1
DEPARTMENT OF HEALTH AND HUMAN
SERVICES
FOOD AND DRUG
ADMINISTRATION
CENTER FOR DRUG EVALUATION AND
RESEARCH
PEDIATRIC ADVISORY
SUBCOMMITTEE
OF THE ANTI-INFECTIVE DRUGS
ADVISORY COMMITTEE
ACS Conference Room
Room 1066
2
PARTICIPANTS
Joan P. Chesney, M.D., Chair
Thomas H. Perez, M.P.H., Executive
Secretary
CONSULTANTS (VOTING):
Mark Hudak, M.D.
David Danford, M.D.
Richard Gorman, M.D.
Robert Nelson, M.D., Ph.D.
Susan Fuchs, M.D.
Victor
Santana, M.D.
Naomi Luban, M.D.
Judith
O'Fallon, Ph.D.
Katherine L. Wisner, M.D.
MEMBER OF THE ANTI-INFECTIVE DRUGS
ADVISORY
COMMITTEE (VOTING):
Steve Ebert, Pharm.D., Consumer
Representative
FEDERAL GOVERNMENT EMPLOYEE (VOTING):
Janet Cragan, M.D.
INDUSTRY REPRESENTATIVE TO
ANTI-INFECTIVE DRUGS
ADVISORY COMMITTEE (NON-VOTING):
Sam Maldonado, M.D., industry
representative
FDA STAFF:
Solomon Iyasu, M.D.
Susan Cummins, M.D.
Shirley Murphy, M.D.
Dianne
Murphy, M.D.
3
C O N T E N T S
Call to Order and
Introductions,
Joan P. Chesney, M.D. 5
Meeting Statement, Thomas H. Perez,
M.P.H. 7
Welcome, Dianne Murphy, M.D. 10
Adverse Event Reports per Section 17 of
Best
Pharmaceutical for Children Act,
Solomon Iyasu, M.D. 15
Fexofenodine, Jane Filie, M.D. 22
Topotecan and Temozolomide, Susan
McCune, M.D. 34
Moxifloxacin and Ciprofloxacin,
Harry Gunkel, M.D. 59
Fosinopril, Larry Grylack, M.D. 66
Fentanyl, ShaAvhree Buckman, M.D. 78
David J. Lee, Ph.D. 89
D.
Elizabeth McNeil, M.D. 91
Discussion of Question 1 92
Adverse Event Reports per Section 17 of
BPCA
(cont.), Venlafaxine, Hari
Sachs, M.D. 115
Pediatric Update, Dianne
Murphy, M.D. 148
Meeting Statement, Thomas H. Perez,
M.P.H. 170
Update on Neonatal Withdrawal Syndrome:
Kathleen Phelan, R.Ph. 174
Robert Levin, M.D. 189
Katherine Wisner, M.D.,
Women's Behavioral Health CARE 216
Discussion of Questions 2 and 3 254
Update on Congenital Eye Malformations in
Infants,
Solomon
Iyasu, M.D.
303
4
C O N T E N T S
(Continued)
Open Public Hearing:
Philip Sanford Zeskind, Ph.D.,
University
of
Pediatric Research Equity Act,
Shirley Murphy, M.D. 326
Overview of
Conduct of Clinical Research Involving
Children," Robert Nelson,
M.D. 340
5
1 P R O C E E D I N G S
2 Call to Order, Introductions
3
DR. CHESNEY: Good morning. I think we
4 are
ready to get started. I would like to
welcome
5
everybody to this meeting which, for those in the
6
room who don't know, and Dr. Murphy will elaborate
7 on
this, this is the last meeting for this group of
8 the
Pediatric Subcommittee as currently
9
constituted. I would like to also
mention that Dr.
10
Mimi Glode will not be with us because her father
11
became ill on Sunday and she had to cancel at the
12
last minute.
13
Tom has just told me that traffic is going
14 to
become very bad this afternoon because of
15
President Reagan's funeral so we want to keep that
16 in
mind as we move on throughout the day.
So, I
17
think we will start with introductions and, Dr.
18
Maldonado, would you like to start?
19
DR. MALDONADO: Sam Maldonado,
from
20
Johnson & Johnson, the industry representative on
21
this committee.
22
DR. FUCHS: Susan Fuchs, pediatric
6
1
emergency medicine physician from Children's
2
Memorial Hospital in Chicago.
3
DR. O'FALLON: Judith O'Fallon,
4
statistics, retired from the Mayo Clinic.
5
DR. SANTANA: Victor Santana,
pediatric
6
hematologist/oncologist from St. Jude's Children's
7
Research Hospital in Memphis, Tennessee.
8
DR. GORMAN: Rich Gorman,
pediatric
9
private practice in Ellicott City, Maryland.
10
DR. EBERT: Steve Ebert,
pharmacist,
11
infectious diseases, Meriter Hospital and
12
University of Wisconsin, Madison.
13
DR. PEREZ: Tom Perez, executive
secretary
14 to
this committee meeting.
15
DR. CHESNEY: Joan Chesney,
pediatric
16
infectious disease at the University of Tennessee
17 in
Memphis, and also St. Jude's Children's Research
18
Hospital.
19
DR. HUDAK: Mark Hudak,
neonatologist,
20
University of Florida, Jacksonville.
21
DR. DANFORD: Dave Danford,
pediatric
22
cardiology, University of Nebraska Medical Center,
7
1
Omaha.
2
DR. NELSON: Robert Nelson,
pediatric
3
critical care medicine, Children's Hospital,
4
Philadelphia and University of Pennsylvania.
5
DR. IYASU: Solomon Iyasu, lead
medical
6
officer in pediatrics, FDA.
7
DR. CUMMINS: Susan Cummins, lead
medical
8
officer, pediatrics, FDA.
9
DR. S. MURPHY: Shirley Murphy,
Division
10
Director, Division of Pediatric Drug Development,
11
FDA.
12
DR. D. MURPHY: Dianne Murphy,
Office
13
Director, Office of Counter-terrorism and Pediatric
14
Drug Development, in the Office of Pediatric
15
Therapeutics.
16
DR. CHESNEY: Thank you. Now Tom Perez
17
will read the meeting statement.
18 Meeting Statement
19
DR. PEREZ: Thank you and good
morning.
20 The
following announcement addresses the issue of
21
conflict of interest with regard to the adverse
22
event reporting session and is made part of the
8
1
record to preclude even the appearance of such at
2
this meeting.
3
Based on the submitted agenda for the
4
meeting and all financial interests reported by the
5
committee participants, it has been determined that
6 all
interests in firms regulated by the Center for
7
Drug Evaluation and Research present no potential
8 for
an appearance of a conflict of interest at this
9
meeting, with the following exceptions:
10
In accordance with 18 USC 208(b)(3), full
11
waivers have been granted to the following
12
participants, Dr. Richard Gorman for ownership of
13
stock in a company with a product at issue, valued
14
between $50,001 to $100,000; Dr. Judith O'Fallon
15 for
her and her sponsor's ownership of stock in a
16
company with a product at issue, between $5,001 and
17
$25,000; Dr. Katherine Wisner, for her speaker's
18
bureau activities for a company with a product at
19
issue for which she receives less than $10,001 per
20
year; Dr. Patricia Chesney for her spouse's
21
ownership of stock in a company with a product at
22
issue, valued from $5,001 to $25,000 and unrelated
9
1
consultant earnings less than $10,001 per year. In
2
addition, Dr. Chesney's spouse owns stock in a
3
company with a product at issue, worth less than
4
$5,001. Because this stock
interest falls below
5 the
minimis exception allowed under 5 CFR
6
2640.202(b)(2), a waiver under 18 USC 208 is not
7
required. Further, Dr. Chesney is
recused from
8
participating from the subcommittee's discussion
9
regarding Duragesic due to a conflict of interest.
10
A copy of the waiver statements may be
11
obtained by submitting a written request to the
12
agency's Freedom of Information Office, Room 12A-30
13 of
the Parklawn Building. In the event that
the
14
discussions involve any other products or firms not
15
already on the agenda for which an FDA participant
16 has
a financial interest, the participants are
17
aware of the need to exclude themselves from such
18
involvement and their exclusion will be noted for
19 the
record.
20
We would also like to note that Dr. Samuel
21
Maldonado has been invited to participate as an
22 industry representative, acting on behalf of
10
1
regulated industry. Dr. Maldonado
is employed by
2
Johnson & Johnson. With
respect to all other
3
participants, we ask in the interest of fairness
4
that they address any current or previous financial
5
involvement with any firm whose product they may
6
wish to comment upon. Thank you.
7
DR. CHESNEY: Thank you. Our first
8
speaker for the morning will be Dr. Dianne Murphy,
9
Director of the Counter-terrorism and Pediatric
10
Drug Development Office.
11 Welcome
12
DR. D. MURPHY: And just as you
all
13
understand how those two got to be combined, we
14
have come to the end of an era.
That was really
15 the
substance of my opening comments this morning
16 and
I am going to talk more about this later in the
17
day, that this is a milestone.
18
But I wanted to take this morning to focus
19 on
the importance of the activity of this committee
20 in
the review of the safety or adverse events that
21
occur after a product has been granted exclusivity.
22 It
has been clearly legislatively mandated that
11
1
this is going to occur and that task has come to
2
this committee.
3
I wanted to make sure that you all
4
realized how much you have contributed to this
5
process. We have received
feedback from you during
6 the
time about what was useful and have tried to
7
maintain a course, as we have to, that obeys the
8
legislative intent and, yet, makes it more
9
scientifically interesting within the constraints
10
that we have. I think probably
years from now we
11
could come and ask you all what are the problems
12
with the AERS data reporting system.
So, you have
13
been mandated to participate in a process in which
14 you
were told every meeting that you come here that
15 the
limitations are numerous with passive
16
reporting; that when we do get reporting it is
17
either poor or limited in nature; that there is
18
little ability to go back and reconstruct in detail
19 any
of that information; and it basically doesn't
20
have the same quality as a prospective surveillance
21 or
active process. Yet, during this time I
think
22 we
have evolved a process, again with your feedback
12
1 and
assistance, that has allowed us to make it more
2
valuable.
3
I would like to say that I think that what
4 we
have been able to identify over the past year or
5 so
has been the benefits of this system, and that
6 is
that it ensures that attention is focused on
7
what is happening postmarketing to these products
8
that the government initiates and rewards for
9
studies being conducted. As most
of you are aware,
10 one
of the largest safety databases that occurs
11
with any product is the postmarketing activities.
12
That is where you find your rare serious events.
13
And, this process has been critical for this
14 committee and this has been a very important
15
activity that I do think has focused and ensured
16
that products that are marketed for children are
17
looked at in a studied way, a reliable way, a
18
predictable way, and I think that that is
19
important.
20
Now, why is it important? Because
I don't
21
know how many times you have sat through these
22
meetings where we said, "well, here are the
13
1
problems and we didn't see anything.
Okay?" But
2
that is good news. We would hope
that the majority
3 by
far, if not 100 percent of these products that
4 are
studied and marketed don't have serious hidden
5
adverse events. So, in a say, it
is like
6
prophylaxis. We hope we don't
find major issues.
7
But I think the other thing that this
8
process has done that I wanted you all to know
9
about that was important is that it has the effect
10 on the agency of re-prioritizing pediatric
safety
11
assessments. As everyone knows,
there are many
12
deadlines the agency has to meet and it is hard
13
often to see the plate for all the things that are
14 on
it. But clearly the legislation, your
15
participation and our coming to you says we are
16
having a public meeting and a discussion and it
17
re-prioritizes this activity for the agency, as I
18
said, and ensures that attention occurs.
19
We are going to hear today about some
20
activities that have evolved during this process,
21
some questions that we want to bring to you because
22 of
information that, in essence, was moved forward
14
1 a
little faster because of this process, not that
2 it
was being neglected but because we basically
3
made sure that we facilitated the assessments of
4
some of these products and some of the issues. In
5 the
past, as you know, we have had some reviews of
6 the
SSRIs and this whole process has been important
7 in
helping facilitate moving that activity forward
8
also.
9
I wanted to just thank you for your
10
scientific input, your thoughtfulness and your
11
feedback which we still would like to receive about
12 the
process on adverse event reporting, knowing
13
that we have to work within the constraints of the
14
systems that we presently have.
With that, I will
15
speak a little more about the contributions of this
16
committee and where we are going in the future
17
later today. Thank you very much.
18
DR. CHESNEY: Thank you, Dr.
Murphy. Our
19
second speaker this morning, Dr. Solomon Iyasu, is
20
going to talk to us about adverse event reports,
21 per
Section 17 of the Best Pharmaceuticals for
22
Children Act. Dr. Iyasu is a
pediatrician, a
15
1 medical
epidemiologist who has fellowship training
2
with the EIS of the CDC and residency training in
3
preventive medicine at the CDC.
Prior to joining
4 the
FDA, just in 2002, he worked for 13 years as a
5
medical epidemiologist at the CDC, in Atlanta,
6
where he led research and programmatic programs in
7
infant health. He also served as
the CDC liaison
8 to
the Committee on the Fetus and Newborn of the
9
American Academy Pediatrics for many years, and has
10 served
on several HHS committees and inter-agency
11
working groups, including the National Children's
12
Study. His research papers have
involved maternal
13 and
child health issues. In his current
position
14 at
the FDA he serves as a medical team leader in
15 the
Division of Pediatric Drug Development and also
16
serves as the lead medical officer for safety in
17 the
Office of Pediatric Therapeutics, which has
18
become--always was but has become a particularly
19 important
office in function. Dr. Iyasu?
20
Adverse Event Reports per Section 17 of Best
21 Pharmaceuticals for Children
Act
22
DR. IYASU: Thank you very much,
Dr.
16
1
Chesney, for that kind introduction.
Good morning.
2
In the next few minutes I will provide you
3
with an overview of today's agenda.
The theme for
4
today is safety, safety of pediatric drugs. A
5
series of presentations will discuss postmarketing
6
reviews of adverse events for drugs that have been
7
granted exclusivity.
8
The review of the post exclusivity adverse
9
events is accomplished through the collaboration
10
with the Office of Drug Safety, Office of Pediatric
11
Therapeutics and Division of Pediatric Drug
12
Development. Therefore, at first
I would like to
13
acknowledge the contribution of the staff in the
14
Office of Drug Safety for these reviews.
15
In the morning you will hear adverse event
16
reviews for eight drug products that were granted
17
pediatric exclusivity. These
reviews will be
18
presented by medical officers within the Division
19 of
Pediatric Drug Development. Several of
these
20
presentations are informational while a few discuss
21
important issues, ranging from a lack of
22
age-appropriate pediatric formulations for
17
1
fosinopril to a preventable safety signal
2
associated with the use of fentanyl transdermal
3
system or Duragesic. You will be
asked to discuss
4 a
question of risk management strategies in
5
relation to fentanyl. The morning
will also
6
include a time for open public hearing, followed by
7 a
short pediatric update by Dr. Dianne Murphy.
8
We are doing the adverse event review a
9
little differently than before.
In addition to the
10
usual format which you are familiar with, we have
11
incorporated some of the clinical trial data
12
available in the public domain into these reviews.
13 You
are not going to see this component for all the
14
drugs because the trial data are not yet in the
15
public domain for some of the drug products that we
16
will be discussing.
17
This is a pediatric page on the external
18 FDA
website where you will find all the publicly
19
available summaries of medical and clinical
20
pharmacology of these pediatric studies for
21
exclusivity. The process of
making these reviews
22
available in the public domain is evolving,
18
1
therefore, some of the reviews that I mentioned
2
before may not be yet available on this website.
3
Nevertheless, I invite you to use it as a resource
4 and
urge you to spread the word about this site.
