that LDL cholesterol
19
is not what it has been for regulatory use, then we
20
can consider what other options will be available
21
to us, and we can take that.
22
And if we hear that discussion, and
But we thought that the discussion today
A Matter of Record (301) 890-4188
355
1
should really focus on is the lowering of LDL
2
cholesterol sufficient to substitute for
3
demonstrating its effect on clinical outcomes,
4
which is where it has been.
5
you voted no, for example -- and we're not to a
6
voting question yet.
7
this discussion and the comments into account, and
8
I don't know what options we'll all consider.
And if not, then -- if
But we're going to take all
But that's where we want the focus of the
9 10
discussion to be.
11
to where it's been for years past, let's hear why
12
you think so. DR. SMITH:
13 14
If something is different now as
Dr. Parks, did you have a
comment you wanted to make? DR. PARKS:
15
Yes.
I just wanted to add to
16
that.
17
written in the regulations, so the accelerated
18
approval pathway is based on a surrogate endpoint
19
where there is some uncertainty as to how that
20
endpoint relates to a clinical outcome.
21 22
What Dr. Smith is referring to here is
So just an excerpt from the regulation is, "Approval under this section will be subject to the
A Matter of Record (301) 890-4188
356
1
requirement that the applicant study the drug
2
further to verify and describe its clinical
3
benefit, where there is uncertainty as to the
4
relation of the surrogate endpoint to clinical
5
benefit or of the observed clinical benefit to the
6
ultimate outcome."
7
So just again reiterating here that if you
8
now in your discussion have concerns, you want to
9
question that surrogate endpoint that the agency
10
now for, what, 20, 30 years has been approving
11
drugs based on that surrogate endpoint, you can put
12
that into your discussion.
13
scope of today's advisory committee meeting to vet
14
an accelerated approval process.
15
of today's discussion.
16
DR. SMITH:
But it's beyond the
That's not part
I'm going to take the chair's
17
prerogative, and I'm going to move to the next
18
question.
19
would like to make comments.
20
opportunity to make those comments in the context
21
of this question or the next question.
22
And I know there are some panelists who And there will be an
So I don't want to stifle comments and
A Matter of Record (301) 890-4188
357
1
input, but I'm concerned to make sure we get
2
through the questions we have in the time we have
3
available.
4
I'll start by reading the question.
5
So I'm going to move to question 2, and
The goal of LDL-lowering therapy is to
6
reduce the risk of cardiovascular disease.
7
Historically, a change in LDL cholesterol has been
8
considered sufficient to establish the
9
effectiveness of the lipid-altering drug intended
10
for use to reduce cardiovascular risk without any
11
regulatory requirement to demonstrate evidence for
12
benefit in a CV outcomes trial, provided the
13
reduction is sufficiently robust and the product or
14
its class does not have safety issues that raise
15
concern that risk exceeds benefit.
16
Discuss whether alirocumab-induced LDL-C
17
lowering is sufficient to substitute for
18
demonstrating its effect on clinical outcomes,
19
i.e., to substitute for investigation in a CV
20
outcomes trial in one or more populations; for
21
example, different degrees of CV risk, familial
22
versus non-familial etiologies of hyperlipidemia,
A Matter of Record (301) 890-4188
358
1 2
use with or without concomitant statins, et cetera. So if there's a critical question to be
3
asked for clarification, by all means, it's on the
4
table, but let's try to focus our discussion around
5
the specifics of this question.
6
launch that.
7
Dr. Geller?
8
DR. GELLER:
9
Okay?
So we'll
So I think that the patients
who discuss familial hypercholesterolemia were very
10
impressive.
11
of favoring approval of the drug for that specific
12
condition.
13
So I have considered the possibility
One of my concerns is its effect on the
14
outcomes trial, because there are people with
15
familial hypercholesterolemia in it.
16
likely to drop out to take the drug since they're
17
blinded.
18
kicking itself in the foot because they'll be the
19
highest-risk patients, and the potentially
20
beneficial effect will be lost because of drop-out.
They are
And I'm afraid that the company will be
21
DR. SMITH:
Dr. Hiatt?
22
DR. HIATT:
So I guess this is the central
A Matter of Record (301) 890-4188
359
1
question.
2
an approval surrogate with the statin drug class is
3
very, very strong and been shown in many
4
meta-analyses that it fulfills all the criteria for
5
surrogacy that you put up there, and I certainly
6
support those criteria.
7
I think the data for LDL cholesterol as
PCSK9s work through very similar mechanisms,
8
so if you were to ask what's the probability that
9
that will show a similar clinical benefit, it would
10
have to be high.
11
convinced that LDL cholesterol is an approvable
12
surrogate for non-statin drugs.
13
recent literature in the big trials, HPS2-THRIVE,
14
for example, support that contention.
15
But I have also never been
I think that the
IMPROVE-IT actually supports the concern
16
that it's not approvable.
17
side of approvability.
And so there's some
18
ambiguity in the data.
But if we have to come to
19
where we land on this, I land on the position of
20
uncertainty.
21 22
IMPROVE-IT is on the
Therefore, I don't believe we have enough data today to say that LDL with a different drug
A Matter of Record (301) 890-4188
360
1
class, a new molecular entity, is sufficient for
2
approval today.
3
DR. SMITH:
Dr. Blaha?
4
DR. BLAHA:
I'll check my mic again.
I'm
5
Mike Blaha, Hopkins.
So this is a great question.
6
And I think you might say, after the IMPROVE-IT
7
study, that the data for LDL is stronger than ever
8
in some ways.
9
after IMPROVE-IT that LDL is validated in a way
Certainly, a lot has been written
10
that we didn't have with the addition of non-statin
11
drugs to statins before IMPROVE-IT.
12
sense, I think that IMPROVE-IT reassured me quite a
13
bit about LDL.
So in that
14
I think the question comes up about in the
15
modern understanding of LDL-C versus LDL particles
16
or apo B, et cetera.
17
matters here.
18
much about LDL-C as we do about atherogenic
19
lipoprotein load, to me, mechanism matters, and the
20
genetic data that we've seen presented definitely
21
matters to me.
22
And certainly mechanism
So while a lot of us don't think as
The mechanism here seems to be supportive by
A Matter of Record (301) 890-4188
361
1
the daily randomization data.
2
presented data that not only does LDL-C go down,
3
but also apo B goes down, not HDL goes down,
4
et cetera, that gives me additional satisfaction
5
that LDL-C is tracking with what we understand as
6
LDL's atherogenic lipoproteins.
7
I'm reassured by LDL in the post-IMPROVE-IT era. DR. SMITH:
8
But since we've been
So in that sense,
So I'll make a comment.
In
9
terms of my way of thinking about this question,
10
I've been a little bit of a literalist here as I
11
look at the question, which I'm looking at right
12
now.
13
is sufficient to substitute for demonstrating its
14
effect on clinical outcomes.
15
And it's really the question of whether LDL-C
So given all the discussion we've had, at
16
least me personally, I kind of feel the need to
17
consider this question as a separate question from
18
the question of how I might vote for approval or
19
non-approval.
20
it's not sufficient.
21 22
And with that caveat, I feel that
I respect the extensive data on statins, and LDL cholesterol, and the power that that has, and
A Matter of Record (301) 890-4188
362
1
the link to actual cardiovascular outcomes.
2
also respect the possibility for very unexpected
3
results when one actually looks at clinical
4
outcomes with a drug.
5
But I
So in answering this question with my
6
blinders on here -- I'm not trying to be a
7
regulator and make decisions, but trying to deal
8
with the data and answer the question -- I would
9
feel personally that LDL-C is not adequate in terms
10
of outcomes, and that knowing the LDL-C for any new
11
class of drugs, including this one, I don't think
12
adequately informs me or completely assures me that
13
I can predict the clinical outcomes.
14
feel that I would need to see those clinical
15
outcomes data.
And I would
16
But again, I'm answering that question in
17
this literalist way that I'm viewing the question
18
here, and I understand it creates some questions
19
down the line about how one would respond.
20
DR. BLAHA:
I just want to clarify one thing
21
I said about that mechanism matters.
22
relevant that -- for example, CETP inhibitors, that
A Matter of Record (301) 890-4188
I think it is
363
1
the Mendelian randomization data doesn't support
2
CETP being a predictor of outcomes.
3
So this is a complicated question just to
4
say LDL-C, because it does matter.
I want to
5
clarify my statements in the record saying I do
6
think it matters how you lower LDL-C.
7
case of alirocumab, the way LDL-C is lowering, in
8
the sense of lowering atherogenic lipoproteins, I'm
9
reassured.
But in the
10
DR. SMITH:
Dr. Sager?
11
DR. SAGER:
I think, like a number of the
12
panel members, I also would have to say that there
13
definitely is uncertainty here.
14
mechanism of action and the genetics are favorable
15
and push me towards thinking that the likelihood
16
that these marked reductions in LDL-C in patients
17
who were on maximal statin therapy and still have
18
LDL-Cs that are not at goal would be more likely to
19
be positive in terms of reducing cardiovascular
20
outcomes than not.
21
be a substitute for an outcomes study.
22
there's just no way to know.
I think that the
But there's no way that it can
A Matter of Record (301) 890-4188
And I think
364
1
How many times have we been fooled by things
2
that we've even sometimes thought may not be
3
ethical to do a study, and then it has results that
4
are the opposite of what we expected?
5
We'll get into this later, but I think there
6
are ways to potentially handle this conundrum, for
7
example making sure, if the drug was approved, that
8
patients on statins don't have their statin doses
9
reduced or stopped.
10
So again, I would say that there's
11
uncertainty.
I personally fall on the side of
12
having optimism, but really needing to see the
13
cardiovascular outcomes study data.
14
DR. SMITH:
Dr. Thomas?
15
DR. THOMAS:
I think the guidance has been
16
LDL cholesterol, and in some ways, it's a little
17
unfair to change that for approval because that's
18
what's been already decided upon.
19
that have been improved really didn't lower LDL
20
cholesterol that much.
21
an agent that lowers LDL cholesterol well and has
22
proven cardiovascular benefits.
Before, agents
But with statins, we have
A Matter of Record (301) 890-4188
365
1
I think the uncertainty is what's the
2
additional benefit of this agent, or what will
3
happen if someone substitutes this agent for a
4
statin, or uses it differently in a way that we
5
wouldn't probably like in terms of guidance.
6
Things happen in practice that we can't control.
7
We know that very well.
8 9
So I think that's the level of discomfort. If you think about it in the heart failure field,
10
we used to use the Johnson Lasix, but then there
11
was ACE inhibitors.
12
you have to be on a background of an ACE inhibitor,
13
or spironolactone, or whatever drug it is.
14
the marginal benefit that becomes smaller and
15
smaller as you add one or more agents versus the
16
first agent that had the indication.
17
Now to get something approved,
It's
So if this was an agent that came out before
18
statins that actually lowered cholesterol like
19
this, we would approve it because the cholesterol-
20
lowering effect is dramatic, and the safety signal
21
is not so strong.
22
statins.
We already have an agent called
A Matter of Record (301) 890-4188
366
To put it in perspective, if we waited for a
1 2
cardiovascular outcomes trial for statins, you
3
would have waited approximately seven years.
4
Right?
5
time.
Before it has to come out for the approval
If I remember, Zetia was 2002, so I chaired
6 7
the SHARP trial with you.
And that was, I think,
8
2010 or so.
9
was renal failure with a combination of Zetia and
I may have the date wrong, but that
10
simvastatin.
And IMPROVE-IT is now.
11
years later.
If you wait for the final outcome
12
trial, when does that happen and how does that
13
affect prescribing practice?
14
care?
15
So that's 13
How does that affect
Now, the good thing is, with cardiovascular
16
outcomes, you can get those fairly quickly as
17
opposed to some outcomes, which, if you're going to
18
look at development of end-stage renal disease and
19
diabetics who are newly diagnosed, we'll be retired
20
before that trial would ever be completed because
21
of the long duration.
22
So I mean, there's a benefit of having a
A Matter of Record (301) 890-4188
367
1
disease where you have lots of events and they
2
happen frequent enough to get the right population.
3
So I think LDL cholesterol is what we've used.
4
think for the purposes of this discussion, that's
5
still what we should be using unless there are some
6
concerns about outcomes.
7
I
But now, let's talk about specific groups
8
because that's the second part of the question.
9
for the FH, I think it's pretty clear we use a lot
10
of therapies that we're not going to be looking at
11
the outcomes in a formalized way, so we do
12
plasmapheresis.
13
few years ago.
14
for those because we're assuming that LDL
15
cholesterol lowering helps these patients, and the
16
benefit outweighs the risk of those agents.
17
So
Other drugs have been approved a And we don't have outcomes trials
For people who are truly statin intolerant,
18
where you can't get them on a statin or you can't
19
get them close to goal, this is a problem we face
20
all the time, and we don't know what to do.
21
agents that are approved for lowering LDL
22
cholesterol that don't lower LDL cholesterol as
A Matter of Record (301) 890-4188
We use
368
1
much and have a fairly high side effect profile. One of the patients described taking niacin.
2 3
It's not the most pleasant drug to take, even
4
before, even when it wasn't the extended-release
5
form.
6
So I think there are issues with certain
7
groups that may affect the decision when people
8
vote, how that is impacted.
9
cholesterol has been a surrogate marker.
But overall, the LDL I think
10
it's been a reasonably good surrogate marker.
11
if we didn't have statins at all, I think that
12
would make the decision easier.
13
this on to some things that we're fortunate have a
14
very effective drug already available for use.
15
DR. SMITH:
16
CAPT BUDNITZ:
And
But you're adding
Dr. Budnitz? Dan Budnitz.
So I'll just
17
briefly repeat a few things just because it sounds
18
like FDA does want comment by folks on the panel.
19
And I think I wrote almost exactly the same thing
20
as Drs. Sager and Hiatt wrote.
21 22
As of today, I think there's uncertainty, same word, in the use of LDL as a surrogate
A Matter of Record (301) 890-4188
369
1
endpoint for a new class of cholesterol-lowering
2
drugs.
3
obvious, is that the sponsor may be a little bit of
4
a victim of its own success in getting a
5
first-in-class product ready for consideration for
6
approval.
7
convincing outcomes trial that is conducted, then
8
LDL might again be a good surrogate endpoint for
9
this class of medications and that subsequent
I think I'll just add what might be the
And moving forward, if there's a very
10
products might not need the same outcomes, and it
11
could go with a surrogate.
12
So I think that leaves us, then, with the
13
decision today.
And I think Dr. Thomas started
14
some of these, where we were going to start talking
15
about maybe specific patient classes in the next
16
question.
17
DR. SMITH:
Dr. Smith?
18
DR. J. SMITH:
If I just may make a couple
19
comments to make sure we're getting as much
20
information on you as we can in the time that we
21
have remaining.
22
So a few of you have commented on
A Matter of Record (301) 890-4188
370
1
uncertainty of LDL cholesterol as a surrogate or
2
you've even said no.
3
us to hear why you think so.
4
considering?