5
In the afternoon we will discuss two
6
pediatric safety issues regarding the use of SSRIs
7 and
SNRIs during pregnancy. As you recall,
we
8
discussed several case reports of neonatal
9
withdrawal syndrome related to the use of Paxil and
10
Celexa during the meeting of this committee last
11
February. At that time you
requested more
12
information on the syndrome and FDA's efforts to
13
address it.
14
To address this issue, we have lined up
15
three presentations for you. Kate
Phelan, from the
16
Office of Drug Safety, will present the
17
postmarketing adverse event review for this class
18 of
drugs. Dr. Bob Levin, from the Division
of
19
Neuropharmaceutical Drug Products, will speak on
20 the
new class labeling regarding neonatal
21
withdrawal toxicity and its rationale.
Dr. Kathy
22
Wisner will address the risk/benefit of treatment
19
1 in
child depression, a critical issue for both the
2
practitioner and the patient. At
the end of this
3
update you will be asked to discuss two questions.
4
Next, I will present an update on
5
congenital eye malformations, again, as a fallout
6 to
the February meeting when we reported a case
7
report about possible congenital eye malformation
8
related to the use of Celexa during pregnancy.
9
This update will review all postmarketing reports
10 of
congenital eye malformations for Celexa and some
11
newer anti-depressants.
12
Before we present the specific adverse
13
events, I will briefly review the data sources used
14 in
this review and their limitations. The
Adverse
15
Event Reporting System is a spontaneous and
16
voluntary system. Because it is a
passive system
17 it
suffers from a number of limitations, listed
18
here on this slide, that you are already familiar
19
with and we have discussed several times during
20
previous presentations.
21
Again the drug use data source and their
22
limitations have also been presented before and are
20
1 not
new to you. IMS National Prescription
Audit
2
Plus is used to estimate the number of outpatient
3
prescriptions but lacks demographic information.
4 The
National Disease and Therapeutic Index can
5
estimate drug mentions during office-based
6
physician visits but pediatric use estimates can be
7
unstable for less frequently used medications.
8
Another outpatient data source is the IMS
9
National Sales Perspectives which provides
10
estimates of units sold from manufacturers to
11
various channels of distribution and, therefore,
12 may
be a possible surrogate measure for drug use.
13 An
important limitation of this data source is
14
absence of demographic information such as age and
15
gender.
16
Important drug use data sources and their
17
limitations are well-known to you.
To refresh your
18
memory, these are described in this slide and the
19
next slide. The main limitation
with all the
20
inpatient data sources, except for Premier, is the
21
inability to make national projections of drug use.
22
However, national estimates from Premier are
21
1
available but are selective.
Furthermore, drug use
2
cannot be linked to diagnosis or procedure and drug
3 use
in hospital or outpatient clinics is not
4
captured in this data system.
Data from CHCA are
5
limited to 29 children hospitals and cannot be
6
projected nationally.
7
This concludes my remarks and now let me
8
turn to the presentations for this morning by
9
introducing the first speaker.
But before I do
10
that, I do want to recognize two individuals who
11
have tirelessly worked behind the scenes to make
12
this meeting possible. Please
stand up and be
13
recognized, Miss Christine Phucas and Rosemary
14
Addy.
15
[Applause]
16
Thank you. Now the next speaker,
Dr.
17
Filie is a general pediatrician and pediatric
18
rheumatologist. She conducted
research on
19
molecular biology, connective tissue disorders and
20
genetics at NIH for many years before going into
21
private practice. She joined the
FDA from private
22
practice about a year ago. She
will discuss
22
1
adverse event reports for fexofenodine.
Dr. Filie?
2 Fexofenodine
3
DR. FILIE: Good morning,
everyone. I
4
will proceed with the adverse event review for
5
fexofenodine during the one-year post-exclusivity
6
period.
7
Fexofenodine, trade name Allegra, is an
8
antihistamine by Aventis Pharmaceuticals. The
9
indications for adults and children are relief of
10
symptoms associated with seasonal allergic rhinitis
11 and
non-complicated skin manifestations of chronic
12
idiopathic urticaria. It was
originally approved
13 in
July, 1996 and pediatric exclusivity was granted
14 in
January, 2003.
15
In order to fulfill the requirements for
16
exclusivity, 3 pivotal studies were conducted and
17 415
children, 6 months to 6 years of age, were
18
treated for allergic rhinitis.
One study was a
19
Phase 1 pharmacokinetic study characterizing the
20
dose for children 6 months to 2 years of age.
21
Another study was a Phase 3 study assessing safety
22 and
tolerability in 2 groups, 6 months to 2 years
23
1 of
age, weighing under 10.5 kg and weighing over
2
10.5 kg.
3
A previous safety and tolerability study
4 on
children 2-6 years of age was also submitted.
5 The
adverse events occurred at similar frequencies
6 as
for placebo, and no new safety signals were
7
observed.
8
Efficacy studies were not conducted due to
9 the
fact that the disease course and
10
pathophysiology of allergic rhinitis and chronic
11
idiopathic urticaria, as well as the drug's effect,
12 are
similar in children and adult patients.
The
13
studies conducted on children 6 months to 6 years
14 of
age utilized fexofenodine powder mixed with
15
apple sauce or rice cereal. There
is no marketable
16
age-appropriate formulation for children 6 months
17 to
6 years of age.
18
Drug use trends for
19
fexofenodine--currently, fexofenodine is the
20
leading prescription for non-sedating antihistamine
21 on
the market since loratadine became
22
over-the-counter in 2002. The
total number of
24
1
fexofenodine product dispensed increased from
2
approximately 20.9 million in 2000 to 29.6 million
3 in
2003. Pediatric patients accounted for
4
approximately 2.5 million prescriptions of
5
fexofenodine dispensed in 2003.
The most common
6
diagnoses associated with the use in pediatric
7
patients in 2003 were allergic rhinitis and
8
allergic disorder.
9
The adverse events from pediatric clinical
10
trials that I just presented are as follows:
11
Headache, accidental injury, cough, fever, pain,
12
otitis media and upper respiratory infection, and
13
least common, insomnia, nervousness, sleep
14
disorders, rashes, urticaria, pruritus and
15
hypersensitivity reactions.
16
During the exclusivity period the total
17
adverse event reports from the AERS database was
18
158, 84 of them in the United States.
Among the
19 158
reports there were 8 unduplicated pediatric
20
reports which included 2 with serious outcomes, 1
21
hospitalization and 1 life-threatening event.
22
There were no pediatric deaths.
25
1
In the 8 pediatric case reports the
2
following unlabeled pediatric adverse events were
3
reported, psychosis exacerbation with suicidal
4
ideation and depression; seizure, visual
5
disturbances; abnormal liver function; fungal
6
urinary tract infection; non-accidental overdose of
7
multiple drugs and prolonged QT, prematurity,
8
maternal experience and medication error.
9 I would like to present you with a
10
synopsis of individual reports. A
15 year-old with
11
schizoaffective disorder and ADD, on multiple
12
medications, experienced exacerbation of psychosis,
13
suicidal ideation and depression which resolved
14
after discontinuation of fexofenodine.
15
A 13 year-old child presented with grand
16 mal
seizures. The patient was also on
multiple
17
medications and one of them was bupropion which has
18 a
warning about the potential to cause seizures.
19
A 7 year-old presented transient loss of
20
color vision and visual disturbances such as black
21
dots and bubbles. It also
resolved after
22
discontinuation of the drug in a few days.
26
1
A 10 year-old patient developed a
2
bacterial UTI and abnormal liver function tests
3
after receiving fexofenodine for one week. The
4
child was on concomitant medications and one of
5
them was labeled for hepatic function impairment.
6 We
do not have the name of the drug on the report.
7 The
child recovered after discontinuation of
8
fexofenodine.
9
A 16 year-old who developed a fungal UTI
10 and
pyelonephritis was hospitalized. This
patient
11 was
also on multiple medications for depression and
12
gastritis.
13
A 13 year-old had an intentional overdose
14 of
fexofenodine, acetaminophen, metoclopramide and
15
tramadol. QT prolongation was
observed in the
16
emergency room which normalized the following day.
17
The last two cases--a 27-week old
18
premature baby, small for gestational age, was born
19 by
C-section due to pre-eclampsia. There
was a
20
history of abnormal alpha-1 fetoprotein.
The
21
mother was on concomitant medications.
22
The last case--a prescription refill was
27
1
mistakenly filled with Zyrtec-D instead of
2
Allegra-D, but no adverse event was reported.
3
Concluding the report, despite the large
4
number of fexofenodine prescriptions, there were
5 few
pediatric adverse event reports during the
6
one-year post-pediatric exclusivity period. It is
7
also very difficult to make any attributions of the
8
adverse events of the drug when there are
9
concomitant medications in the reports.
In this
10
case, the FDA will continue to monitor the adverse
11
event reports in all populations.
Any questions or
12
comments?
13
DR. CHESNEY: Dr. Santana?
14
DR. SANTANA: Do you know if there
are any
15
similar adult reports with the use of this
16
medication and concomitant anti-psychotic
17
medications in adults?
18
DR. FILIE: I don't know that I
can
19
respond to that adequately. From
the information
20
that we have on the label, the adverse events are
21
very similar in both populations.
They resemble
22
pretty much the two groups.
28
1
DR. S. MURPHY: Pete Stark I think
is here
2
from the Division. Do you have
any comments about
3
adult report?
4
DR. CHESNEY: Dr. O'Fallon?
5
DR. O'FALLON: It seems to me that
the way
6 you
keep your data may help you to find things.
7 So,
I am wondering when you have these reports, are
8 you
keeping track of the various concomitant
9
medications so that you could be looking for trends
10
developing that may be subtle, that there may be
11
interactions, or something?
12
DR. FILIE: Yes. The hope is to
13
accumulate this data over a long time.
14
DR. O'FALLON: Yes, but I mean in
a way so
15
that you are able to go back, search and find those
16
combos? I am asking about how the
data is being
17
collected so that you are going to be able to
18
search on it.
19
DR. FILIE: Yes, it is possible
and we are
20
doing that collecting and the Office of Drug Safety
21 is
also involved in this. This is something
that
22 has
accumulated and we can keep all this data
29
1
without losing it.
2
DR. O'FALLON: But in a computer
file that
3 you
can search?
4
DR. FILIE: I don't know.
5
DR. IYASU: Let me respond to this. The
6
AERS database has been in existence for a long time
7 and
the database is searchable both by high risk
8
event terms as well as by the drug name or the
9
trade name. So, it is searchable
by a number of
10
parameters and there is an accumulated database
11
which resides at FDA so you can look at one year or
12 you
can look at several years since the first time
13 a
report comes into existence for a particular
14
product. Once there is approval,
there are going
15 to
be postmarketing reports that come in.
So,
16
there is a way to look at that.
But there isn't a
17
whole lot of information to try to look at multiple
18
permutations of different confounders or looking up
19
interactions. It is a limited
database in that
20
way.
21
DR. D. MURPHY: I did want to
respond that
22 in
your package it does tell you that fexofenodine
30
1 has
been looked at with the co-administration of
2
acetyl console and erythromycin, the sip
3
interactions. So, what the agency
does is where we
4
know that a metabolism uses a certain sip enzyme
5
that will cause increases or decreases, they will
6
frequently look at that interaction but they can't
7
look at all of them. That often
is actually a
8
negotiated activity as to how many of them they do
9
look at, and whether there are ones that are more
10
likely to give serious adverse events by the normal
11
drugs that might be used with this specific
12
disease. So, you could see that
with an allergic
13
indication you might think that antibiotics would
14 be
one of the set of drugs that they would look at.
15
So, I just wanted to put on the table that
16
prospectively the agency will sometimes ask,
17
knowing what the metabolism is, for these
18
interactions. But, you can
imagine that the list
19 could
get endless so the agency does not do all
20
possible combinations. Certainly,
I think from
21
allergic rhinitis to antidepressants--I mean,
22
unless you had a mechanistic reason for doing that,
31
1 you
wouldn't up front do it. Your question,
I
2
realize, was looking at statistical analysis post
3 but
up front there is a certain amount of activity
4 in
that area.
5
DR. O'FALLON: It seems to me that
since
6 you
only have a handful of reports it might be
7
worth it, that when you see something showing up
8 you
would say they took drug A, drug B, drug C,
9
let's look and see if we have any reports in the
10
database, especially in the adults or something, to
11 see
if you are seeing if that has been reported
12
before.
13
DR. D. MURPHY: As noted, ODS has
the
14
database and it will have that information in it.
15 So,
you could go back and plug in certain drug
16
names. I think, as always, the
caveat is that
17
there are those who didn't enter that and were on
18 it
so there is always that question of what does it
19
mean when you do it. But, you are
right, if you
20 kept
seeing that pattern, then it would be
21
something you might wish to pursue further and ask
22 for
some additional studies.
32
1
DR. CHESNEY: Dr. Gorman?
2 DR. GORMAN: This is mainly for
3
clarification from my reading of the labeling. On
4
page 7 of the label for this product there is a bar
5 on
the side and I wanted to know whether this was
6
edited out of the label or is the present labeling
7
wording which says that the safety and
8
effectiveness of fexofenodine in pediatric patients
9
under 6 years of age has not been established. Is
10
that in the label now or out of the label?
11
DR. D. MURPHY: It is not labeled
under 6.
12 Is
that right?
13
DR. GORMAN: It is a question of
the bar
14
because it comes up several times later on in
15
labeling.
16
DR. D. MURPHY: Right, right. We will
17
verify this but I think the point was that because
18
there was no formulation that was available, it is
19 not
labeled under 6.
20
DR. GORMAN: I think one of the
issues
21
that was raised at the last meeting, and I would
22
like to have it reemphasized again is that there is
33
1 now
data. When we started this process two
decades
2
ago, that statement meant that there were no
3
studies. Now it means there may
well be studies
4 but
it is not included in the label. I
noticed in
5 the
executive summary, which will be available on
6 the
web-based FDA data, that there is information
7
about its use in children less than 6 months of
8
age.
9
DR. D. MURPHY: I think you
referred to
10 the
clinical pharmacology and biopharm study.
11
Unfortunately, it doesn't have a page number but it
12 is
after the label. It does say in there
that no
13
labeling changes for pediatric indication or dosing
14 for
children less than 6 years old will be made at
15
this time because there are no age-appropriate
16
formulations for fexofenodine for these children,
17 and
your point being that it was studied. And,
18
that is not going to be put in the label and I
19
think that is an issue.
20
DR. GORMAN: That is the issue I
wanted to
21
raise and it will now be raised by others for the
22
rest of the meeting.
34
1
DR. CUMMINS: Can I just provide
one point
2 of
clarification? The labels that we
provide to
3 you
are ones that are publicly available and are
4 the
most recent labels. Often the strikeouts
are
5
still present. We download them
from the labels
6
that are posted on the web often--you know, that we
7
post on the FDA website. If you
see a strikeout,
8 as
you see on page 7, then that strikeout will be
9
removed in the published label by the company.
10
DR. GORMAN: Thank you.
11
DR. CUMMINS: You are welcome.
12
DR. FILIE: Given there are no
further
13
comments or questions, let me introduce the next
14 speaker, Dr. Susan McCune. Dr. McCune is a
15
neonatologist whose previous experience includes
16
academic neonatal practice at Johns Hopkins and
17
Children's National Medical Center.
She recently
18
received her masters degree in education and has
19
worked on computer-based education models for
20
pediatrics. She will discuss two
oncology
21
products, topotecan and temozolomide.
Dr. McCune.
22 Topotecan and Temozolomide
35
1
DR. MCCUNE: Thank you very much,
Dr.
2
Filie. Ladies and gentlemen of
the committee and
3
guests, Drs. Murphy told me to try to keep things a
4
little bit light to keep you all awake and my Irish
5
ancestry would allow me to tell shaggy dog stories
6
but, unfortunately, I don't do very good jokes so I
7
think we will just move along.