5
Dr. Hiatt mentioned a couple earlier, HPS2- THRIVE.
6
But it would be helpful for us to know -- it's very
7
easy to say you're uncertain, and we recognize
8
that, but what's making you uncertain?
9
It would be very helpful for What data are you
What trials are you thinking about?
Then the second issue -- Dr. Thomas, thanks
10
for bringing us back to a little bit -- is I want
11
to make sure that there's sufficient discussion
12
about the populations and whether or not the
13
population influences your thinking about LDL
14
cholesterol as a surrogate.
15
As I mentioned in my introductory remarks,
16
when we approved Juxtapid and Kynamro for
17
homozygous FH, a disease that we know the phenotype
18
results from deranged LDL metabolism, LDL
19
cholesterol seemed to be very reasonable to
20
consider that as a surrogate.
21 22
On the flip side, one might say -- and I'm not telling you how to think, but I'm just giving
A Matter of Record (301) 890-4188
371
1
examples of non-familial causes where there could
2
be multifactorial reasons for cardiovascular
3
events.
4
giving you sort of two ends of a spectrum.
5
One might think differently.
So I'm just
We're not talking about homozygous FH today,
6
but we do have heterozygous FH.
As you know, it's
7
been widely studied in this program.
8
wondering if there's any interaction between the
9
population and you're thinking about LDL
So I'm just
10
cholesterol as a surrogate.
11
sure that we're fully getting at those issues.
12
DR. SMITH:
So I just want to make
Thank you.
And what I'd like to
13
ask the panelists to do, I'm concerned again about
14
just time here, and I don't want to leave this
15
question until we've adequately dealt with it.
16
if you would focus on new points.
17
know, actually, if people support what's already
18
been said, but I think for time reasons, if we
19
would focus on things and points that haven't been
20
made, or remove a point that's been made, argue
21
with it if that's what you wish to do.
22
Dr. Shamburek, you're next.
A Matter of Record (301) 890-4188
But
It's helpful to
372
1
DR. SHAMBUREK:
I think 10 years ago, as we
2
heard, the LDL wouldn't have been a bit
3
controversial.
4
my view over the years, that I think LDL is a
5
biomarker.
6
think it still stands as a surrogate marker.
7
think we've learned a lot about the CETP, the
8
estrogens, fenofibrate.
9
multifactorial.
And I think I've kind of switched
However, there are situations where I I
In my mind, those are
And that's where it's more of a
10
biomarker, because many other things are going on.
11
It's not primarily the LDL effect.
12
I think Dr. Blaha made a very important
13
point, the mechanism is the key.
And I think it
14
still is the key in this disorder.
15
what strengthens that also is -- I'm a big
16
believer, and Mother Nature tells you something
17
when it makes PCSK9 mutations, when it makes FH
18
mutations.
19
fenofibrates, the CETP as strong indicators.
And I think
We don't see the estrogens, the
20
So to me, we have to make a hard decision.
21
But we may make it to where there's risk in safety
22
for a particular group, not all out.
A Matter of Record (301) 890-4188
But I still
373
1
believe, with this particular mechanism, I would
2
still favor it as a surrogate marker with some
3
limitations to availability, based on natural
4
history.
5
where we felt we've been burned over the last
6
10 years.
And I think it differs from all those
7
DR. SMITH:
8
DR. WILSON:
9
Dr. Wilson? I would bring up the point that
the dose-response curve for LDL cholesterol levels
10
and atherosclerosis is S-shaped.
11
talking about the more flatter part of the curve.
12
This is seen in observational science.
13
to some element in the clinical trials.
14
showed a straight line.
15
And we're now
It's seen Dr. Cannon
Many of us would say that probably should be
16
logarithmic, and it should be S-shaped.
17
want really in the future is more information at
18
this flatter part of the curve.
19
Dr. Geller has been asking for, more person-years
20
on the product, for instance, to really see safety
21
and efficacy.
22
So what we
And that's what
On the other hand, the patients with
A Matter of Record (301) 890-4188
374
1
hypercholesterolemia like heterozygous FH, they're
2
more on the stronger gradient.
3
different question.
4
that there really is a place for the product there,
5
but less confident where there's much less
6
information in the flatter part of the curve.
7
DR. SMITH:
8
DR. BURMAN:
9
So it's a slightly
And we feel more comfortable
Dr. Burman? Thank you.
Just a comment that
I don't think has been brought up yet.
The FDA
10
guidelines indicate that for a surrogate to be
11
authentic, it only has to reach the bar of being
12
reasonably likely that a surrogate can predict
13
long-term specific CV events, and it doesn't have
14
to be highly likely or certain.
15 16
DR. J. SMITH:
Just to clarify, that's in
the setting of accelerated approval.
17
DR. SMITH:
Dr. Everett?
18
DR. EVERETT:
Thanks.
I just was a little
19
bit confused both by the comments that Dr. Smith
20
and Dr. Parks made earlier -- not your most recent
21
comments -- and those of this Dr. Smith, because
22
the sentence is, discuss whether alirocumab-induced
A Matter of Record (301) 890-4188
375
1
LDL lowering is sufficient to substitute for
2
demonstrating its effect on clinical outcome.
3
other words, is it a substitute for investigation
4
in a CV outcomes trial?
5
In
The answer I heard from some people is no,
6
but that's not the end of the sentence, because the
7
sentence continues, as Dr. Smith mentioned just a
8
moment ago -- and this makes, I think, the next
9
question very challenging.
Are there populations
10
where it does constitute a substitute for
11
investigation in the CV outcomes trial?
12
If you stop the sentence, the answer might
13
be no, but when you start to consider some of these
14
other populations, at least for me, I think that
15
the heterozygous FH population is one where I would
16
be comfortable with LDL being a surrogate in place
17
of a CV outcomes trial.
18
In a general population, the population with
19
mixed dyslipidemia, a population at high
20
cardiovascular risk who does not have heterozygous
21
FH, or average cardiovascular risk, I would not be
22
as comfortable using LDL as a -- I do not think LDL
A Matter of Record (301) 890-4188
376
1
is a substitute for investigation in a CV outcomes
2
trial in a first-in-class medication in spite of
3
the comments by Dr. Blaha and others about the
4
mechanism of this particular drug class. DR. SMITH:
5
Yes.
And I would just push back
6
a little bit in that, when I expressed my points
7
about whether it is adequate in my opinion as a
8
surrogate, even in a population with FH, I'm giving
9
in a sense a scientific answer and not a practical
10
answer.
11
didn't mean that I wouldn't be comfortable using
12
this drug with FH.
13
feel that the question has been answered.
14
And I tried to make that point, that that
But it did mean that I don't
What do I think the probability is?
15
Extremely high that it would have benefit in that
16
population.
17
be surprised by the outcome?
18
that's considering all the arguments about
19
closeness of mechanism, which I think they're
20
compelling, and I agree with those arguments.
21 22
Do I know that?
No, I don't.
Might I
Yes, I might.
And
So maybe I was being a little rigorous there, but that was sort of where that was.
A Matter of Record (301) 890-4188
So I'm
377
1
actually not really disagreeing with you at all,
2
but I'm sort of putting in context the nature of
3
the answer I gave to that question because it
4
leaves open the question of ultimate outcomes.
5
don't know.
6
Dr. Geller?
7
DR. GELLER:
We
I have many things going
8
through my head.
9
talk about surrogate, a lot of times, you just talk
10
about marker, LDL lowering, and outcome, CVD, MACE,
11
whatever you want to define it as.
12
third aspect, and it's drug.
13
My first comment is, when you
But there's a
So a surrogate needs to have a drug as
14
part -- to have a good surrogate, you need to be
15
talking about a specific drug.
16
lowering, we know statins -- statins induce LDL
17
lowering, and that induces a positive outcome.
18
know that.
19
of drugs.
20
that LDL lowering by this drug is a surrogate.
21
That's one thing.
22
And for LDL
We
But we don't know this for a new class So that's my first point.
We don't know
I think if we would vote for approval for
A Matter of Record (301) 890-4188
378
1
FH, it may have an effect on the outcomes trials as
2
I said earlier.
3
that the outcomes trial is positive as I am that it
4
doesn't bring up something perhaps that we've seen
5
possible signs of or something bad.
6
really my concern about the outcomes trial.
7
But I actually am not so concerned
So that's
I guess I'm also concerned about the people
8
who haven't been looked at, minorities in
9
particular and that the effect on women is smaller.
10
But I am generally optimistic, and I guess I
11
wouldn't mind if we voted approval for familial
12
hypercholesterolemia.
13
But I really do want the outcomes trial to
14
be completed.
I want to know if there really is a
15
cardiovascular benefit.
16
wonderful, and I am cautiously optimistic.
I think that would be
17
DR. SMITH:
Dr. Sager?
18
DR. SAGER:
Yes, I think, just to add to my
19
previous comments, I think, here, we're thinking
20
about what's the potential risk now of approving a
21
drug that may not actually reduce cardiovascular
22
outcomes versus the risk of delaying it, and then
A Matter of Record (301) 890-4188
379
1
patients who could have really benefitted not
2
getting it.
3
I guess one thing that I factor in here is
4
when I looked at all the data, the risk that the
5
drug is actually going to be significantly
6
deleterious strikes me as pretty low, and
7
particularly quite low from a cardiovascular
8
standpoint.
9
cardiovascular signal at all.
10
I just didn't see any kind of
I know we've talked a little bit about maybe
11
something, a weak signal in diabetes or cognition.
12
In terms of having something that's really from a
13
public health standpoint, it would be
14
deleterious -- while we wait for -- deleterious if
15
the drug was approved now, and then we have the
16
outcomes study in a few years, it seems pretty
17
unlikely to me.
18
DR. SMITH:
Dr. Hiatt?
19
DR. HIATT:
So very focused on the question
20
you raised, what makes me uncertain about LDL's
21
approval surrogate?
22
inhibitors and results with niacin clearly
The results of the CETP
A Matter of Record (301) 890-4188
380
1
challenge that hypothesis.
2
currently, I think, are consistent with the
3
hypothesis, but the right phenotype has not been
4
clearly studied.
5
and high triglycerides were studied, that might
6
support the hypothesis.
7
The results of fibrates
And if patients with very low HDL
But currently, it's not really supportive
8
without further outcome data.
Omega 3 fatty acids,
9
not fully studied, bile resin binding agents,
10
historical data, not in a contemporary setting.
11
All those different drug classes in the associated
12
trials challenge the LDL hypothesis.
13
it is ezetimibe.
14
What supports
The second half of the question are
15
populations.
So in sitting through the discussion
16
on lomitapid and mipomersen, that was interesting
17
because the population was estimated to be 300
18
patients, and it's the most severe manifestation of
19
the genetic abnormality.
20
uncomfortable at the time we recommended approval,
21
but it seemed that there was no other options for
22
those patients but to move those drugs forward.
And I was slightly
A Matter of Record (301) 890-4188
381
1
If you then move down that pyramid to a
2
broader slice, which is the heterozygous patients,
3
you could make the case that the unmet need over
4
the next two and a half years is great enough that
5
you may want to make that drawing available because
6
it could be a harder population to study and at
7
much more greater risk of an event.
8 9
But then if you go to the bottom of that pie and take kind of the average patient with poorly
10
controlled LDL cholesterol, that doesn't seem like
11
a good option to me at all with this surrogate as
12
the basis of approval, nor does the
13
statin-intolerant population seem good because it's
14
such a hard thing to define.
15
So I think if you feel that there may be a
16
selective way to approach this, homozygous is off
17
the table because it wasn't studied and there are
18
alternatives for that.
19
make a case for.
20
patients where you have the exposure, the health
21
risk, the risk/benefit becomes uncertain until the
22
outcomes trial is done.
Heterozygous, you might
And you get broader, you get more
A Matter of Record (301) 890-4188
382
1
DR. SMITH:
Dr. Blaha?
2
DR. BLAHA:
I appreciated the discussion
3
between Dr. Smith and Dr. Everett about the strict
4
scientific interpretation of the phrase, can LDL-C
5
lowering be a substitute for event lowering?
6
recognize certainly that that shouldn't matter on
7
the population.
8
that's a yes or a no.
9
there's more subtlety we have to consider, and
10 11
And I
That's a scientific question But I really do think
that's the patient population like we talked about. I do think it would be tragic in a way to
12
not have these drugs available to people with
13
heterozygous FH who clearly have an unmet need.
14
can't separate from that patient population.
15
I'm very concerned about extending this to people
16
with so-called statin intolerance or mixed
17
dyslipidemia and people with diabetes, which is
18
vague also and not as much of an unmet need.
19
I
And
But to me, it does matter, the population.
20
I think LDL -- what I said before.
21
lowering is a suitable substitute for someone with
22
a pure cholesterol disorder, genetic disorder like
A Matter of Record (301) 890-4188
I think LDL-C
383
1
heterozygous FH.
2
DR. SMITH:
3
who want to make comments.
4
couple more quick comments, then I'm going to
5
summarize.
6
summary, that's how we'll handle another couple
7
comments.
8 9 10
There are a couple more people I guess I'll do a
And if anybody wants to add to my
So Dr. Orza, did you have a comment you wanted to make? DR. ORZA:
My thinking is along two lines,
11
which are not dissimilar to what other people have
12
been saying.
13
a meta-issue.
14
discussions that I've been a part of where the
15
drugs or the drug classes have been around for a
16
long time, and the regulatory standard has kind of
17
conglomerated over time.
18
where we wish we could reconsider some of the
19
assumptions, or some of the surrogate endpoints, or
20
some of the other things that made that up.
21 22
One is kind of what I see as kind of There are a lot of these drug
And we get to a point
So I think the opportunity that FDA is taking here to really kind of rethink and revisit
A Matter of Record (301) 890-4188
384
1
this, now that we have so much more science behind
2
it, is really a good healthy thing. I think, perhaps unfairly, of surrogates as
3 4
something that you settle for because you don't
5
have something better.
6
instance, on the one hand, we do have something
7
better.
8
cholesterol lowering because we do know that we can
9
look at actual cardiovascular endpoints.
And I think in this
We don't have to settle for simply LDL
On the other hand, it hasn't worked out all
10 11
the time that a drug that lowers LDL cholesterol
12
does result in improved cardiovascular outcomes
13
without worsening other outcomes.
14
kind of call it into question, just in the
15
abstract, as to whether or not it's still a good
16
surrogate.
So that does
With respect to this drug and this
17 18
population in specific, I wonder if what we're
19
saying is not so much that it still might be a good
20
surrogate for the FH population, as maybe it's not
21
a surrogate.
22
itself.
Maybe it's an endpoint in and of
Maybe LDL cholesterol lowering in and of
A Matter of Record (301) 890-4188
385
1
itself in this population is a good thing, and
2
we're not really talking about it being a surrogate
3
anymore.
4
Then I just had questions for FDA in terms
5
of your options.
6
on the surrogate endpoint, and then the results of
7
the outcomes study were negative, what would your
8
options be then to pull it back for any population?