8
As Dr. Filie mentioned, I will talk about
9 two
oncologic agents this morning. The first
is
10
topotecan. Topotecan, trade name
Hycamtin, is an
11
anti-tumor oncologic agent produced by
12
GlaxoSmithKline. The indication
in adults is
13
metastatic carcinoma of the ovary after failure of
14
initial or subsequent chemotherapy and small cell
15
cancer sensitive disease after failure of
16
first-line chemotherapy. There
are no approved
17
pediatric indications. The
original market
18
approval was May 28, 1996 and the pediatric
19 exclusivity
was granted on November 20, 2002.
20
I am going to tell you about the studies
21 for
exclusivity for this drug. As you all
22
mentioned, in terms of data that is available for
36
1 the
label, these studies were done based on what
2 Dr.
Iyasu told you already. BPCA mandates
that
3
this information be available on the website and
4
this information is available on the website,
5
however, there were no changes to this label based
6 on
this information.
7
The studies that were submitted for
8
exclusivity were summaries of studies that were
9
previously performed by the Pediatric Oncology
10
Group. They were initiated in
1992 and 1993. This
11 was
a Phase 2 study in pediatric solid tumor that
12
enrolled 108 patients that were less than 16 years
13 of
age. The tumor types were Ewing's
sarcoma,
14
peripheral neuroectodermal tumor, neuroblastoma,
15
osteoblastoma and rhabdomyosarcoma.
The study
16
endpoint was tumor response rate.
Eighty-six
17
percent of patients died, with 10 percent dying
18
within 30 days of the last dose of topotecan. The
19
overall response rate was 8 percent but the
20
response rate for patients with neuroblastoma was
21 18
percent. Of note, it is important to
know that
22 for
alternative regimens using combinations of
37
1
available drugs in pediatric patients with relapse
2
neuroblastoma the response rates were 35-50
3
percent. In this case, no
patients less than 2
4
years of age showed any response.
5
Eight of the 11 patients that died within
6 30
days of the last dose of topotecan had
7
progressive disease and 3 died with infection which
8 is
a known complication. Forty-four percent
of
9
patients were hospitalized with adverse events,
10
primarily febrile neutropenia, fever or sepsis.
11
The Phase 2 study did determine a
12
different dose from adults, a daily infusion for 5
13
consecutive days every 21 days.
The adult dose is
14 1.5
mg/m
2/day and the
pediatric
dose that was given
15 was
either 1.4 mg/m
2/day without granulocyte-colony
16
stimulating factor or 2 mg/m
2/day with
17
granulocyte-colony stimulating factor.
18
In terms of drug use trends in topotecan
19 in
the inpatient setting, between July, 2001 and
20
June, 2003 there were 10.6 percent of discharges.
21
Just to give you a rough idea, compared to the last
22
drug which had a number of prescriptions, this was
38
1
only 425 of 4,001. Pediatric
topotecan did
2
increase annually in that time period, from 6.8 to
3
18.6 percent. It accounted for
407 discharges from
4 29
CHCA free-standing pediatric hospitals, with the
5
most frequent diagnosis being chemotherapy
6
encounter followed by malignant neoplasm of the
7
adrenal gland. A significant
limitation, as we
8
have already discussed, of the analysis is that the
9 FDA
does not currently access data capture in the
10
outpatient hospital clinic setting where most
11
chemotherapy is administered.
12
Now I am going to tell you about the
13
adverse event reports for topotecan for the
14
one-year post-exclusivity period.
There were 29
15
total reports for all ages, 18 in the United
16
States. There were no pediatric
reports that were
17
submitted during this time. Of
note, in the 7-year
18
period from 1996 there were some unlabeled
19
pediatric reports, none of them during that 1-year
20
post-exclusivity period. There
were 4 reports of
21 convulsion,
hypotension, edema and speech
22
disorder, and 3 reports each of arachnoiditis,
39
1
ascites, Budd Chiari syndrome, caecitis and
2
confusional state.
3
In summary, the FDA will
continue its
4
routine monitoring of the adverse events in all
5
populations. I will stop here and
take any
6
questions on this particular drug.
7
DR. CHESNEY: Dr. Santana?
8
DR. SANTANA: I think I have made
this
9
point before and I will try to reinitiate it again.
10 In
contrast to some of the other drugs that we have
11 in
front of us, the oncology drugs are usually used
12 in
the setting of clinical research. They
are not
13
used in the setting of common practice.
So, there
14 is
a wealth of data from protocols either initiated
15 by
the historically previous oncology groups or the
16
current Children's Oncology Group and certainly by
17 other
large institutions like St. Jude's that do
18
research in these drugs. How is
that data captured
19 and
reflected in these reports? Because
there is a
20
wealth of adverse event data that is generated
21
through that clinical research that will not show
22 up
through these voluntary reporting mechanisms but
40
1
will show up in the databases of the clinical
2
research infrastructure.
3
DR. MCCUNE: A lot of the reports
that we
4 get
for these particular drugs are actually from
5
study reports. In terms of the
studies that were
6
done for exclusivity for this drug, they actually
7
were, as you mentioned, part of the research
8
protocols so they were independent studies
9
conducted by the company.
10
DR. SANTANA: But I guess the
point is
11
that that is true but there is a lot more usage of
12
this drug now, as you indicated in your brief
13
summary of the trends of usage of this drug in
14
pediatric oncology. How is that
data eventually
15
going to make it into the adverse event reporting?
16
Because it is not really part of the exclusivity
17
because those studies have not been submitted for
18
exclusivity. Am I correct?
19
DR. MCCUNE: That is correct.
20
DR. SANTANA: These are studies
that are
21
ongoing.
22
DR. MCCUNE: That is correct. This is the
41
1
one-year post-exclusivity period.
2
DR. SANTANA: How will that data
show up
3 in
the current study?
4
DR. S. MURPHY: It would have to
come
5
through the AERS. It would have
to be submitted to
6
AERS for us to have that information.
Dr.
7
Maldonado may want to comment, but the companies
8
have to report any adverse events to the FDA. So,
9 the
companies, you know, keep very close tabs on
10 the
medications, especially the medications that
11 are
in trials that are using their drugs.
So,
12
there is a sort of cross-reference thing. Then, it
13 is
even global with the pharmaceutical companies
14 and
with the international organizations with the
15
FDA. So, I think it is a very
good question. I
16
think Don Mattison might want to make a comment,
17
from NIH.
18
DR. MATTISON: Just a brief
comment. We
19 are
currently working with NCI and COG to develop
20
full access to their databases and that information
21
will be shared with FDA.
22
DR. D. MURPHY: Dr. Santana, I
think if
42
1 you
look at what is in the label now, it just says
2
that the effectiveness in children has not been
3
demonstrated. Then it goes ahead
and it does
4
describe the studies. As you
know, for cancer this
5 has
been a real issue because of the reasons you
6
have stated. The label is
marketing approval and
7 if
it is not approved for that indication, you
8
know, the agency is in this quandary of how do you
9
make information available when you don't want to
10
give a de facto indication that doesn't exist? So,
11
that is the tension here.
Depending on the
12
product, depending on what comes out of the
13
exclusivity studies if we don't have sufficient
14
evidence to say it is efficacy and, as you know,
15 and
I don't want to say this over and over again,
16 but
these studies are not powered to do that.
So,
17 how
do we make that information available has been
18
difficult.
19
I think what they have done here is that
20
they have been able to put into--by saying it has
21 not
been demonstrated, first, and saying yet we
22
looked and here is what we found in a very limited
43
1
way, and then having some adverse event reporting
2
that came out. Now, does it
happen for every
3
product, every time? Not always
because it may be
4
that there were other issues with the studies and
5
then what you may end up with in the label if there
6 is a particular safety thing, they would say
it was
7
studied in so many kids; it wasn't effective or we
8
couldn't determine effectiveness but we are going
9 to
tell you about these adverse events. So,
that
10 can
happen. The adverse events in those
studies
11
could be put into the label if it is a safety
12
issue.
13
DR. SANTANA: I guess what I am
getting at
14 is
that the information that is derived from
15
granting exclusivity is for the studies that the
16
sponsor has put forth to reach that point.
17
DR. D. MURPHY: Right; that is
correct.
18
DR. SANTANA: But there is another
wealth
19 of
data that is being generated. As I
understand
20 it,
unless it is throught the sponsor or through
21
some other mechanism that data becomes available to
22 the
FDA it is not part of the information that we
44
1
have in front of us today or in the future.
2
DR. D. MURPHY: Well, it is
required to be
3
reported to the FDA. It is
required to be reported
4 and
if the agency sees a signal, then there is a
5
re-review of the data and a determination if that
6
additional information needs to be entered into the
7
label. I would say that if a
researcher had access
8 to
data that they were concerned about and saw that
9 it
wasn't in the label, it is perfectly appropriate
10 to
ask--you know, again, it is a requirement.
11
Companies get into big trouble if they have adverse
12
events that they don't report to us.
13
The other issue--I am not saying it
14
happens, but if somehow you thought something
15
wasn't getting reported, it is perfectly
16
appropriate to call the agency and say I am aware
17 of
this; make sure you got those reports.
18
DR. SANTANA: I want to make it
clear for
19 the
public record that I am not raising issues with
20
this drug or the next oncology drug.
I am trying
21 to
understand the process. I just want to
make
22
that clear.
45
1
DR. D. MURPHY: Yes, and we want
to make
2 it
clear that it is part of companies' standard
3
reporting activity. Sam, maybe
you could say
4
something about the routine things that go on in
5
reporting both during a trial and after a product
6 is
marketed.
7
DR. MALDONADO: Both of you are
completely
8
right. Companies are not going to
get in trouble
9 by
not reporting. That is very
enforceable. A lot
10 of
not reported events happen when physicians don't
11
report to companies. So, that is
where the problem
12 is;
it is the education. We are not only
talking
13
about sending in the reports, but sending them
14
within 15 days of occurrence.
Most of the
15
non-reporting happens because of lack of education
16
from clinicians. In clinical
trials it happens
17
much less, or probably very, very close to zero
18
because there is monitoring by the company. Actual
19
people go there and make sure they are doing it.
20
Outside clinical trials it is more difficult
21
because you cannot police physicians so it is up to
22
them to report. But once it is reported to the
46
1
companies, it is reported to the FDA and the FDA,
2 of
course, can always come to a company and check
3 if
we are doing it and actually FDA does that.
4
DR. D. MURPHY: I think what Sam
has said
5 is
really important. If a physician sees an
6
adverse event on a product, particularly if you put
7
them on a product, take them off and put them back
8
on--you know, if you have evidence, but even if you
9
don't, if you put a child on a product and you have
10
some serious event and you are not sure whether it
11 is
related or not, you don't have to make
12
attribution. This is one of the
problems I think
13
physicians don't understand. You
don't have to
14
determine individually that this product caused
15
this adverse event. If physicians
would, please,
16
make it part of their public health rule to report
17
adverse events that they think are serious to the
18
agency and to the company, I mean, that is a double
19
way--or either way, you know, whichever way you
20
know how to get that information in.
It will get
21 to
us if it gets to the company or it can come to
22 us
directly. So, I would like to keep
adding that
47
1
commercial. It is a very important
part of
2
activity. I have been out there;
I have practiced
3
medicine and I know I haven't done it when I should
4
have. So, it is just a plea that
we keep putting
5
that out there because you can see how important it
6 can
become.
7
DR. CHESNEY: Dr. O'Fallon?
8
DR. O'FALLON: There is one other
issue
9
that is a possible problem. I
don't know these
10
particular studies that COG is doing but if they,
11
indeed, have closed patient accrual before the
12
exclusivity period it is entirely possible that the
13
acute toxicities wouldn't be available at this
14
time. You know, not all the data
in these clinical
15
trials gets reported out until the final study is
16 done. I mean, the company had to know about it
17
ahead of time, but during this exclusivity period
18
there maybe weren't any from those trials.
19
DR. D. MURPHY: I think that
brings up the
20
other issue just of any follow-up post-trial. As
21 you
know, there was a legislative mandate also to
22 put
the 1-800 MedWatch number on labels and that
48
1
process is proceeding. I don't
have any idea when
2
actually you will see it but it is continuing to
3
move forward.
4
DR. CHESNEY: Dr. Ebert?
5
DR. EBERT: Just a follow-up to
that, is
6 it
feasible or even reasonable with these drugs
7
that are specifically under exclusivity for the FDA
8 to
make pediatricians more aware of the fact that
9
they are under this particular scrutiny?
And,
10
would it heighten their level of interest with
11
regards to reporting adverse events?
12
DR. D. MURPHY: Joan has a suggestion for
13 you
later today I think about maybe one way of
14
doing it. We have been trying to
do that in a
15
number of ways by working with the American Academy
16 of
Pediatrics newsletter that goes out and doing
17
annual updates of changes in the label, talking
18
about exclusivity, but I think you bring up a good
19
point--have we really made an issue in that
20
reporting about changes in label about reporting
21
adverse events? No. And, that is a good point and
22 we
will take that back and pursue that as an
49
1
additional piece of information we should try to
2 get
out to pediatricians, family practice, people
3 who
are taking care of children. We are
working
4
with the Academy on the CME activity so that we can
5 put
in some case studies that might bring that up.
6
DR. S. MURPHY: Joan, just one
more point,
7 there
are really two ways of reporting adverse
8
events. One is to the FDA and the
other is to the
9
companies. The larger
pharmaceutical companies
10
have these 1-800 numbers and if you call and you
11 say
you have an adverse event, you are immediately
12 put
in touch with the Pharm.D. who has a whole
13
scheme of questions to ask you right away. All
14
those reports, like Sam said, do go back to the FDA
15 and
the seriousness of the report triggers certain
16
times to report it. Having been
on the other side
17 in
a pharmaceutical company, I was in charge of a
18
drug that had a lot of adverse reactions and we
19
were constantly reviewing all the cases that came
20
in. The company will often send
somebody out to
21 the
hospital to look at the records and make sure
22 of
the accuracy of the reporting. So, it is
taken
50
1
incredibly seriously on both sides.
2
DR. CHESNEY: Thank you.
We can move on
3 to
the next speaker.
4
DR. MCCUNE: Actually, I am doing
the next
5
drug. You get to listen to me
again. The next
6
drug I am going to talk about is temozolomide. The
7 trade
name for this is Temodar. Once again,
this
8 is
an oncologic agent produced by Schering Plough
9
Research Institute. The
indication in adults is
10
that the capsules are indicated for the treatment
11 of
adult patients with refractory anaplastic
12
astrocytoma, in other words, patients at first
13
relapse who have experienced disease progression on
14 a
drug regimen containing a nitrosourea and
15
procarbazine. In pediatrics there
are no approved
16
pediatric indications. The
original market
17
approval was August 11, 1999; the pediatric
18
exclusivity was granted November 20, 2002.
19
Once again, I am going to tell you about
20 the
studies for exclusivity. These are
available
21 on the
website. In addition, for this
particular
22
label safety information is included in the
51
1
pediatric section of the precautions part of the
2
label and it does include a description of the
3
clinical studies that were completed.
4
The studies that were submitted for
5
exclusivity were one Phase 1 and two Phase 2
6
open-label, multicenter studies.
The Phase 1 study
7 was
dose escalation in 27 patients with advanced
8
non-CNS and CNS cancers. The
first Phase 2 study
9 was
in 63 patients with recurrent brain stem glioma
10 and
high grade astrocytoma. The second Phase
2
11
study, a cooperative group-sponsored study, was in
12 122
patients with various recurrent CNS tumors.
13 The
patients ranged in age from 1 to 23 years of
14
age, with the majority of patients between 3 and 17
15
years of age.
16
The primary endpoint for these studies was
17
tumor response rate. In the first
Phase 2 study
18
there was 1 complete response and 3 partial
19
responses among 27 patients. In
the second study
20
there were no complete responses or partial
21
responses in the brain stem glioma patients and no
22
complete response and 12 percent partial responses
52
1 in
the high grade astrocytoma patients. In
the
2
third study the overall response rate, combined
3 complete response and partial response rate,
was 5
4
percent. Only 1 patient achieved
complete response
5 and
5 patients had partial responses.