9
If you were to approve it based
DR. J. SMITH:
So if it were approved under
10
our traditional approval pathway, where LDL is
11
accepted as the full basis of approval, if the
12
cardiovascular outcomes trial were not to show the
13
expected reduction in cardiovascular risk, we would
14
have a conversation with the sponsor.
15
depend obviously on the review of that trial, the
16
safety of that trial.
It would
17
But if the sponsor agreed that the trial
18
changed the way that labeling should appear, the
19
indications should appear, It might be very easy to
20
change things.
21
problems with the trial that made it uncertain as
22
to how well the trial actually answered the
If they disagreed or there were
A Matter of Record (301) 890-4188
386
1
question, without getting into all the details, it
2
might not be so straightforward to do something
3
with the approval.
4
So there are options, but it's hard to dwell
5
on the hypotheticals.
6
straightforward, oh, well, clearly the drug would
7
just be withdrawn, or the indication would just be
8
withdrawn.
9
It wouldn't be a
It's not so straightforward.
DR. ORZA:
Then are there other mechanisms?
10
Is there such a thing as a conditional approval
11
where it's pending the outcome?
12
something on the sponsor's side where, if it were
13
not to be approved, they could make it available
14
through open access or something to the people who
15
desperately need it while we're waiting for the
16
trial results?
Or is there
17
DR. J. SMITH:
So those are a couple
18
completely different issues.
19
approval that you're referring to is the same
20
accelerated approval that we've been discussing.
21
Regarding open access, if a drug is not approved,
22
there are mechanisms for a drug to be made
The conditional
A Matter of Record (301) 890-4188
387
1
available to patients of unmet need.
2
conversations that the sponsor would have with us.
3 4 5
DR. SMITH:
And that's
Dr. Budnitz, final, quick
comment? CAPT BUDNITZ:
Just to pick up on what
6
Ms. Orza had mentioned And other folks have said,
7
for LDL-C lowering, it might be a surrogate
8
endpoint for more populations, the familial
9
population -- maybe the thing that we're trying to
10
get at is that it could be an endpoint in and of
11
itself by analogy.
12
If we think about like PKU, there was a drug
13
out there that would help those patients clear
14
phenylalanine, they wouldn't wait for the outcome
15
of brain damage.
16
outcome.
17
think is going on in the familial cases.
18
We'd just use that as the
And maybe that's what we intuitively
DR. SMITH:
I will try to summarize.
We
19
heard views expressed -- and this is, again,
20
focusing on the questions and the elements that
21
have been asked here by the FDA, that we heard
22
views expressed that, ultimately, rigorously, the
A Matter of Record (301) 890-4188
388
1
LDL cholesterol lowering itself is not equivalent
2
to clinical outcomes, and, thus, as a surrogate,
3
one might raise questions about that. We have also heard a lot of modification of
4 5
that circumstance in that we've heard expressed
6
that there are very strong data for LDL cholesterol
7
being a very good surrogate in the context of
8
statin use, and on the one hand, the recognition
9
that that doesn't validate the surrogate for other
10
drugs.
11
which were mentioned, where the data we have
12
available have not been supportive of LDL as a
13
surrogate.
14
And there are examples of drugs, separate,
On the other hand, the point has been made
15
that this particular drug, in its mechanism of
16
action, is operating at a locus that is really very
17
close to the locus in action of statins, and that
18
that tends to shift one's thinking in seeing this
19
as more likely being effective as a surrogate.
20
Other points that have been discussed at
21
length are the importance of recognizing that for
22
different patient groups, a willingness to use a
A Matter of Record (301) 890-4188
389
1
drug or approve a drug like this in the setting of
2
perhaps some residual questions about the
3
effectiveness of LDL as a surrogate, that it may be
4
different for different patient groups.
5
been stressed with familial hypercholesterolemia as
6
a group, where one would be looking aggressively
7
for new drug therapies and/or this particular drug.
8
Then alternatively, the point was made that
9
there are other patient groups, which might include
And that's
10
the broad spectrum.
This might or might
11
not -- this was expressed by a member of the panel,
12
might anchor the broad spectrum of high
13
cardiovascular risk patients, might include
14
patients who are "classified" as statin intolerant.
15
A note was made that in terms of the
16
dose-response effect of LDL lowering against
17
outcomes, that that's something that we have some
18
uncertainties about, and that that dose-response
19
curve may be an S-shaped curve, meaning that when
20
levels get pretty low, pushing LDL levels lower, it
21
may not have the same effects on outcomes as when
22
starting with higher levels in lowering dose
A Matter of Record (301) 890-4188
390
1
levels.
2
sense, understanding the data and the uncertainties
3
or questions that might remain in the data.
4
Again, in attempting to answer this
5
question, the point was again raised or made about
6
the level of concern about the evidence for serious
7
adverse effects of the drug, and the fact that,
8
again, how one's willingness to accept a
9
surrogate -- or a marker as a surrogate is
10 11 12
And again, that's just a point of, in a
influenced by that context as well. So I think I'll stop there if anybody wants to add to that or modify those comments.
13
(No response.)
14
DR. SMITH:
So I'd like to move to the
15
voting question.
And again, I'll read it.
Has the
16
applicant sufficiently established that the
17
LDL-C-lowering benefit of alirocumab exceeds its
18
risks to support approval in one or more patient
19
populations?
20
regulatory pathway, it would not be required to
21
successfully demonstrate an effect of alirocumab on
22
CV outcomes after an approval based on changes in
We remind you that, under the current
A Matter of Record (301) 890-4188
391
1 2
LDL-C. So there are two potential votes here.
A
3
would be a yes in response to that question above.
4
And if yes, please explain your rationale and
5
describe the patient population or populations for
6
whom you believe that the benefit/risk is
7
favorable.
8
studies you believe the applicant must conduct to
9
establish a favorable benefit/risk to support
10 11
If no, please describe what further
approval. So first of all, any discussion or questions
12
about the wording of this question?
13
that why your hand is up?
14
DR. GELLER:
Dr. Geller, is
I was going to suggest to
15
divide the question into the three risk groups so
16
you can do a yes/no for the three, and then worry
17
about the second part.
18
DR. SMITH:
Based on my prior experience
19
working with the FDA, I think that, very often, as
20
we look at questions like this, individuals of us
21
would like, for good reasons, to ask a question a
22
different way, but it creates a lot of
A Matter of Record (301) 890-4188
392
1
complications for trying to process this meeting.
2
So there is an opportunity, and it's very
3
important, in comments that people make after their
4
vote to perhaps express amplifying opinions about
5
their vote, whether it was a yes or a no vote.
6
I think that's the avenue we need to take for this. DR. GELLER:
7
Can you help me, then?
And
If
8
somebody believes that we should recommend approval
9
for one of the subgroups and not the others, do you
10
vote yes or no? DR. J. SMITH:
11
You would vote yes.
We're
12
asking about approval in one or more populations,
13
so in your comments, please address -- if you've
14
identified three risk groups, you could comment on
15
all three.
16
you could comment on 10.
17
caveats and suggestions as you'd like after your
18
vote.
19
approval is supported for at least one.
20
If you've identified 10 risk groups, You could address as many
But really, the question is whether or not
DR. SMITH:
So if there are no further
21
discussions on this question, no further
22
discussion, we'll begin the voting process.
A Matter of Record (301) 890-4188
And
393
1
the way to do this is to press the button on your
2
microphone that corresponds to your vote.
3
should be a yes or a no.
4
would encourage you to not abstain.
5
uncertainty, you have the opportunity to vote yes
6
or no and then comment on how hard it was --
7
(Laughter.)
8
DR. SMITH:
9 10
There
There's an abstain, but I If you have
-- and why you chose one or the
other, and how you want to amplify or qualify that vote. Please press the button firmly.
11
After
12
you've made your selection, the light may continue
13
to flash.
14
want to change it or what you did, just press the
15
button again before the vote is closed.
16
everyone has voted, we'll have a projection of what
17
the votes were, and they'll be read into the
18
record.
19
each person to briefly comment on their vote and
20
why.
If you're unsure of your vote, or you
And after
And then there will be an opportunity for
So go ahead with your votes, please.
21
(Vote taken.)
22
DR. BAUTISTA:
The vote is now closed.
A Matter of Record (301) 890-4188
I'll
394
1
now read the vote into the record:
2
3 noes, zero abstentions.
3
DR. SMITH:
13 yeses,
What we'll do now is we'll go
4
around the room, and each person who voted, we'll
5
ask you to state your name into the microphone for
6
the record, state your vote into the microphone.
7
And then you have the opportunity to just briefly
8
share what you might wish to, to amplify your
9
voting decision.
10
And maybe we'll start with
Dr. Nason. DR. NASON:
11
I was afraid you'd say that.
12
name is Martha Nason.
I voted no.
13
mentioned, I did not find this an easy decision,
14
especially for the population of familial
15
hypercholesterolemia -- FH.
16
that.
My
As was
Let's just go with
17
I'm very on the fence for that population.
18
For the other populations, I think it's more clear
19
that the safety data I think is just not there,
20
necessarily.
21
is the right surrogate in a new class of drugs, as
22
has been discussed.
I'm not fully convinced that this LDL
A Matter of Record (301) 890-4188
395
I really don't want to see the
1 2
cardiovascular outcomes trial derailed, which I
3
think it would be if the drug is licensed possibly
4
for anyone, but certainly for a wide swatch of
5
people.
6
important to get that data.
I think that's going to be really
7
I think no matter what the sponsor did to
8
encourage people, I think even a blinded trial, I
9
think people are likely to know and doctors are
10
likely to have a pretty good guess, given that
11
we've seen this result on the LDL, as to which arm
12
people are on. So I think you would systematically lose a
13 14
lot of placebo people on the outcomes trial if it
15
were licensed.
16
impossible to adjust for.
17
go.
18
And I think that would be pretty So I'd hate to see that
I think that data is necessary. Again, the FH population, I feel like I
19
could have pushed either button.
I'm still really
20
on the fence about that one.
21
would say for anybody with FH, or for anyone with
22
FH whose LDL is still above a certain limit.
I don't know if I
A Matter of Record (301) 890-4188
I
396
1
think that could even be fine-tuned more as far as
2
just the idea of lowering LDL, if it were to turn
3
out not to show an effect in a CV trial, would
4
still be a benefit. DR. SAGER:
5 6
9
Philip Sager.
I think this drug
addresses --
7 8
I guess that's all.
DR. SMITH:
If you would, state your vote,
DR. SAGER:
Philip Sager.
please. I voted yes.
I
10
think this drug potentially addresses a really
11
significant unmet medical need in patients who are
12
at high cardiovascular risk and their
13
hypercholesterolemia is not adequately controlled.
14
In terms of waiting for the outcomes study,
15
I think the likelihood that having people on the
16
drug now would subject them to high risk is quite
17
unlikely based upon the data sets that we examined.
18
The patient groups that I would look for
19
this in are the heterozygous FH patients, patients
20
at high cardiovascular risk, and with both of those
21
populations on statins whose LDL-Cs are not
22
adequately being treated.
A Matter of Record (301) 890-4188
397
I'd also consider it for those populations
1 2
who are on maximally tolerated statins, but I think
3
that has to be really dealt with very carefully in
4
the labeling.
5
patients having their statins reduced or stopped.
6
I think that's something that would really have to
7
be addressed very strongly in the label.
8
stop there.
I've got great concerns about
DR. SHAMBUREK:
9
Robert Shamburek.
And I'll
I voted
10
yes.
I think the risk and benefit is a major
11
factor in my decision.
12
particular first-in-class drug, the mechanism of
13
action, was an important driving force here.
I also think this
I also agree there's an important unmet
14 15
need.
And I think the target group in my mind
16
right now would be high LDL in a high
17
cardiovascular risk patient on a high-dose statin,
18
which in many of these times is FH with the
19
qualification that these patients could also -- and
20
I don't like the term statin intolerant.
21
the maximally tolerated statin therapy.
22
particular group, I think, should also be
A Matter of Record (301) 890-4188
I like This
398
1 2
considered. DR. EVERETT:
Brendan Everett.
I voted yes.
3
I voted yes because I think this drug could offer
4
substantial benefit to patients with heterozygous
5
familial hypercholesterolemia, and I'm worried
6
about the opportunity cost for them of waiting two
7
or so years until potentially approval of the drug
8
based on outcomes.
9
I would restrict this approval to this
10
population and would not allow broader use for
11
patients with mixed dyslipidemia, statin
12
intolerance, or very high or high CV risk, absent
13
the presence of heterozygous FH.
14
for me, absent the CV outcomes data, the safety
15
profile, while absent of any large signals of harm
16
or concern, it's not sufficiently robust to balance
17
a lack of real established benefit on the outcomes
18
side of the scale.
19
This is because,
I believe that the unmet medical need, as I
20
said, is most substantial in the patients with FH.
21
If the FDA and the sponsor can't or could not -- if
22
we knew now that they could not figure out a way to
A Matter of Record (301) 890-4188
399
1
deliver the drug just to those patients with
2
heterozygous FH, then I probably would have voted
3
no because I think the outcomes trial is of
4
fundamental importance for all the other
5
populations in question.
6
DR. GELLER:
Nancy Geller.
I voted yes
7
because when you repeated the questions, you said
8
at least one of the subgroups.
9
is an unmet need for the heterozygous FH group.
And I think there I
10
don't want to deprive them of this drug for another
11
two years.
12
want the outcomes trial results before we approve
13
it for those.
14
And the others, I think that we really
DR. WILSON:
Peter Wilson.
I voted no.
And
15
I'll say it's an interesting response because I
16
wasn't sure where the vote was going to go, but I'm
17
certain that I'm uncertain.
18
(Laughter.)
19
DR. WILSON:
So I vote conservatively.
And
20
I no longer think we're in an LDL surrogate era.
21
That's number one.
22
think we need clinical outcomes.
I'll say that very strongly.
A Matter of Record (301) 890-4188
We've got to get
I
400
1
new LDL-affecting molecules and lipid-affecting
2
molecules for cardiovascular disease
3
outcomes -- we've got to get them right, right out
4
of the box with the initial reviews, I feel.
5
The other thing I had -- so Brendan Everett,
6
I really respect his synopsis, and I came out
7
almost the same way, but I came on the other side
8
of the vote.
9
program that really targets FH patients.
And I think what's missing is a And I
10
think such a thing could be developed rather
11
rapidly to highly recruit FH patients and address
12
some of our concerns, and perhaps come back to this
13
committee, or the FDA would move with the sponsor
14
to create a framework where the product could get
15
approved quicker.
16
But we don't have such a program outlined
17
that really targets the FH patients at this point,
18
and I'm concerned they might not be that highly
19
represented in this critical large trial that's
20
under way.
21 22
DR. HIATT:
William Hiatt.
I voted no.
In
trying to add to my colleagues' comments, I think
A Matter of Record (301) 890-4188
401
1
we're going to be saying much the same thing.
But
2
as Dr. Wilson, I landed on the no side.
3
reason is, the question was framed, focusing on the
4
LDL surrogacy issue.