6
Safety was assessed in 204 patients at
7
doses of 100-200 mg/m
2/day daily for 5 days every
8 28
days. The toxicity profile that was seen
was
9
similar to adults. The most
common adverse events
10
that were reported were dizziness, neuropathy,
11 paresthesia, nausea/vomiting, constipation
and
12
myelosuppression.
13
Just to give you an idea of the drug use
14
trends in the outpatient setting for temozolomide,
15 the
number of prescriptions dispensed has nearly
16 doubled
over the past 3 years from 50,000 in 2001
17 to
93,000 in 2003, with the top prescribers, as you
18 can
imagine, being oncology/neoplastic, neurology
19 and
hematology. Of note, only 1 percent of
20
temozolomide prescriptions were written by
21
pediatricians.
22
The pediatric population of 1-16 years of
53
1 age
accounted for a small number of temozolomide
2
prescriptions, 3.1 percent in 2002 and 3.9 percent
3 in
2003, with the most frequent diagnosis being
4
malignant neoplasm of the brain both in adults and
5
pediatric patients.
6
In terms of outpatient sales, they have
7
been on the rise, from 1.8 million capsules to 2.2
8
million capsules in the last 2 years, with the
9
majority of sales through retail channels, 80
10
percent of them going to chain and independent
11
pharmacies and other retail channels.
12
CHCA data demonstrated from 2002 to June,
13
2003 that there were only 17 pediatric discharges
14
associated with this drug.
15
The limitations to drug use data in the
16
outpatient setting for these drugs are important to
17
note because we don't have sources that
18
specifically examine outpatient hospital clinics
19
where chemotherapy treatments are provided. What
20 is
important to note though is that the retail
21
sales do capture a number of those sources and it
22 is
felt that most of the use of this drug is
54
1
captured through assessment of outpatient use.
2
In terms of adverse event reporting for
3 the
post-exclusivity period from November, 2002 to
4
December, 2003 there were 250 reports in all ages,
5 160
of them in the United States. There were
5
6
unduplicated pediatric reports, 2 of them in the
7
United States, all with serious outcomes and 1
8 death. There were 4 females and 1 male. Three of
9 the
patients were aged 2-5 years; 2 of the patients
10
6-11 years. There was one patient
each for the
11
diagnoses of blastoma, adrenal metastatic
12
neuroblastoma, anaplastic astrocytoma,
13
medulloblastoma and brain stem tumor.
14
The clinically significant unlabeled
15
adverse events could be divided into 5 groups. One
16 was
brain edema; 1 was death. Another,
hemangioma
17
acquired; another ITP and another myelodysplastic
18
syndrome. All of these, although
not specifically
19
delineated in the label, are potentially related to
20
either a labeled process or the underlying disease
21
state.
22
Just to take each one of these
55
1
individually, brain edema in the patient was
2
associated with concomitant radiation therapy. The
3
death was potentially due to the underlying
4
condition. The acquired
hemangioma was potentially
5
related to either the underlying condition, the
6
concomitant medication or the radiation therapy.
7 The
ITP was a potentially labeled event or
8
secondary to the underlying condition.
The
9 myelodysplastic
syndrome was also a potentially
10
labeled event or secondary to the underlying
11
condition.
12
Just to give you a brief synopsis of these
13 5
cases, the first was a 3 year-old that was
14
treated for pineal blastoma who died of an
15
unspecified cause.
16
The second was a 6 year-old who was
17
treated for recurrent anaplastic astrocytoma, was
18 on
concomitant medications including radiation
19
therapy, and following temozolomide use, a
20
cavernous hemangioma was noted on MRI.
Of note, it
21 was
not previously seen on prior MRIs.
Following
22
temozolomide treatment, this patient also had
56
1 thrombocytopenia
requiring transfusions and was
2
diagnosed with ITP and myelodysplastic syndrome.
3
This patient was discharged with an improved
4
clinical status 18 days after admission.
5
The third case is a 4 year-old treated for
6
medulloblastoma who suffered an infection and there
7 was
no outcome of the event that was documented.
8
The fourth case is a 4 year-old treated
9 for
metastatic neuroblastoma who developed
10
thrombocytopenia, anemia and fever which were
11
managed with transfusions and antibiotics. She
12
recovered without sequelae and was given a second
13
cycle of temozolomide without recurrence.
14
The final case is an 8 year-old who was
15
treated for brain stem tumor.
Routine MRI revealed
16
radiation-induced cerebellum edema requiring
17
hospitalization for intracranial drainage. This
18
patient was subsequently discharged in stable
19
condition.
20
In summary, for temozolomide there have
21
been described both labeled and unlabeled adverse
22
events. The unlabeled events have
also been
57
1
reported in adults and are not unique to
2
pediatrics, and the FDA will continue to do routine
3
monitoring of adverse events in all of the
4
populations.
5
DR. CHESNEY: Thank you very
much. I just
6
wanted to bring to the committee's attention the
7
fact that at 9:30, although we are getting
8
significantly behind with the very full agenda, the
9 FDA
has asked us to address question one, which is
10 at
the back of the packet that we were given today
11
with the agenda on it, which involves process
12
issues. So, I think unless you
have specific
13
questions related to this drug, if they are process
14
issues, we will have an hour to discuss that later
15
on. So, does anybody have
specific comments
16
regarding this drug? Shirley?
17
DR. S. MURPHY: Dr. Chesney, Dr.
Starke
18
from the Pulmonary Division has some late-breaking
19
information on the first drug that we discussed.
20 He
was just going to tell us a follow-up on a
21
question that the committee had, what the bar was
22
beside the label.
58
1
DR. STARKE: I am Dr. Starke, from
2
Pulmonary and Allergy Division. I
am a medical
3
team leader. I went upstairs and
double-checked
4 the
label for you since there was a cross-out
5
there. That was simply something
that was caught
6 as
the final label was approved. The
current
7
labeling does say for 6 months and older.
8 I just want to make the comment
that even
9
though the studies were done down to 6 months of
10 age
and, as you know, certain other antihistamines
11 may
be approved down to 2 for SAR and 6 months for
12
PAR, this drug was not approved below age 6 because
13
there was no marketed formulation.
A
14
non-marketable formulation was used which, of
15
course, is an issue which you may want to address.
16
Thank you.
17
DR. CHESNEY: Thank you. If there are no
18
additional questions on your presentation, which I
19
thank you for, I think we can move on to the next
20
speaker.
21
DR. MCCUNE: It is my privilege to
22
introduce Dr. Harry Gunkel to you.
He is the only
59
1
person standing between me and the privilege of
2
saying that I am the most junior member of the
3
Pediatric Drug Development Office.
Like me, he is
4 a
neonatologist who has extensive experience in
5
private practice, the pharmaceutical industry and
6
academic medicine. Many of you
may know him for
7 his
significant work on surfactant. He is
going to
8
talk to you today about two ophthalmologic
9 anti-infective
agents.
10 Moxifloxacin and
Ciprofloxacin
11
DR. GUNKEL: Thank you,
Susie. Hello. As
12
Susie said, the next two products on the list are
13
both ophthalmic antibacterials, both
14
fluoroquinolones. The first is
ciprofloxacin,
15
known under the trade name Ciloxan and sponsored by
16
Alcon Laboratories. It is
indicated in adults and
17
children greater than 1 year of age in a solution
18
dosage form, and adults and children greater than 2
19
years of age in the ointment dosage form for the
20
treatment of bacterial conjunctivitis caused by the
21
organisms shown on the slide. The
solution form is
22
also indicated for corneal ulcer.
The original
60
1
market approval was in 1990 and pediatric
2
exclusivity was granted in January, 03.
3
Drug use data shows that dispensed
4
prescriptions for Ciloxan decreased slightly over
5 the
period of exclusivity. Almost half of
the
6
prescriptions for this drug were for children
7
between 1 and 16 years of age, and pediatricians
8
wrote about a third of the prescriptions during the
9
exclusivity period.
10
The most common indication for the
11
prescription was conjunctivitis, other or
12
unspecified, and Ciloxan was the most mentioned
13
product for this indication in pediatric patients.
14
During the exclusivity period there were 9
15
total reports for all ages; 3 were from the U.S.
16 The
age was not specified for 2 of the 9 reports.
17
There were no pediatric reports.
We will continue
18 to
monitor the adverse event reports, of course.
19
The next drug is moxifloxacin,
also
20
sponsored by Alcon Laboratories, also an ophthalmic
21
antibacterial drug. It is
indicated for adults and
22
children 1 year of age or greater for the treatment
61
1 of
bacterial conjunctivitis caused by a number of
2
susceptible organisms, aerobic gram negative and
3
gram positive organisms. The
market approval for
4
this product was April of '03, less than a year
5
ago. So, that will become
pertinent when we look
6 at
the data in just a moment. Exclusivity
was
7
granted before market approval, in January of '03.
8
Since approval didn't occur until April of
9
last year, the drug use and adverse event data
10
cover less than a 1-year period, unlike the other
11
products you are reviewing today.
About 800,000
12
prescriptions were dispensed since approval in
13
April, '03. About a quarter of
the prescriptions
14 were for pediatric patients. Ophthalmologists
15
wrote most of the prescriptions for this agent,
16
just over half of the prescriptions, followed by
17
pediatricians who wrote about a quarter of them.
18 The
most common indication, as for ciprofloxacin,
19 was
for conjunctivitis, other or unspecified and
20
Vigamox, the trade name of the product, accounted
21 for
4.6 percent of the mentions for children.
22
There was 1 report in the exclusivity
62
1
period and it was a pediatric report.
It was an
2
incidence of subconjunctival hemorrhage in a 6.5
3
year-old female that occurred 24 hours after the
4 use
of Vigamox. The child was also using
5
Augmentin. The child recovered
after
6
discontinuation of the drug and this event,
7
subconjunctival hemorrhage, is a labeled adverse
8
event occurring in 1-6 percent of patients. We
9
will continue to monitor this product as well, of
10
course.
11
One study was done for the exclusivity and
12 it
actually involved both products. It was
a
13
multicenter, randomized, double-blind, parallel
14
group comparison of moxifloxacin and ciprofloxacin
15 in
neonates, with the endpoints of clinical cure at
16 day
5 and the microbial eradication rate.
17
From the data that is available in the
18
public domain, these are the results.
The rates of
19
clinical cure are shown for both the agents. These
20
rates are less than the generally expected vehicle
21
rate, and the difference between the two was not
22
significant. Thank you.
63
1 DR. CHESNEY: I have two questions. What
2 do
you mean by expected vehicle rate?
3
DR. GUNKEL: If you apply a
vehicle to a
4
case of bacterial conjunctivitis the expected cure
5
rate is 70 percent.
6
DR. CHESNEY: That is what I
thought you
7
meant; I just wanted to be sure.
And, what were
8 the
side effects of Ciloxan? There were 9
reports.
9
DR. GUNKEL: They weren't
pediatric so I
10
didn't see them. I don't know.
11
DR. CHESNEY: Other
questions? Dr.
12
Murphy?
13
DR. D. MURPHY: Go ahead and
finish up
14
with this topic because I was asked to make a
15
clarification on the last one.
16
DR. CHESNEY: Dr. O'Fallon?
17
DR. O'FALLON: If I were the
statistician
18 on
this study I would be very concerned. I
would
19 be
talking to the docs and saying, "wait a minute
20
guys, this looks like it's doing harm." Both of
21
these agents look like they are not helping. If
22
they have a lower response rate or success rate,
64
1
whatever you want to call it, than the placebo
2
which is the vehicle without anything in it I would
3 be
worried that it is contra-effective.
4
DR. GUNKEL: I don't know whether
that is
5 the
case. The information that is in the
public
6
domain doesn't allow us to deduce that the rates
7
that were shown in the study that I showed were
8
significantly less than the expected vehicle cure
9
rate. But your point is well
taken I would think.
10
DR. CHESNEY: Dr. Murphy?
11
DR. D. MURPHY: Dr. McCune has
said that I
12 may
have confused things in efforts to answer Dr.
13
Santana's question about how we get information in
14 the
label because you were talking about the
15
topotecan when I read to you the information that
16 was
in the Temodar label. I was trying to
point
17 out
that there are various approaches depending on
18 the
quality of the data. So, for the
topotecan the
19
actual information that is in the label now in
20
pediatrics is that there is no safety or
21
effectiveness that has been established versus the
22
Temodar, which is the one that I read you. I
65
1
thought I read the product but they both start with
2
T. So, I want to make it clear
that it is the
3
Temodar that has all that information in it.
4
DR. SANTANA: My question was a
process
5
issue; it didn't relate to any specific--
6
DR. D. MURPHY: Yes, and I was
trying to
7
give a process where there can be different types
8 of
information put in. Anyhow, I just
wanted to
9
make sure that I didn't confuse the committee with
10 the
Ts when I started talking about the second
11
label before it was actually presented.
Thank you.
12
DR. CHESNEY: I think we are all
looking
13 at
your last two slides and puzzling over the last
14
one, but I think that wasn't really the issue of
15
this morning's discussion so we will leave that for
16 the
moment and move on to the next speaker.
17
DR. GUNKEL: The next speaker is
Dr. Larry
18
Grylack. Dr. Grylack began a
career in the
19
Commission for U.S. Public Health Service from
20
1971-73. His training is in
pediatrics in
21
neonatal/perinatal medicine. He
was in the
22
practice of neonatal medicine at Columbia Hospital
66
1 for
Women, in Washington, for 26 years with a
2
particular interest in neurodevelopmental follow-up
3 of
high risk newborn and apnea during infancy.
Dr.
4
Grylack?
5 Fosinopril
6
DR. GRYLACK: Thank you, Dr.
Gunkel, for
7 the
introduction. It is a privilege to speak
to
8 the
committee this morning. In case there
has been
9
anything said so far this morning that has caused
10
your blood pressure to rise, I will be discussing
11 an
antihypertensive drug at this time.
12
The name of the drug is fosinopril, with
13 the
trade name of Monopril. Its sponsor is
14
Bristol-Myers Squibb. Fosinopril
is in the renin
15
angiotensin antagonist subclass of
16
antihypertensives. Its mechanism
of action is
17
inhibition of angiotensin converting enzyme.
18
Although fosinopril is approved for use in adults,
19
there are no approved pediatric indications.
20
Pediatric exclusivity was granted early last year.
21
Despite a 20 percent increase in the
22
prescribed use of renin angiotensin antagonist
67
1
drugs in the outpatient setting, there was a 25
2
percent decrease in the use of fosinopril during a
3
recent 3-year period. Conversely,
there was a 33
4
percent increase in the use of the combination drug
5
fosinopril/hydrochlorothiazide during that same
6
time period. The ratio of the
number of pediatric
7
prescriptions for fosinopril alone to prescriptions
8 for
the combination drug was approximately 10:1.
9
Let's focus on the inpatient usage data
10 for
fosinopril. Two databases from recent
3-year
11
periods report a very low percentage of pediatric
12
inpatients using fosinopril during their hospital
13
stays. There were no pediatric
adverse event
14
reports submitted during the post-exclusivity
15
period.
16
Two studies were done for the purpose of
17
achieving exclusivity. A
single-dose
18
pharmacokinetic study showed an age-dependent
19
increase in bioavailability in a population of 43
20
patients between the ages of 1 month and 16 years.
21 An
oral solution containing a dose of 0.3 mg/kg of
22
body weight was used.
68
1
Secondly, an efficacy and safety dose did
2 not
demonstrate a dose-response relationship in a
3
population of 253 patients between 6 and 16 years
4 of
age. A tablet form of medication was
used in
5
this study. No deaths or cases of
angioedema were
6
reported, the latter being an adverse event
7
reported in adults.
8
Pharmacokinetic parameters in the children
9
studied are similar to those found in adults.