5
comment about the number of trials that have not
6
supported that hypothesis, that drove my no vote.
7
The other thing that drove it is that the
And the
And based on my previous
8
sponsor intentionally targeted a very broad
9
development program of a number of different
10
clinical types of patients and different
11
comparators.
12
this into a very broad population of patients.
So the intention is clearly to put
13
Had we asked this question three years ago
14
in the homozygous, as I said earlier, LDL-C seemed
15
to be sufficient.
16
broad.
17
But now, we're going quite
I think it's insufficient. But to mirror my colleagues, I would like to
18
see a path forward in some way for a very high-risk
19
population, as we said, the heterozygous patients
20
that might benefit in this interim period between
21
now and the time the cardiovascular outcomes trial
22
is completed.
A Matter of Record (301) 890-4188
402
1
My final comment is, I think the sponsor got
2
put in a little bit of a bind here because, as
3
mentioned earlier, discussed earlier of why this is
4
a difficult question, if this had been accelerated
5
approval, the question would have been an easy
6
vote, but it wasn't.
7
DR. SMITH:
I'm Robert Smith.
I voted yes,
8
and I certainly share many of the views that have
9
been expressed by the prior panel members who have
10
commented.
11
vote, I do feel that there is a substantial unmet
12
need, and I feel that's very true for familial
13
hypercholesterolemia.
14
true for less well-defined groups of patients.
15
But to emphasize a few points behind my
But I feel that it's also
Perhaps I'm being a little bit anecdotal
16
here as well, but in my own practice, I have
17
certainly had patients with galloping
18
cardiovascular disease and LDL cholesterol levels
19
that are very difficult to control, who may be
20
taking substantial levels of statins, may be
21
compromised in their statin dose because of issues
22
of tolerance, but are facing very serious disease
A Matter of Record (301) 890-4188
403
1 2
outcome issues. For those groups of patients, I don't feel
3
comfortable with extending an interval, when they
4
would not have access to a medication like this,
5
while we wait for outcomes data to come.
6
already expressed my views that I'm very much
7
unconvinced, absolutely certain, that LDL is an
8
adequate surrogate, but I'm still unwilling to
9
subject people to that wait.
10
And I've
Part of the reasons for that or the further
11
reasons are, in part, the mechanistic
12
similarities -- or closeness I guess is more how I
13
would put it -- to statins.
14
It's hypothetical.
15
benefit, which is hypothetical.
16
word "potential," and also the potential risks,
17
which based on the data that had been collected and
18
the limitations in them, nonetheless are not
19
alarming risks.
20
That encourages me.
And also, the potential And thus I use the
So that puts me in a circumstance where I
21
was pretty comfortable voting yes.
22
mean I don't totally support the importance of
A Matter of Record (301) 890-4188
That doesn't
404
1
completing the cardiovascular outcomes trial and
2
even doing others, which I certainly do.
3
on the optimistic side, separate from negotiations
4
that can occur and regulatory pressures that can be
5
brought, I can envision some motivators for the
6
sponsor.
7
come out favorably, that can have tremendous impact
8
in terms of their utilization of the drug.
9
it's also recognizing that this is an injected
Sort of
One of them is, obviously, if the trials
10
medication, and that that represents often a
11
barrier for patients.
But
12
There may be at least one, who knows.
13
may be other PL alternatives as other ways to
14
approach patients who are not adequately
15
satisfactorily controlled on a statin.
16
there are other pretty powerful motivators for the
17
sponsor to complete that trial, which I hope will
18
play out in reality.
19
DR. BLAHA:
I'm Michael Blaha.
There
So I think
I voted yes.
20
And I'm really basing my decision as my background
21
as a scientist, as an epidemiologist, but very
22
importantly as a clinician who deals with patients
A Matter of Record (301) 890-4188
405
1
like this and experiences the unmet need that we've
2
heard throughout the day from both providers and
3
patients.
4
indication of heterozygous FH, or FH in general,
5
because of the genetic nature of the disorder, the
6
isolated LDL elevation that is the primary driver
7
of risk in these patients, and the unmet need.
8
I would vote for a strong yes for the
I also voted yes for high-risk secondary
9
prevention, not primary prevention, but secondary
10
prevention in patients who have an "insufficient"
11
response to statins or its response to maximally
12
tolerated statins.
13
I like how it's worded in the new ACC AHA
14
guidelines.
In those patients, additional options
15
are needed.
There is an unmet need in these
16
patients who are very high risk.
17
Based on my decision, based on the totality
18
of the evidence, but including the IMPROVE-IT
19
study -- which I think bolstered the idea of LDL
20
lowering as a primary way to lower cardiovascular
21
events -- but also based on mechanism, as we talked
22
about, because I don't think lowering LDL via any
A Matter of Record (301) 890-4188
406
1
mechanism is equivalent.
2
genetic data, the mechanistic data, the safety
3
data, are very compelling for this drug, for
4
lowering risk in high-risk secondary prevention
5
patients who just can't get a maximal response to
6
statins.
7
However, I do find the
But I would very strongly say I would not
8
endorse approval for this drug for what would be
9
construed as statin intolerance.
I think that it
10
could be interpreted in a variety of different ways
11
that could become concerning.
12
anecdotally, in my experience, what people call
13
statin intolerance is widely different,
14
particularly amongst non-cardiologists, or
15
endocrinologists, or whoever deals with these
16
patients the most, like some of us in this room do.
17
I know just
Certainly, I get referrals for statin
18
intolerance in people who have had one dose of a
19
statin, and that's not statin intolerance.
20
much worry about this drug getting into a much
21
wider circulation for which it's not ready without
22
a large cardiovascular outcomes trial.
A Matter of Record (301) 890-4188
I very
And I also
407
1
feel the same way about the idea of mixed
2
dyslipidemia. DR. THOMAS:
3
Abraham Thomas.
I voted yes.
4
I think some of the other drugs were mentioned, but
5
the other drugs that haven't really progressed have
6
had safety signals, which this one has not had so
7
far, so I think it's really not a fair comparison.
8
So I think the LDL lowering is appropriate in this
9
case.
10
For FH, I think it's very clear if it was
11
just FH alone, we wouldn't really have the need for
12
even a meeting potentially to discuss the approval
13
because of the safety profile that's been presented
14
so far in the efficacy in terms of lowering LDL
15
cholesterol.
16
I do think there's a concern, if it does get
17
approved, that FH patients won't be enrolling in
18
the trial because they'll want to get the
19
medication now that it's approved.
20
For the other groups, there is an unmet
21
need, not reaching targets, not tolerating statins.
22
However, I think there are going to be clear
A Matter of Record (301) 890-4188
408
1
guidelines that need to be in place by the sponsor
2
of how to guide a physician on how to use these in
3
these situations.
4
to be a difficult issue on the FDA's side to do
5
that as well.
And I'm sure labeling is going
There are three things that I think are
6 7
going to be important that will limit some of the
8
concerns I have, and one is patients.
9
enough to get patients to take statins when we know
10
they have evidence and they're oral medications you
11
take once a day.
12
someone to take an injection every two weeks
13
without convincing data.
It's hard
So it would be harder to get
It's going to be hard to get prescribers to
14 15
prescribe this without convincing outcomes data.
16
And what drives our healthcare unfortunately is the
17
people who pay for it.
18
benefit managers and insurances are going to really
19
want to pay for something unless they have outcomes
20
data.
21 22
So I don't think pharmacy
So I feel strongly that the trial will get done, and there will be enough patients in it to
A Matter of Record (301) 890-4188
409
1
find out the outcomes.