10
Dosing information is available for children
11
weighing more than 50 kg.
However, the
12
formulations used in children in the exclusivity
13
studies are not currently commercially available.
14
This leads me to the broader
issue of the
15
need for age-appropriate formulations.
As
16
physicians and parents know, non-liquid forms of
17
medications are not appropriate for infants and
18
preschool children, as for some school age children
19 as
well. Therefore, sponsors are being
encouraged
20 to
develop age-appropriate commercially available,
21
marketable pediatric formulations during their
22
exclusivity studies.
69
1
The goal of the FDA, and especially of our
2
Pediatric Drug Development Division, is to have
3
commercially available formulations for the
4
pediatric patient population. If
this cannot be
5
done for certain drugs in a pharmacy--and I
6
underscore pharmacy--compounded recipes should
7
appear in the drug label.
8
This concludes my remarks for today.
9
Thank you for your attention.
10
DR. CHESNEY: Thank you very
much. Any
11
non-process questions for the speaker?
Dr. Hudak?
12
DR. HUDAK: The slide that showed
that
13
there was no dose-response relationship in
14
children, is that sort of a euphemism for no
15
efficacy?
16 DR. D. MURPHY: Yes.
It is in our written
17
request as one way for the cardiorenal drugs,
18
hypertensive drugs, to demonstrate efficacy and it
19 is
a long description about what you have to do if
20 you
don't choose a placebo-controlled trial and you
21
choose a dose effect trial and what sort of effect
22 you
have to demonstrate and, if you don't, then you
70
1
failed.
2
DR. CHESNEY: Dr. Nelson?
3
DR. NELSON: With your indulgence,
it is a
4
process comment but it is not about risk process.
5 We
have heard two presentations where there has
6
been a lack of an adequate formulation.
I guess my
7 question,
which may not be answerable today or we
8 may
not want to answer it today is that my
9
understanding is a company doesn't get exclusivity
10
unless the FDA determines--or doesn't get a
11
request--that there is a significant health
12
benefit. It is unclear to me how
you can decide
13
that there is a significant health benefit to the
14
population when at the end of the day there is no
15
formulation available for them.
16
DR. D. MURPHY: Again, they have
to fairly
17
meet the terms of the written request.
A written
18
request is based on what the public health benefit
19
would be and it often will say that you must
20
conduct this trial with an age-appropriate
21
formulation. If they conduct the
trial with the
22
age-appropriate formulation it does not say, nor do
71
1 I
think we would be allowed to legally say, you
2
must market it.
3
DR. NELSON: Well, I guess I would
go back
4 to
the attorneys and ask them to reflect on that
5
because--
6
DR. D. MURPHY: We have.
7
DR. NELSON: --I guess I don't
think that
8 was
the intent of Congress, that they would get the
9
money and then have nothing available for that
10
population.
11
DR. D. MURPHY: Yes, we have gone
back
12
actually because, as you can see, this is becoming
13 an
issue. We have brought this back to them
and we
14 are
in the process of discussing again, within our
15
legal regulatory authority, what we can and cannot
16 do.
17
In balancing that, the other effect, the
18
unintended effect is that you don't issue any
19 written
request because they aren't going to do
20
them, or you can issue them and they won't do them
21 at
all. So, is there a way we can balance
the kind
22 of
information that we need--and I really can't
72
1
give a final answer on this right now--is there a
2 way
that we can set it up so that we say you need
3 to
develop a marketable formulation that would be
4
appropriate for children? We have
always had
5
criteria that if you can't do that you have to tell
6 us
why but make that clear, more definitive.
7
Then, if you can't--because there are
8
reasons sometimes why you cannot develop certain
9
formulations--the solvents become too large or
10
other reasons, as you all I think know, with some
11 of
the proton pump inhibitor types of
12
products--then we are looking at trying to define
13
requirements that have to be met having to do with
14
stability, bioavailability, for kids' use that
15
would be appropriate. We get into
other issues for
16
compounding and how do you avoid those issues.
17
So, the bottom line, Dr. Nelson, is that
18 we
are very aware that this is an issue and we are
19
trying to find a resolution that promotes
20
development of products while, at the same time,
21
does not end up in the situation where we have
22
products that are then not available.
73
1
DR. NELSON: I appreciate the
2
complexities. In my simplistic
view, I suspect
3
that if you went back to those that drafted and
4
then passed the Best Pharmaceuticals for Children
5
Act, they would not interpret significant health
6
benefit to mean that at the end of the day there is
7 no
formulation and nothing in the label.
8
DR. CHESNEY: Dr. Hudak?
9
DR. HUDAK: Can I just clarify
this
10
because I am trying to understand exactly what the
11
data show. The formulations used
in children less
12
than 50 kg were not commercially available?
13
DR. GRYLACK: That is correct.
14
DR. HUDAK: These are the same
15 formulations used that assessed the PK issues?
16
DR. GRYLACK: Yes, the initial
singe-dose
17 PK
study was done in patients between the ages of 1
18
month and 16 years so, as you can determine, a
19
number of those were less than 50 kg.
Then, the
20
second study, the efficacy and safety study, was
21
done in patients between 6 and 16 years and, again,
22 a
certain number of those would be less than 50 kg.
74
1
DR. HUDAK: So, essentially, the
drug with
2
this non-available preparation showed that, as
3
given, it was absorbed and available in the
4
bloodstream like in adults, but showed no efficacy.
5
DR. GRYLACK: Well, there was the
6
age-dependent increase in bioavailability.
7
DR. HUDAK: I understand, but
giving
8
adequate levels of the drug, there was no
9
level-related efficacy, no dose response--
10
DR. GRYLACK: No dose response.
11
DR. HUDAK: No dose response but
if you
12
control for the level of the drug in the blood
13
there was still no response. See
what I am saying?
14
There may be a difference depending upon the age.
15 DR. GRYLACK: Yes.
16
DR. HUDAK: So, the bottom line is
that
17
this drug did not work with the best possible
18
formulation in this population and, therefore,
19
there doesn't seem to be any reason to have a
20 formulation
available for pediatric patients. Is
21
that correct? For this drug?
22
DR. D. MURPHY: Correct.
75
1
DR. CHESNEY: Dr. Danford?
2
DR. DANFORD: To Dr. Hudak's point, I
3
wonder if the group in which this drug was studied
4
actually had hypertension or not.
Hypertensive
5
children, the younger you get, are harder and
6
harder to come by and if you were just studying the
7
bioavailability of the drug and giving it to
8
volunteer children you would not necessarily expect
9 a
drop in blood pressure in a pediatric population.
10 Do
you know who these children were?
11
DR. GRYLACK: I would have to take
a
12
minute and go back and look at the detailed
13
description of the studies. I am
sorry, I can't
14
answer that off the top of my head.
Perhaps I can
15 get
back to you a little later.
16
DR. D. MURPHY: Was the question
did we
17
give it to normal children?
18
DR. DANFORD: Or children without
19
hypertension.
20
DR. D. MURPHY: That is what I
meant,
21
children without hypertension.
22
DR. DANFORD: There could be a
group that
76
1
might conceivably benefit from this, who have
2
congestive heart failure who would not have
3
elevated blood pressure. If you
were looking at a
4
response in blood pressure and it were given to a
5
group of patients with VSD you might not be able to
6
determine much of a change in their blood pressure.
7
DR. D. MURPHY: I think the first
part of
8 it
is that we would not have done the studies in
9
children who were not hypertensive.
Now, could we
10
have selected a different population so that
11
potentially mechanistically you could postulate a
12
benefit? You possibly could have
but it was felt
13
that the need was in this population so that is why
14 it
was written for this population. Again,
as this
15
committee has discussed, it would have to be
16
children who had the disease under study.
17
DR. HUDAK: I am happy to hear
that
18
because testing this antihypertensive medication in
19
normotensive children I think would be a real--
20
DR. GRYLACK: I have some comment
here.
21
Thank you for waiting for me. The
patient
22
population in the efficacy and safety study
77
1
consisted of patients with hypertension or high
2
normal blood pressure.
3
DR. CHESNEY: Dr. Nelson, one more
4 question
and then we really need to move along.
5
DR. NELSON: It just occurs to me
that
6
that question is answerable if you have the
7
pharmaceutical review that is on the website. So,
8
maybe in the future just including that as part of
9 the
packet would enable us to have that at hand.
I
10 am
looking to see if that one is in here.
11
DR. PEREZ: Use the mike, please.
12
DR. D. MURPHY: It is in here in
what is
13
called the critical pharmacology and
14
biopharmaceutics review; summary of findings--
15
DR. S. MURPHY: Just to remind you
that we
16 can
only put what is in the public domain so, as we
17
look at what is being posted on the web I think
18
some of these are more extensive than others. So,
19 it
is giving us an opportunity to see what is going
20 on.
21
DR. GRYLACK: The PK study was
done on all
22
hypertensive patients. Are we
going to take a
78
1
break now for the vote or are we going to pursue to
2 the
next one?
3
DR. CHESNEY: Assuming there are
no more
4
questions on this particular issue, Dr. Santana
5
will cover the next drug as I am recused for stock
6
reasons. So, Dr. Santana?
7
DR. SANTANA: Let's go ahead and
get
8
started. Dr. Buckman?
9
DR. GRYLACK: Yes, it is my
pleasure to
10
introduce Dr. ShaAvhree Buckman. Dr.
Buckman is a
11
pediatrician who is not a neonatologist, who also
12 has
a Ph.D. in molecular cell biology and
13
pharmacology. Dr. Buckman has
been a medical
14
officer with the Division of Pediatric Drug
15
Development for nearly two years, and I will add
16
that Dr. Buckman has been a valued colleague of
17
mine during the time I have been here at the FDA.
18 Fentanyl
19
DR. BUCKMAN: Good morning. I will be
20
discussing the one-year post-exclusivity adverse
21
events for the fentanyl transdermal system.
22
The fentanyl transdermal system or,
79
1
trademark Duragesic, is marketed by Johnson &
2
Johnson and its subsidiary ALZA.
It is indicated
3 for
the treatment of chronic pain such as that of
4
malignancy that cannot be managed by lesser means,
5
such as acetaminophen-opioid combinations,
6
non-steroidal anti-inflammatory drugs or PRN dosing
7
with short-acting opioids, and pain that requires
8
continuous opioid administration.
It is approved
9 for
pediatric use in children down to the age of 2
10
years. The drug obtained original
market approval
11 in
August of 1990 and pediatric exclusivity was
12
granted in January of 2003.
13
The Duragesic label carries a boxed
14
warning that specifically states that due to the
15
possibility of serious or life-threatening
16
hypoventilation Duragesic is contraindicated in the
17
management of acute or postoperative pain,
18
including use in outpatient surgeries.
It is also
19
contraindicated in the management of mild or
20
intermittent pain responsive to PRN or non-opioid
21
therapy. It is also
contraindicated in doses
22
exceeding 25 mcg/hour at the initiation of opioid
80
1
therapy.
2
I have also outlined in red the pediatric
3
safety information that is in the boxed warning,
4
which specifically states that the safety of
5
Duragesic has not been established in children
6
under 2 years of age. Duragesic
should be
7
administered only if they are opioid-tolerant at
8 age
2 years or older.
9
There is selected additional safety
10
labeling which states that Duragesic should be
11
prescribed only by persons knowledgeable in the
12
continuous administration of potent opioids and the
13
management of patients receiving potent opioids for
14
treatment of pain and in the detection and
15
management of hypoventilation, including the use of
16
opioid antagonists.
17
Now, the total number of prescriptions
18
dispensed for the fentanyl transdermal systems in
19 the
United States have increased by 20 percent in
20 the
past 2 years, from 4.5 million in 2002 to 5.4
21
million in 2003. The top
prescribers in 2003 for
22 the
fentanyl transdermal systems were internal
81
1
medicine, family practice and anesthesiology.
2
Approximately 0.2 percent of fentanyl transdermal
3
system prescriptions dispensed were written by
4
pediatricians.
5
In the outpatient setting children and
6
adolescents have accounted for very few dispensed
7
fentanyl transdermal system prescriptions over the
8
past 2 years, 4,535 prescriptions from February
9
2002 to January of 2003 to 5,422 prescriptions from
10
February, 2003 to January, 2004.
In both the
11
outpatient and inpatient settings, adolescents age
12
12-16 years accounted for 60 percent of the
13
pediatric fentanyl transdermal system use over the
14
past 3 years.
15
In the outpatient setting the most
16
frequent diagnoses associated with the fentanyl
17
transdermal systems in the pediatric, as well as
18 the
adult, population were associated with diseases
19 of
the musculoskeletal system and connective
20
tissues. In the pediatric
population the most
21
predominant musculoskeletal diagnosis was spinal
22
stenosis, followed by injuries involving fractured
82
1
bones. One must be mindful though
that these are
2
very small numbers that we are capturing.
3
In the inpatient setting the primary
4
discharge diagnoses most frequently associated with
5
billing during hospitalization in the pediatric
6
population were for cholesterol encounters and
7
various blood disorders, including sickle cell
8
disease.
9
There was a total of 1,917 adult and
10
pediatric adverse event reports for the fentanyl
11
transdermal system during the 1-year
12
post-exclusivity period. Of
these, there were 8
13
unique pediatric cases. Seven
were from the U.S.
14 and
1 was a foreign report. All of these
cases
15
were described as serious outcomes, including 5
16
deaths. There were 4 reports in
females and 4
17
reports in males, and the ages ranged from 4-16
18
years of age. Of these 8
pediatric reports, most
19
adverse events were mentioned only once.
The
20
labeled adverse events that were captured twice
21
included overdose drug abuser and medication error.
22
Again, these are labeled adverse events.
83
1
Of the unlabeled adverse events that were
2
captured more than once, they included cardiac
3
arrest, respiratory arrest and self-medication.
4
There were 5 deaths that were reported
5
during the 1-year post-exclusivity period for
6
Duragesic and I would like to describe these
7
reports to you. The first was the
case of an 8
8
year-old female who was diagnosed with
9
rhabdomyosarcoma who died 2 months after being
10
switched from the fentanyl transdermal system to IV
11
morphine. This was a foreign case
and it is
12
believed that this child's death was due to
13
progression of her underlying disease and not due
14 to
the patch itself.
15
The second case is that of a 9 year-old
16
male who was 2 days post tonsillectomy and
17
adenoidectomy, who was treated with the fentanyl
18
transdermal system 25 mcg patch with subsequent
19
respiratory arrest resulting in death.
Concomitant
20 medications
that were given included acetaminophen
21
with codeine elixir, although the timing of
22
administration of this dosing is unclear from the
84
1
report.
2
This was a U.S. case and I have
a couple
3 of
comments about this case. One is that
this is a
4
case where a non-opioid tolerant patient was
5
prescribed the drug for an acute postoperative pain
6
situation. As you recall from the
boxed warning,
7
Duragesic is contraindicated in the management of
8
acute or postoperative pain.
9
The next case was that of a 4 year-old
10
female who died from cardiac arrest after having
11 the
fentanyl transdermal system applied by her
12
grandmother for pain relief. The
details of this
13
case are largely unknown. This is
a U.S. case, and
14 the
only additional information that we have is
15
that the child had marks on her body that indicated
16 that
she may have had more than one patch applied
17
because there was adhesive residue on her skin.
18
The next case was that of a 16 year-old
19
male with a history of drug abuse, including
20
marijuana, methylphenidate and dextropropoxyphene,
21 who
was reported to have been using the fentanyl
22
transdermal system several days prior to death and
85
1 was
found wearing a 100 mcg patch. This was
a U.S.
2
case.
3
The last of the 5 reported deaths was that
4 of
a 16 year-old male, with a history of alcohol
5 and
marijuana use, who died of cardiac arrest after
6
using 100 mcg patches obtained from another
7
student. He was found wearing a
100 mcg/hour patch
8 and
this was a U.S. case.
9
Now, there were 3 non-fatal adverse events
10
that were reported during the 1-year
11
post-exclusivity period. These
included a patient
12 who
experienced euphoria, hallucinations and weight
13
loss after initiation of therapy with the fentanyl
14
transdermal system.
15
The second case was a child who
16
experienced withdrawal symptoms from what was
17
considered a loose patch, meaning that the patch
18 had
become non-adherent to the skin and the patient
19
experienced withdrawal symptoms which resolved
20
after replacement of a new patch.
21
The last case was that of respiratory
22
depression in a patient who had intentional misuse
86
1 of
the fentanyl patch.
2
We have reported the adverse events that
3
occurred during the 1-year post-exclusivity period.
4 Due
to our concern regarding the pediatric deaths
5
occurring with this product, we decided to
6
investigate the adverse events which occurred since
7 the
approval of Duragesic for adults, in 1990.
8
There were 4 pediatric deaths before initiation of
9 the
pediatric exclusivity period. There have
been
10 3
additional pediatric deaths since the end of the
11
1-year post-exclusivity period.
Although we are
12
continuing to monitor for adverse events, for the
13 purpose of this presentation we set our
internal
14
cut-off for reporting to you at May 15th.
15
Now I would like to describe briefly those
16
deaths that occurred outside of the exclusivity
17
reporting period. The first was a
case of
18
accidental exposure. The second
was a case of
19
misuse or abuse. Most concerning
are these cases
20 of
off-label use. One is a case of a child
with
21
post-tonsillectomy and adenoidectomy pain. Another
22 is
a case of a child with infectious mononucleosis
87
1 and
sore throat pain; a child with chronic
2
headaches and infectious mononucleosis; and a child
3
with acute migraine. The last case
that was
4
reported was that of a child with rhabdomyosarcoma
5
and, again, this was another situation where it was
6
thought that the child died due to disease
7
progression and not due to administration of the
8
patch itself.
9
In summary, the cumulative pediatric
10
adverse events for the fentanyl transdermal system
11
since original market approval in 1990 totaled 35
12
unique cases. Of these, 22
reports were for
13
children who used the product appropriately for an
14
indication of chronic pain. Of
these 22 reports,
15
there were 2 pediatric deaths and these were both
16
children with rhabdomyosarcoma which I described.
17
By comparison, there were 13 reports in
18 children using the medication for a
non-chronic
19
pain management indication. Of
these 13 reports,
20
there were 10 pediatric deaths.
It is important to
21
remember that the Adverse Event Reporting System is
22 a
voluntary reporting system which is subject to
88
1
under-reporting and other influences, which you
2
have heard described multiple times this morning.
3
In conclusion, several of the serious
4
pediatric adverse events captured occurred in
5
patients who administered the product for an
6
unlabeled indication, for example, treatment of
7
acute pain in a non-opioid tolerant patient. There
8 is
need for additional education regarding the
9
proper use of the fentanyl transdermal system to
10
help further minimize abuse, misuse and off-label
11
use.
12
In conclusion, instead of answering
13
questions right now, because we have two subsequent
14
presentations that deal with the same product, I
15
would like to introduce the next speaker and then
16 we
can take questions at the end of all three
17
presentations. So, Dr. Lee will
address the
18
fentanyl pharmacokinetic characteristics following
19
Duragesic application. Dr. Lee is
a clinical
20
pharmacology and biopharmaceutics reviewer with the
21
Office of Clinical Pharmacology and
22
Biopharmaceutics, currently working with the
89
1
Division of Anesthetic, Critical Care and Addiction
2
Drug Products. Dr. Lee?
3
DR. LEE: Thank you, Dr.
Buckman. Good
4
morning, ladies and gentlemen. I
would like to
5 present
to you this morning on unique features of
6
fentanyl pharmacokinetics after Duragesic patch
7
application, but first, before I go into my slides,
8 I
would like to give you some overall background
9
information on the Duragesic patch.
10
First on the patch strengths, Duragesic
11
patches are available as 25 mcg, 50 mcg, 75 mcg and
12 100
mcg fentanyl delivered per hour patches.
13
Secondly on the site of application, patches are
14
applied mostly on a flat skin surface, mostly on
15 the
upper torso, such as chest, back, flank or
16
upper arms. In young children,
however, the upper
17
back is a preferred location to minimize the
18
potential for the child to remove the patch.
19 Lastly
on the intended use, as we all know, each
20
patch can be worn continuously up to 72 hours but,
21 if
analgesia for more than 72 hours is required, a
22 new
patch should be applied to a different skin
90
1
site after removal of the previous patch.
2
Gollowing the patch application the
3
fentanyl drug molecules move from the patch
4
reservoir through a rate-controlling membrane and
5
continue to be absorbed into the skin.
At this
6
juncture a depot of fentanyl concentrates in the
7
upper skin layer and fentanyl then becomes
8
available to the systemic circulation.
Peak serum
9
concentrations of fentanyl generally occur between
10 24
and 72 hours.
11
However, after patch removal the serum
12
fentanyl concentrations decline slowly, falling
13
about 50 percent in approximately 17 hours, which
14 is
the elimination half-life of the fentanyl patch
15
drug delivery system. Due to the
continued
16
absorption of fentanyl from the skin because of the
17
skin depot effect, fentanyl disappearance from the
18
serum is slower than is seen after an IV infusion.
19 The
elimination half-life for the IV infusion route
20 is
approximately 7 hours compared to that of 17
21
hours.
22
So, what are some of the potential
91
1
implications? With respect to
initial patch
2
application, the full drug benefit, analgesic
3
effect, may not be seen immediately.
Thus, there
4 is
a potential situation for applying another
5
patch. This can become a safety
issue I think.
6
With respect to post-patch removal,
7
substantial drug effect may be felt for a
8
significant period of time. Thus,
there is a
9
potential safety situation for a patient who will
10 be
switching over to another opioid therapy.
11
If you have any questions, I
will be happy
12 to
answer any, otherwise I will introduce Dr.
13
McNeil. Dr. McNeil is a medical
reviewer with
14
HDF-170. Prior to coming to the
agency she trained
15 in
pediatric neurology and oncology. Dr.
McNeil?
16
DR. MCNEIL: Good morning. I am with the
17
Division of Anesthetic, Critical Care and Addiction
18
Drug Products and, in collaboration with our
19
pediatric colleagues, we have been considering ways
20 to
manage the risk of off-label use.
21
We have been coming up with preliminary
22
strategies for managing this risk, and the
92
1
preliminary strategies that we have come up with
2 are
labeling changes; prescriber education through
3 the
company or, one thing that has been used in the
4
past, are "dear healthcare professional" letters,
5 or
prescriber education through physician
groups.
6 We
will, of course, be in contact with the company
7 and
with our colleagues in pediatrics as we try to
8
come up with a method of managing this risk.
9
DR. SANTANA: Did you have further
10
comments, Dr. Buckman?
11
DR. BUCKMAN: We can go ahead and
12
entertain questions at this time.
13 Discussion of Question 1
14
DR. SANTANA: Good. I do have a question
15 for
Dr. Lee. Is there any data either in
16
pediatrics or in adults that other concomitant
17
problems, like fever or skin rashes, change the
18
absorption? I was struck by a
couple of the deaths
19 in
patients who had infectious diseases or had
20
postoperative conditions that could be associated
21 with fever or some of these associated skin
rashes.
22 So,
is there any data to suggest that there is a
93
1
different pharmacokinetic profile under those
2
circumstances?
3 DR. LEE: As far as I know, I don't think
4 we
have any information from the pediatrics which
5
were involved with PK studies.
However, Dr. McNeil
6
may--she says no.
7
DR. SANTANA: Do we know from the
adults
8
about postoperative fever and things of that
9
nature?
10
DR. MCNEIL: No, we don't. It is actually
11 in
the label that if you apply heat externally to
12 the
drug patches you can increase the serum
13
concentration of fentanyl, but that is what is
14
known about it.
15
DR. SANTANA: Dr. O'Fallon?
16
DR. O'FALLON: I have been
watching these
17
things because my 93 year-old mother-in-law has
18
been outcome this--now she is 96 and a half--for
19
three and a half years. When I
first looked at it
20
what bothered me was the slow--she is allergic to
21
lots of different things; as it turns out she is
22
fine with this, but with something that moves so
94
1
slowly what would happen? It
would seem to me that
2
after 24, 36, 48 hours, something like that, a
3
person might reach a level where they would not be
4
able to tolerate it. Then, there
is this 17-hour,
5
which is really up to a whole day--before you can
6
drop the levels down sufficiently.
What do you do
7 if
somebody--I don't see anything in the label
8
about how to manage somebody that has a bad effect.
9 How
do you do it when it is in your system for so
10
long?
11
DR. LEE: My first answer could
have been
12
that the person who is experiencing adverse events
13 may
just peel off the patch and then for 17 hours,
14 for
that I don't have any answers.
15
DR. O'FALLON: It is actually up
to 24
16
almost. I mean, there is a
terrific range on these
17
things.
18
DR. LEE: Yes, the range is very
large.
19
Yes.
20
DR. S. MURPHY: Dr. Lee, could you
show
21
your backup slides with the kinetics?
I think they
22 are
very helpful.
95
1
DR. LEE: I would just like to
remind you
2
that the information in this study is from a
3
limited number of patients and the pediatric
4
subjects were non-opioid tolerant subjects. This
5
study had full pharmacokinetic profiling and,
6
therefore, it was a very useful study for me.
7
The Y axis is in nanograms per milliliter
8
concentration versus time. We put
a patch on and
9
take it off at 72 hours. This is
the adult
10
population where we see the increase in the
11
fentanyl concentration at approximately 22 to about
12
40-some odd hours.
13
Compared to the adults, this is a
14
pediatric population and I would just like to
15
mention at this time that the patch strength size
16 was
50 mcg/hour for adults and 25 mcg/hour for the
17
non-opioid tolerant pediatric patients.
As you can
18
see, time to maximum concentration has shifted at
19
earlier time points and it is higher.
Where I have
20
marked it with the shaded ovals, that is where we
21
need to kind of think again as far as having the
22
pain relief because it takes so long in order to
96
1
reach that plasma concentration.
And, then for the
2
black square, because it takes so long in order to
3
have the fentanyl concentration either eliminated
4
from the system or what-have-you, it takes so long,
5
even up to maybe 140 hours you could have some of
6 the
residual fentanyl concentration after patch
7
removal. So, I guess this is what
we have.
8
DR. MCNEIL: Excuse me, I should
mention
9
that my answer on fever was related to the
10
information we have, actual data from patients, but
11 in
the label, by PK modeling, there has been some
12
association that fentanyl doses could theoretically
13
increase up to a third but, again, that is from PK
14
modeling and not from actual patients.
15
DR. SANTANA: And we have no
postmortem
16
information from any of these deaths regarding
17
measurement of drugs in these patients?
18
DR. BUCKMAN: In looking at a
couple of
19
MedWatch reports we do have a couple of cases where
20 we
did get levels. In one case that I
reported to
21
you, the 16 year-old that died from an overdose of
22 the
fentanyl patch had an autopsy that was
97
1
performed. The cause of death was
cardiac arrest
2 due
to highly toxic levels of fentanyl. The
3
fentanyl level was 16 ng/ml. He
also had
4
cannabinoids in his bloodstream as well.
So, that
5 was
one case where we did actually have a
6
concentration.
7
DR. SANTANA: Dr. Fuchs?
8 DR. FUCHS: Well, two things that strike
9 me
from reading all your cases are that three of
10
them were used in kids with tonsillectomy or mono,
11 and
if you have ever looked at kids with
12
mononucleosis, those are kissing tonsils and, yes,
13
they do hurt. That may be
something where we might
14 add
a warning, "do not use when there is any airway
15
problem or tonsillitis or mono" because that is an
16
airway issue to begin with and then if you have
17 respiratory depression, which this drug is
known to
18
cause, and you have no airway obviously you will
19 get
hypoxia and that will then lead to respiratory
20
arrest and then lead to cardiac arrest.
21
The second thing is that in the cases
22
where you mentioned cardiac arrest, I suspect
98
1
mostly in kids this is respiratory arrest. Once
2
again, we can't really tell from the reports but I
3 suspect they are all respiratory related.
4
DR. BUCKMAN: That is a very good
point.
5 We
can only report to you exactly what is captured
6
there but I agree with you that in most pediatric
7
cases it is respiratory arrest leading to cardiac
8
arrest. But that is how it was
captured in the
9
reports.
10
DR. SANTANA: Dr. Nelson?
11
DR. NELSON: Looking at the label,
my
12
understanding is the strongest warning that the FDA
13 can
do is the black box, which is what you have at
14 the
front. So, unless we think it needs to
be
15
worded differently, there is already a black box.
16 The
only thing that doesn't seem to be in that
17
black box that is elsewhere is the comment about
18 the
qualifications of who should prescribe this.
19
Working in critical care and having used this in
20 the
past and no longer using it in the way that I
21 had
used it simply because of the labeling change,
22 it
is unclear to me how much stronger you could
99
1
make this, other than perhaps moving the
2
information about prescribers to the black box.
3
And, it is unclear to me--I am assuming there was a
4
pretty good malpractice loss for these deaths, or
5
there should be--so it is unclear to me how much
6
more you can do in your labeling if, in fact,
7
people are going to use it when it says not to use
8 it
that way. It seems pretty
straightforward to
9 me.
10
DR. BUCKMAN: Can I respond to
that
11
briefly? That was why in the
presentation we
12
wanted at the outset to show you exactly what was
13 in
the labeling because that is what we need to
14
hear from you as far as comments as far as how can
15 we
get that word out there anymore so. It
seems as
16 if
the greater propensity of what is happening is
17
that patients are being administered this product
18 for
an unlabeled or contraindicated use. So,
that
19 is
the feedback that we want to get from you all.
20 You
know, what other things can we do? That
is
21
going to be the question that we will be asking.
22
DR. SANTANA: Dr. O'Fallon first
and then
100
1 Dr.
Hudak.
2
DR. O'FALLON: I think that the
letter to
3
patients is very helpful that is included in our
4
packet. Is that normally given to
the patients? I
5
don't know your process here.
Between the
6
executive summary and the actual label there is a
7
patient information thing.
8
DR. MCNEIL: Patient information?
9
DR. O'FALLON: Yes, and I don't
know where
10
that comes into the Act.
11
DR. MCNEIL: In theory, what
happens is
12
when you buy your box of Duragesic is that you
13
should get this.
14
DR. O'FALLON: Yes, I don't think
we did.
15 She
said theoretically we should get it when we buy
16 our
box but I don't remember that we did.
17
DR. D. MURPHY: Remember, it is
not
18
required to be given to every patient.
19
DR. O'FALLON: That is what I was
20 wondering.
21 The
other thing is that I don't think it addresses
22 the
issue. You see, I was worried the first
time I
101
1 saw
this about the long-lastingness of this drug.
2
What happens if they get into trouble?
Is there
3 any
information that the patients could have if
4
they are seeing something? I
don't even see that
5 it
says call your emergency room immediately, or
6
something. I don't see anything
about what to do
7 in
case the kid gets in trouble or the person gets
8 in
trouble, the 90-odd year-old gets in trouble.
9
DR. MCNEIL: Under "how do I
use
10
Duragesic" in the patient information section it
11
does say if you use too much Duragesic or overdose
12 get
emergency medical help right away. But I
guess
13
from what you are saying that is not enough.
14
DR. O'FALLON: Well, I didn't see
it in
15 the
patient letter but maybe I missed it.
16
DR. SANTANA: Dr. Hudak?
17
DR. HUDAK: I guess in comment to
Dr.
18
Nelson's comment, I am not sure what can be done
19 for
language and what the limitations are but I
20
think many physicians are sort of jaded when they
21 see
"serious" or "life-threatening" written down
22
somewhere because that seems to be on a lot of
102
1
different drugs, and maybe something very specific
2 about deaths have occurred due to, you know,
3
inappropriate use in these situations should be in
4
there in some form that makes it very concrete.