2
don't think it's right to deprive the FH patients
3
that really could benefit from this drug in the
4
short term. DR. BURMAN:
5
But in the meantime, I
Ken Burman.
I voted yes.
I
6
think the sponsor has demonstrated that the
7
LDL-lowering benefit of the medication exceeds its
8
risk, and I support its approval.
9
postmarketing studies relative to safety and
I think
10
benefit need to be performed, with a focus on
11
adjudicated CV events and the potential benefits of
12
a lower LDL value. I suggest that, presently, the optimal
13 14
benefit of the medication is when given to
15
heterozygous FH patients, as noted, to
16
statin-intolerant patients who really are statin
17
intolerant, intolerant given the vagaries of that
18
definition; or to patients with high risk who have
19
not reached goal LDL or continue to have CV events,
20
or a high likelihood of having CV events on statins
21
alone.
22
It is never possible or not possible to know
A Matter of Record (301) 890-4188
410
1
all the benefits, risks, and side effects of a new
2
class of medications at approval, and postmarketing
3
studies are required with the adjudicated study, as
4
noted.
5
In summary, the studies that have been
6
performed have noted safety and efficacy, and I
7
recommend approval, but note the caveat that
8
possible new adverse or safety effects may be noted
9
in larger, more diverse populations.
10 11
This
medication seems to meet an important unmet need. MS. MCCALL:
Debra McCall.
I voted yes.
I
12
believe this is a drug that needs to be in the
13
patient's toolbox.
14
patients with FH, those who are statin resistant,
15
and that would be decided with a discussion between
16
patient and cardiologist, and also for those with
17
high cardiovascular risk that are not being well
18
treated otherwise.
19
I believe it's appropriate for
I have a small amount of unease because of
20
the incidence of atrial fibrillation the short time
21
that the study was actually conducted and the lack
22
of diversity in the overall population.
A Matter of Record (301) 890-4188
As to some
411
1
of the other members who have commented that
2
patients might be resistant to taking an injection,
3
also understand that many of us are already
4
suffering pill fatigue, and one more pill might be
5
one too many.
6
be a plus.
7
So an injectable every two weeks may
I'd also like to say to those patients that
8
e-mailed the FDA and then took the time to come in
9
today and talk to us, thank you.
10
a lot of time and took a day.
11
you're the reason we're here.
12
DR. ORZA:
I know this took
I appreciate it, and
Michelle Orza.
I voted a
13
reluctant yes.
14
wording of the vote that said one or more patient
15
populations.
16
FH population, two years is too long to wait.
17
represents a tremendous burden of cardiovascular
18
morbidity and mortality for that population.
19
I essentially took the out in the
I think that for the very high-risk That
But I'm really only comfortable with the
20
risk/benefit equation in that population because I
21
really don't think that LDL-C lowering is a
22
sufficient surrogate any more.
A Matter of Record (301) 890-4188
Writ large, I think
412
1
we're saying it's not really a surrogate in this
2
population.
3
And that's the basis on which I think it's
4
acceptable.
It's an endpoint in this population.
I think that very narrow population needs to
5 6
be dealt with through labeling.
The indication
7
needs to be much more restrictively worded than it
8
currently is.
9
signals that we're concerned about and getting more
10
information about those needs to be handled through
11
a robust REMS.
And through REMS, I think the safety
12
I think we also need assurances from the
13
sponsor, and I think that they've given them and
14
that they will do everything they can to see that
15
that outcomes trial is completed.
16
In terms of the field as a whole, I think
17
there needs to be consideration given to what the
18
non-cardiovascular endpoints should be in these
19
trials
20
have to have a little bit more humility.
21
still learning about statins, for that matter.
22
because I think we're still learning we
CAPT BUDNITZ:
Dan Budnitz.
A Matter of Record (301) 890-4188
We're
I voted yes for
413
1
a very narrow indication of FH in combination with
2
a high dose or high-potency statins.
3
I think, a little bit by some mental contortions in
4
thinking about, again, in this patient population,
5
LDL as an outcome as a surrogate endpoint.
6
I was swayed,
I'm also reassured by the safety data for
7
the time period of the study, 12 to 18 months, with
8
the hope and strong encouragement of the sponsor to
9
complete the outcomes trial for safety monitoring,
10
so then with that data in hand, we might be able to
11
go for a more broader indication, and also might be
12
able to demonstrate, as a first-in-class product,
13
that LDL might indeed be a good surrogate that
14
could be used in the future for this class.
15
DR. STANLEY:
I'm Charles Stanley.
I voted
16
yes, obviously, for many of the reasons that have
17
been talked about, but particularly because I think
18
that this discussion today is an illustration of
19
the need to individualize the way we think about
20
drugs to specific genetic mechanisms.
21 22
In a way, this discussion has almost been a review of an orphan drug indication, except that
A Matter of Record (301) 890-4188
414
1
familial hypercholesterolemia is much more common.
2
And as a pediatrician, I am also particularly
3
encouraged that this is a new therapy that's going
4
to make a big difference for the children in these
5
families with familial hypercholesterolemia.
6
So I think, for all these reasons, I'm very
7
enthusiastic about supporting its use, particularly
8
for the FH indication. DR. COOKE:
9
I'm David Cooke, and I voted
10
yes.
11
surrogate for a medication for a cardiovascular
12
improvement for all classes of medication.
13
However, I think for this medication, the data is
14
sufficient to justify its approval for at least
15
some indications.
16
of LDL lowering as well as the genetic and animal
17
studies of the mechanism of this medication.
18
I do feel that LDL is no longer an absolute
I base that both on the degree
As with others, I would certainly support
19
the indication for this medication in heterozygous
20
FH as well as those with high-risk cardiovascular
21
disease in the setting of the maximally tolerated
22
statin therapy.
A Matter of Record (301) 890-4188
415
I'm much less certain about its approval for
1 2
broader indications.
And in those, I would support
3
awaiting the cardiovascular outcome trial before
4
extending the approval for those.
5
DR. SMITH:
6
Dr. Albrecht, you didn't have the option to
7
vote, but would you like to make any comment? DR. ALBRECHT:
8 9
Thank you.
That's very kind.
Per my
earlier comments, I would have voted yes for the
10
reasons that I think many of the other voters
11
stated.
12
heterozygous FH, statin-resistant population; I
13
think also the high-risk cardiovascular population.
14
I would have certainly included
One thought that wasn't stated -- and I
15
wonder whether that's a possibility -- is the
16
recruitment for the outcomes trial I don't think,
17
is complete yet.
18
deficiencies in the data, for example the racial
19
mix and so on, can be fixed to a point by some
20
over-involvement in the trial just for some
21
populations.
22
DR. SMITH:
And I wonder whether some of the
Would the FDA like to make any
A Matter of Record (301) 890-4188
416
1 2
additional comments? DR. J. SMITH:
We would just like to thank
3
you all very, very much for a very long day.
4
know it's been challenging.
5
this is just day one of two.
6
much for your time.
9
And for many of you, So thank you very
We really appreciate it. Adjournment
7 8
We
DR. SMITH:
I finally would like to thank
everyone for their patience and energy all through
10
the day.
11
That extends to the people who spoke in the open
12
public hearing.
13
your input.
14
clear presentations and cooperation with all our
15
many questions, and the FDA for also all of your
16
communications with the panel, and all the panel
17
members for such good work.
18
adjourned.
19 20
I know we finished a few minutes late.
We very much appreciate and value
And thanks to the sponsor for very
And so this meeting is
(Whereupon, at 5:17 p.m., the meeting was adjourned.)
21 22
A Matter of Record (301) 890-4188
View more...
Comments