5
DR. SANTANA: Do we know from
these
6
adverse event reports if, in the cases that were
7
postoperative, those were actually prescribed by a
8
person before the procedure or subsequently by a
9
pediatrician or family physician?
I mean, what is
10 the
sequence of prescriptions here?
11
DR. BUCKMAN: In one case it was
12
prescribed by a family physician.
In another case
13 it
was prescribed by the pain control team in the
14
ICU, and the mother had asked--and Joe Wyeth, our
15 ODS
person, please correct me if I am wrong; she
16 has done an incredible job of helping us get
all
17
these reports together--but in another case it was
18
prescribed in the ICU by the pain control team for
19
this child. The mother asked that
two of the vital
20
checks be suspended. They were
overnight vital
21
checks. She wanted the child to
rest, and by the
22
morning when they did the next vital check the
103
1
child was dead.
2
DR. SANTANA: Dr. Gorman, I would
like to
3
hear your opinion on this since you are a
4
practicing pediatrician in the community.
5
DR. GORMAN: First of all, all of
these
6
patches as they have come out, these long-acting
7
patches--I think I remember the same event with a
8
patch that came out for hypertension with another
9
product where children had it applied
10
inappropriately or retrieved it from wastepaper
11
baskets. There were several
adverse outcomes which
12
were slightly different than these.
13
I would have to echo Dr. Hudak's very
14
explicit comments. I think the
hypothetical that
15 is
put in the black box warning now is a reality
16 and
there should be a statement--and I understand
17
that labels are a negotiated legal document between
18 the
FDA and the pharmaceutical company, but a
19
simple statement that deaths have occurred through
20 the
inappropriate use of this in the following
21
settings, and then a listing that you have
22
contraindicated would take this out of the realm of
104
1
hypothetical and say it is real.
2
Then to echo a little bit of what Dr.
3
Nelson said, there could be a little asterisk on
4 the
bottom--which I know you are not allowed to
5
use--that says and big malpractice awards were
6
awarded.
7
[Laughter]
8
DR. SANTANA: We don't want to get
into
9
that! Any other comments? Dr. Lee?
10
DR. LEE: I just wanted to make a
11
clarification that for the data that I presented
12 for
the non-opioid tolerant patients, the age range
13 was
from 1.5-5. I just wanted you to
understand
14
that. It doesn't give us an
overall 2-16 year-old
15
range.
16
DR. SANTANA: Dr. Luban, you deal
with
17
patients with sickle cell who have chronic pain
18
issues. Would you like to comment
on this issue?
19
DR. LUBAN: I think the biggest
issue
20
there is the complex use of more than one
21
analgesic, and the occasional failure of families,
22
when discharged, to follow the pain team's
105
1
recommendation and to really abuse the medications
2
because of continuing needs of the child. So, I
3 see
sickle cell disease and the use of this as a
4
real avenue of education that really should be
5
followed up on.
6
DR. SANTANA: Dr. Murphy?
7
DR. D. MURPHY: Do you think that
it is
8
clear--getting back to Dr. O'Fallon's
9
question--from the patient insert that after you
10
remove this product it is still absorbed? Do you
11
think that is clear enough in here, for longer
12
periods of time?
13
DR. SANTANA: Dr. Luban?
14
DR. LUBAN: I think that is not at
all
15
clear. I think that this is
written at a very
16
sophisticated level for some families to interpret.
17 We
certainly don't have high level language use
18
when we are doing informed consent, so why should
19 we
if we are trying to educate patients and
20
families?
21
DR. MCNEIL: Thank you. We will talk more
22
with the folks--there is actually a whole team of
106
1
people who help us write these patient information
2
inserts and they are supposed to be geared to the
3
sixth to eighth grade level. By
the giggles in the
4
room, I guess we have not hit that mark so I will
5
talk with people and we will see what we can do.
6 If
I understand you correctly, we should make it
7
slightly simpler.
8
DR. LUBAN: Speaking for our
patient
9
populations, I would say yes. The
use of the term
10
"opioid tolerant" is not a term that most parents
11 can
understand.
12
DR. SANTANA: And I am not even
sure a lot
13 of
physicians understand it.
14
[Laughter]
15
No, that is a fair observation.
16
DR. MCNEIL: The reason that we
used
17
"opioid tolerant"--I mean, I understand your
18
comment but the reason that we used "opioid
19
tolerant" was just to reflect the language in the
20
boxed warning, but I do understand what you are
21
saying and we will try to come up with something.
22
DR. SANTANA: Dr. Gorman and then
Dr.
107
1
Nelson.
2
DR. GORMAN: It strikes me how
attractive
3
this product would have to be to people doing ear,
4
nose and throat surgery on tonsils.
You have a
5
population with generally poor options for oral
6
medications in terms of their taste and
7
tolerability and adverse events of vomiting. So,
8 you
have a product that looks really attractive to
9
them because it is applied to the outside to an
10
obstreperous 4 year-old and you don't have to try
11 to
get them to drink something. If I was
targeting
12 my
educational process, ear, nose and throat
13
physicians and ambulatory surgery centers would be
14 at
the top of my list.
15
DR. MCNEIL: Excuse me, may I just
go back
16 to
Drs. Luban and O'Fallon? I just want to
make
17
certain that what we were speaking about before,
18
that the language is a bit too sophisticated is in
19 the
patient information section, not the actual
20
label? Correct?
21
DR. LUBAN: Correct.
22
DR. O'FALLON: The statement
"call your
108
1
healthcare provider right away of get emergency
2
help if you have trouble breathing or have other
3
serious side effects," that is in there on the
4
fourth page, without a bullet in a wholly bulleted
5
thing. I think you should move it
up to what is
6 the
most important information I should know.
It
7
should go there.
8
DR. MCNEIL: Thank you.
9
DR. CHESNEY: Dr. Nelson, you had
your
10
hand up?
11
DR. NELSON: Well, I think my
comment
12
follows from both of the last two, which is to also
13
look at the order within which you are putting
14
things, particularly given Dr. Gorman's comment.
15 The
first thing probably shouldn't be only use it
16 in
the way that your healthcare professional tells
17 you
to.
18
[Laughter]
19
Because we are talking about healthcare
20
professionals not using it appropriately.
21
DR. SANTANA: Dr. Hudak?
22
DR. HUDAK: I guess this is sort
of
109
1
getting at the question here, but the other avenue
2 for
education, it seems to me, since some of these
3
more egregious events occurred in the hospital
4
setting, is perhaps to have a letter that goes out
5 to
the hospital pharmacies, pediatric pharmacies
6
about this, and in this day and age where there are
7
computerized physician order entry systems it seems
8
that this would be a big way to sort of capture
9
that before it might become an issue.
10
DR. DANFORD: Does the FDA ever
11
communicate directly with risk management
12
individuals for hospitals and clinics?
Several of
13 the
speakers have suggested that the adverse events
14
might most likely be prevented by having a general
15
understanding that lawsuits can happen over misuse.
16
Perhaps the lawyers from hospitals and clinics who
17 try
to reduce their exposure to big settlements, if
18
they received something from the FDA about the
19
misuse of such products might actually do a lot of
20
work of educating the people who work in their
21
institutions.
22
DR. SANTANA: Dr. Nelson?
110
1
DR. NELSON: I think, at least in
my
2
setting, if a letter went out to the pharmacist you
3
effectively would accomplish that because it would
4
then go to the control mechanisms for prescribing
5
that would be used within a facility at least to
6
establish risk management strategies.
I would
7
probably prefer going that way because it is at
8
least then directed to the provision of care rather
9 than the other way.
10
DR. SANTANA: I would support
that. It is
11
within the scope of their care of what they should
12 be
doing with patients in terms of educating as
13
they get prescriptions filled, and so on and so
14 forth. So, I would support that too. Any other
15
comments? Dr. Murphy?
16
DR. D. MURPHY: I just wanted to
summarize
17 and
ask the Division to also pitch in here if they
18
don't think I have summarized correctly what we
19
have heard from the committee.
20
DR. SANTANA: I took some
notes. Would
21 you
allow me to do that? I think the
committee
22
would like the FDA to move in three directions that
111
1 you
have pointed out in this slide. One of
the
2
comments I heard very strongly from the committee
3 is
that the label needs to be re-looked at in the
4
context of maybe providing stronger statements,
5
regarding the inappropriate use resulting in deaths
6
that have already been observed, somewhere earlier
7 in
the actual label so that physicians and others
8
prescribing this can see that clearly early on.
9
I also heard a comment that there is a
10
section about qualifications of the prescriber and
11
those qualifications were kind of hidden in the
12
back of the information, and it should be brought
13
forward into the label too. So,
it is not a matter
14 of
re-writing the label but maybe providing some of
15 the
information in different sections, particularly
16 at
the beginning that would be more evident to
17
those that are prescribing. Those
are the comments
18
that I heard about the label.
19
I heard a lot of comments about patient
20
information and using the patient as an advocate
21 for
him or herself. I heard comments that
probably
22 the
reading language was inappropriate for the
112
1
populations that are being targeted in which this
2
medication could be used. So,
that needs to be
3
looked at very carefully.
4
I also heard some comments that I think
5
were very appropriate about clearer statements in
6 the
patient information regarding how, when this
7
medication or patch is removed, there will be
8
sustained levels that may continue to put you at
9
risk of having respiratory depression and
10
associated side effects.
11
Related to the patient information, I also
12
heard some comments about how the information in
13
that patient information leaflet should be
14
reorganized to put some of the highlights earlier
15 on
and make them more self-evident.
16
Then I heard a brief discussion about
17
education, primarily to prescribers.
I heard
18
various comments about some of the incident cases
19
that received care that had been by ENT, by
20
anesthesiologists, by pain teams.
I didn't hear a
21 lot
of discussion about how we could accomplish
22
that so I am going to seek a little bit more advice
113
1 from
the committee on how potentially that could be
2
accomplished. But I did hear that
there needs to
3 be
reeducation of people prescribing this and
4
potentially starting with some target populations
5 and
then moving it more openly, including
6
pharmacists, of course.
7
Then I also heard a very strong statement
8
about educating our patients who are using these
9
products and parents, and how we can best
10
accomplish that.
11
So, maybe the committee wants to spend
12
maybe one more minute probably advising the agency
13 on
potentially what educational systems may already
14 be
in place that they could target or the company
15
could target. If anybody wants to
add to that?
16 Dr.
Murphy?
17
DR. D. MURPHY: Thank you very
much. I
18
only have one question I want to clarify, and that
19 is
the label--the statement you had, Dr. Santana,
20 was
that we want a stronger statement concerning
21 the
deaths early in the label. I thought I
heard
22
that you wanted it in the black box.
114
1
DR. SANTANA: Yes, in the black
box. That
2 is
what I meant. That is correct.
3
DR. D. MURPHY: Thank you.
4
DR. SANTANA: Any further sort of
advice
5 to
the agency on this issue? I think we
have
6
discussed question one actually.
Am I correct?
7
DR. MCNEIL: Thank you for your
comments.
8 It
is very helpful to us. I am going to
take them
9
back for further discussions with the company.
10
DR. SANTANA: Thank you so
much. I think
11 we
are going to take a ten-minute break and start a
12
little bit after 10:30. Thank
you.
13
[Brief recess]
14
DR. CHESNEY: While everybody is
finding
15
their seats, I wanted to thank the FDA for
16
clarifying one issue which had to do with the use
17 of
ciprofloxacin and moxyfloxacin in ophthalmic
18
preparations. In the last two
slides the expected
19
cure rate of 70 percent, is that for conjunctivitis
20 in
adults? This study was actually done in
21
neonates. So, probably one can't
extrapolate from
22 one
to the other, just for clarification.
Dr.
115
1
Iyasu is going to introduce our next speaker.
2
DR. IYASU: Thank you. Our next speaker
3 is
Dr. Hari Cheryl Sachs. Dr. Hari Sachs is
a
4
professor of pediatrics at GW and
Children's
5
Hospital National Medication Center.
She has over
6 15
years of experience in private practice.
She
7
also served on the FDA non-prescription drug
8
advisory committee and is one of the FDA liaisons
9 to
the AAP committee on drugs. She will be
10
presenting the adverse events for venlafaxine.
11
Adverse Event Reports per Section 17 of BPCA
12 (cont.), Venlafaxine
13
DR. SACHS: Thank you very
much. I am
14
glad to be here to talk to you, guys.
It is
15
actually nice to see some familiar faces among the
16
crowd.
17
I will be discussing the adverse events
18 for
venlafaxine, and I think you, guys, are
19
familiar now with the basic organization of the
20
talk. Venlafaxine, or trade name
Effexor, has been
21 on
the market since December, 1993 for the
22
treatment of major depressive disorder, generalized
116
1
anxiety disorder and social anxiety disorder.
2
Although these are the indications in adults, there
3 are
no approved pediatric indications despite the
4
fact that exclusivity was granted in December,
5
2002, and the sponsor now goes by the name of Wyeth
6
Pharmaceuticals.
7
Venlafaxine and its active metabolite,
8
whose name I am not going to try to pronounce, is a
9 potent
inhibitor of both serotonin and
10
norepinephrine reuptake so this is actually an SNRI
11 but
for convenience I am going to refer to the
12
whole class as SRIs. It also is a
weak inhibitor
13 of
dopamine reuptake. These actions, along
with
14 the
lack of significant muscarinic cholinergic and
15
histaminergic effects, do alter the side effect
16
profile of the drug. The
half-life, which is about
17 5
hours for venlafaxine and 11 hours for the active
18
metabolite, is relevant for the timing of potential
19
discontinuation symptoms when they occur.
20
Venlafaxine was the fourth most commonly
21
prescribed antidepressant in the U.S. during 2003
22
and, as with other SRIs, prescriptions have been
117
1
rising in both pediatric and adult populations.
2
Although pediatric use seems to account for only
3
about 2.5 percent, this represents almost half a
4
million prescriptions. This use
is all off-label.
5
Disorders of mood and anxiety, along with ADHD are
6 the
most common indications that kids have been
7
treated for with venlafaxine.
8
I will now briefly describe the results of
9 the
studies performed for exclusivity. There
were
10 4
large, multicenter, double-blind,
11
placebo-controlled, parallel group, flexible dose
12
studies for each indication, that is, major
13
depressive disorder and general anxiety disorder.
14 The
dose used was flexible dosing between 37.5 mg
15 and
225 mg, and the age was 6-17 years. None
of
16 the
studies in major depressive disorder showed a
17
significant difference in placebo and,
18
interestingly enough, only one of the studies for
19
generalized anxiety disorder showed a positive
20
study result.
21
The endpoints in both the trials were
22
age-appropriate clinical symptom rating scales for
118
1
major depressive disorder. It was
the CDRS
2
revised. For the GAD trial it was
the Columbia
3
Kiddy Scale for Affective Disorders, or the KSADS.
4
Since only one study showed efficacy, that is why
5 no approval was granted.
6
Safety information was based in part on
7
these 4 studies, as well as 2 open-label trials, a
8
6-month trial in major depressive disorder and
9
another study in conduct disorder.
In these
10 studies
decreased weight gain and growth was noted
11
which was unrelated to treatment emergent-anorexia.
12 You
can see the numbers for the approximate weight
13
loss and weight gain in the placebo population, and
14 the
height. If you actually read the results
of
15 the
studies posted on the web, these numbers differ
16
slightly from that because the FDA received
17
additional information and analyzed it.
The other
18
thing that is kind of interesting is that it is
19 actually
important that the height effect was seen
20 in
the exclusivity studies. It is pretty
21
significant because it was a very short period of
22
time that the drugs were studied.
119
1
The other adverse event that was noted,
2
mild adverse event, is that there were elevations
3 in
cholesterol and blood pressure that were similar
4 to
those seen in adults. That also was
added to
5 the
label.
6
Since we are speaking of the label, let's
7
turn to the relevant safety labeling.
I would like
8 to
highlight several things about the labeling,
9
some of which is relevant to the safety discussion
10 or
the adverse events that we see, but also many of
11 the
changes are physically highlighted in yellow to
12
kind of emphasize that these are the new changes
13
that have been added actually since March.
14
In terms of neonates and pregnancy,
15
venlafaxine is considered a category C drug. That
16
means it should be used in pregnancy only if
17
clearly needed. Language
regarding the
18
discontinuation syndrome in newborns was added
19
fairly recently, first in January, 2003 and then
20
updated last month. It is found
in the section
21
under "non-teratogenic effects." What the labeling
22
describes is that you can see discontinuation
120
1 effects
in newborns with complications that may
2
require prolonged hospitalization or respiratory
3
support that may arise even as soon as delivery.
4 You
may also see some clinical findings, including
5
neurologic, respiratory and systemic symptoms.
6
These symptoms are consistent either with
7
discontinuation of the drug or potentially actually
8 a
direct toxic effect of the serotonin.
9
As you know, in part because of the
10
deliberations in February, a warning recommending
11
close observation for deterioration or emergence of
12
psychiatric symptoms in patients that are treated
13
with these SRIs was added in May, 2004 to the
14
label, and this warning actually supersedes and
15 replaces
some of the information that was in the
16
label previously regarding the association of
17
suicide with depression and other co-morbid
18
disorders, and a statement that Wyeth had added on
19 its
own that there were some suicidality seen in
20 the
pediatric trials. The other thing that
is
21
mentioned is the occurrence of sustained
22
hypertension.
121
1
These slides show an extensive list of
2
precautions which are listed in the order that they
3
appear in the label. Most of
these things were
4
seen in the top 20 adverse events that you have in
5 the
main report for venlafaxine, and we see some of
6
these in the post-exclusivity adverse events that
7
occurred during the year. I draw
your attention to
8 the
risk of bleeding and also the problems with
9
seizures, and then in the new labeling which is
10
regarding the weight loss and slower rate of growth
11 in
children.
12
In addition, under adverse reactions the
13
risk of symptoms with discontinuation, and this is
14 in
adults and older children, include both physical
15 and
psychiatric symptoms. In March there was
some
16
labeling added to the postmarketing reports on
17
dyskinesia and rhabodomyolysis.
18
So, now that you are familiar with all the
19
changes in the label, let's look at the adverse
20
events for the year. These are
actually the raw
21
counts of all the adverse events.
There were
22
approximately 1,500, of which about half are in the
122
1
U.S. and they may represent some duplicates.
2
Pediatric reports are really a relatively small
3
proportion, less than 4 percent of total reports
4 and
this has been pretty consistent over the past
5
several years since marketing.
There were only 2
6
deaths that occurred, and I will be discussing them
7 in
a few moments.
8
Turning to the specific 1-year
9
post-exclusivity period, there were 49 unduplicated
10
events. I apologize for the busy
nature of this
11
slide. There were 19 events that
involved in utero
12 or
maternal exposures and 30 that were direct
13
pediatric exposures. The gender
and age
14
distribution is seen here. Of
interest, there is a
15
male predominance of the adverse events in the
16
infants and neonates while the gender distribution
17 is
pretty similar in the older children.
Outside
18 the
neonatal period, most of the direct exposures
19
involved adolescents and children, as would be
20
expected.
21
Looking more closely at the in utero
22
events, there were actually no deaths reported
123
1
among the 19 in utero events, 3 of which also had
2
concomitant breast feeding. There
were 4 unrelated
3 congenital
anomalies; 2 had cardiac malformations,
4 1
with an ASD and the other with dextrocardia.
5
Another infant had hypospadias and the last infant
6 had
extra syndactyly, which is a fusion and webbing
7 of
the digits. Co-morbidities and other
8
medications were actually involved in all these
9
congenital anomalies.
10
Neurologic events were described in 11
11
infants. We saw 2 infants that
had hypotonia; 3
12
infants who developed seizures; and 6 infants who
13 had
disordered movements.
14
Just to illustrate how difficult it is to
15
sort out causality for these events, on follow-up
16 for
one of the infants who had seizures the event
17 was
considered unrelated to venlafaxine because the
18
patient had experienced a subarachnoid hemorrhage.
19
But, if you will recall, abnormal bleeding can be
20
associated with venlafaxine. So,
it is potentially
21
possible that the baby had subarachnoid hemorrhage
22 related
to the medication. Of course, that is
124
1
conjecture but just to show how hard it is to sort
2 out
the information.
3
The losartan in utero exposure events were
4
really 4 reports that detail complications that
5
occurred in babies with co-morbid conditions and
6
medications. They are described
here. While it is
7
less serious, it is something that has emerged in
8 the
literature so it is interesting that we did see
9 2
cases here as well.
10
Co-morbid disease and medications, as I
11
have explained, may contribute to some of these
12
events. Although 2 events were
coded specifically
13 as
neonatal withdrawal, there are actually up to 5
14
others where symptoms that emerged, like
15
jitteriness or tremor or seizure, could have
16
reflected neonatal withdrawal but it was just not
17
coded that way.
18
Prematurity was reported in 4 infants but
19 in
8 cases there actually was insufficient
20
information in the case report to determine whether
21 or
not a baby was premature. Three infants
were
22
breast fed. One mother reported
smoking and
125
1
drinking but that information about tobacco or
2
substance exposure or illicit drugs is actually not
3
present in the majority of the reports.
About half
4 the
infants were exposed to concomitant
5
medications, 4 of which were psychotropic. When I
6
looked back, 5 included benzodiazepines.
In only 2
7 the
case report expressly stated that there were no
8
other medications involved, and whether or not
9
medicines were involved in 7 of the other cases is
10 not
known.
11
So, in looking at the neonatal events,
12
they do seem to reflect the ones that are labeled
13 in
adults, for example tremor, convulsion and
14
hypotonia. The role of
concomitant medicines and
15
co-morbid conditions, such as prematurity, is
16
unclear. And, whether or not
symptoms such as
17
jitteriness or tremor or seizure are related to
18
withdrawal or serotonin toxicity is also unclear,
19 and
this will be discussed later today by Miss
20
Phelan and Dr. Levin.
21
Now I am going to turn to the 30 adverse
22
events that have been associated with direct
126
1
exposure. The majority of these
were neurologic,
2
psychiatric or related to overdose.
Of the 30
3
direct exposures, dose range and indication for use
4 is
not available for all of them but the doses were
5
generally within the approved adult dosing. Note
6
again that there is no pediatric dosing.
The
7
indications were combinations of depression,
8
anxiety or ADHD.
9
Health providers were responsible for the
10
majority of the reports. You
might recall that
11
that is in contrast to the information presented at
12 the
last session. Many patients were on
other
13
medicines in addition to venlafaxine.
Twenty were
14 on
concomitant medication, about two-thirds, and 10
15 patients
were on other psychotropic medications.
16
The majority of adverse events were
17
psychiatric. The 2 deaths in the
sample were due
18 to
completed suicides. Four attempted
suicides
19
also occurred. All of them were
overdoses, 1 case
20 of
suicidal ideation and 2 cases of self-injury
21
where there was no clear intent of suicidal
22
ideation. There are also 3 cases
of aggression and
127
1 agitation and 2 cases of behavior
changes. Once
2
again, there were concomitant medications in a
3
majority of these cases, 6 of which were
4
psychotropic.
5
Neurological events were the next most
6
frequent adverse events reported, and seizures,
7
loss of consciousness and motor or sensory
8
impairment are all labeled events and many of these
9
cases, again, involved other medicines or
10
underlying medical conditions.
11
There were 4 patients that developed
12
symptoms of serotonin toxicity, such as hyperexia
13
and/or neurological symptoms that were related to
14
overdoses. One of them, a 3
year-old, may have
15
accidentally ingested his sister's medication. The
16 other 3 were deemed non-accidental as the
intent
17 was
unclear. The remaining adverse events
occurred
18
singly and are labeled or related to labeled
19
events, with the possible exception of the specific
20
drug interaction between Augmentin and venlafaxine
21 in
this one case where the effectiveness of the
22
antidepressant was decreased.
128
1
As with the SRIs, you can see symptoms
2
with discontinuation or decrease in dose, and we
3 did
see physical symptoms or emergence of
4
psychiatric symptoms in 6 patients in these adverse
5
event reports. So, the adverse
events really are
6
related to labeled or already labeled events, with
7 the
possible exception of the events seen in
8
neonates.
9
Warning concerning the increased risk of
10
suicidal behavior did exist in venlafaxine's label
11
prior to our advisory committee in February but, as
12 you
know, there is now a new class warning for the
13
SRIs. The new labeling was added
as a result of
14 the
exclusivity studies regarding the effect of
15
growth and that was very recent.
The new class
16
labeling regarding maternal exposure and potential
17
occurrence of neonatal withdrawal with serotonin
18
toxicity will be discussed further this afternoon.
19
Routine adverse event monitoring for venlafaxine,
20 as
with all other drugs, will continue. I
hope you
21 are
not depressed.
22
[Laughter]
129
1
Are there any questions?
2
DR. CHESNEY: Thank you very much.
3
Questions? Dr. Ebert?
4
DR. EBERT: Is a non-accidental overdose
5
similar to a suicide attempt? Are
they classified
6
together or separately?
7
DR. SACHS: The reason these
actually were
8 not
classified as suicides, although, truthfully, I
9 made every attempt to do so, is that you
can't
10
tell. In one case, for example,
the kids may just
11
have been exchanging medication.
In another case
12 you
don't know if they were taking it to get high.
13
That is actually why.
14
DR. CHESNEY: Dr. Nelson?
15
DR. NELSON: It may be because it
is more
16
recent, but I haven't been able to find in the
17
label that we have in our book the growth change.
18
DR. SACHS: Yes, that change is
very
19
recent. What I think you got the
handout of the
20
whole statement of the adverse events that is kind
21 of
a summary of what our presentation is and these
22
sections of the label are included in that in
130
1
entirety. But I think this issue
is the point. I
2
mean, we are trying very hard, at FDA, to make the
3
labels available and there is a website now called
4
drugs at FDA. You can see that
for just this drug
5
alone there have been three changes to the label
6
since March.
7
DR. NELSON: Just as a follow-up
question
8 to
reinforce something I think Dr. Gorman said at
9 the
beginning of the day, if there is information
10
about that placed in because of pediatric
11
exclusivity, I would strongly recommend that the
12
standard comment about safety and effectiveness not
13
being established in children is changed to where
14 it
actually says 2 randomized, controlled trials
15
involving 353 children failed to demonstrate safety
16 and
effectiveness. If you look right under
it,
17
under geriatric use, it actually does list a number
18 of
adults[sic]. So, I think it is somewhat,
in my
19
mind, duplicitous to leave that general statement
20 in
which many pediatricians interpret as there were
21 no
studies.
22
DR. SACHS: And I think there is
always a
131
1
tension between kind of somehow having some tacit
2
approval if you put too much information in the
3
label.
4
DR. D. MURPHY: I think the other
issue
5 too
is that in this field particularly 2 studies
6
does not mean the product doesn't work.
I think
7
your point is that at least there is information
8 and
we ought to indicate that there is information.
9 Is
that really what you are trying to get at?
And,
10 I
think we ought to be able to find a way to do it.
11 The
issue is getting agreement within the Division
12
that they think that is an appropriate thing to do.
13 So,
you are telling us to get feedback to the
14
Division that you think it is an appropriate thing
15 to
do?
16
DR. NELSON: Yes, and no matter
how many
17
footnotes you want to put in about assay
18
sensitivity, I would still put it in there. I
19
mean, you can quote everything Bob Temple ever
20
wrote on the topic, as far as I am concerned, in
21 the
label if you want to do that.
22
[Laughter]
132
1
DR. D. MURPHY: Joan, is there
additional
2
concurrence with Dr. Nelson's comment?
3
DR. CHESNEY: Dr. Santana would
like to
4
comment.
5
DR. SANTANA: I want to take it
further.
6 I
have heard a lot of comments about this at
7
various meetings and I wonder if the direction that
8 we
should be taking--and this is a suggestion--is
9
that we start thinking about creating an area in
10 the
labels that is related to pediatric
11
exclusivity. It is here. It is being done. There
12 is
data. It doesn't imply that there is
enough
13
data to provide an indication or that there is
14
enough data to do all these other things that are
15 in
the label, but I wonder if a part of our mission
16 is
to educate practitioners and to educate the
17
public whether having a section in labels that
18
relates to pediatric exclusivity studies and trying
19 to
explain what those mean, obviously, in the
20
context of what those are really done for, would be
21
helpful.
22
Because, if not, the information is not
133
1
going to get there. The label is
not created for
2
that information. So, unless we
use the label in a
3
different way that information is not going to get
4
there. I don't know what the
challenges or the
5
barriers for doing that in the label are. I grant
6 I
am ignorant on that, but maybe that is something
7 we
should be aiming towards.
8
DR. CHESNEY: Can I comment? I think this
9 is
almost a slippery slope. I think by
putting
10
studies and results in the label--we already have
11
people prescribing for totally unapproved
12
indications, in fact contraindications, and if they
13 see
something in there, that there were really no
14 bad
side effects seen, and even though it is not
15
efficacious and approved, "hey, the studies were
16
done." I don't know if I am
expressing myself very
17
well but I think this is a very difficult area.
18
On the surface I would agree with what Dr.
19
Nelson is saying, but I think that it has to be
20
worded very carefully because I think when people
21 say
there are studies and, "hey, they haven't said
22 not
to use it so why don't I just go ahead," to me,
134
1
this is a little more complicated.
On the surface
2 it
seems like a no-brainer, but I think maybe there
3 are
some other issues. Dr. Hudak?
4
DR. HUDAK: Well, I would echo
prior
5
comments here. I look at this and
there are half a
6
million prescriptions of this in the pediatric
7
population a year, and I presume by the other
8
information you shared that most of that is in
9
association with treatment for depression, although
10 you
don't have any hard figures on that. You
know,
11 you
look at the labeling here and in terms of the
12
adult efficacy it talks about 5 studies that
13
demonstrated efficacy in adults that showed
14
improvement in at least 2 of 3 different clinical
15
measures. I think it is critical
to have the
16
pediatric trial information in there because you
17
have 2 studies with a significant number of
18
patients for this type of disorder where you have
19 no
effect. I think any information like
that is
20
very important to get out there.
21
I guess I may take a different tone than
22 Dr.
Chesney on this, but I think, you know,
135
1
information is good and I think the reason we have
2 a
half million prescriptions in pediatrics is
3
because we are looking at the adult studies and, as
4
someone else said, "well, it worked in adults, you
5
know, the same disease and it might work in
6
children; let's use it." On
the other hand, if you
7
have specific information that says we now have 2
8 studies that cannot find efficacy in 353
patients,
9
providers might have a different philosophy in
10
terms of what medication they will use in this
11
population.
12
DR. CHESNEY: I agree with
that. I think
13 it
is just that you have to be very careful about
14 the
wording because it does provide a very subtle
15
endorsement in a sense, just because it is there.
16 I
am probably not expressing this very well.
Dr.
17
O'Fallon and then Dr. Gorman.
18 DR. O'FALLON: One of the issues here is
19
that a lot of these were pharmacokinetic studies.
20 You
know, if a child is not treated at an effective
21
dose level it is not fair to count them as not
22
being very effective. In a certain
sense, by just
136
1
quoting, you know, 500 children were treated, that
2 is
not quite fair if they were being treated at
3
lower levels that were being used for
4 pharmacokinetic studies and that sort of
thing.
5
What I am saying is I think you have to be
6 a
little careful how you do it. I vote to
have
7
that information in there, but I am not an M.D.; I
8
don't treat these patients. But I
think you should
9
have the data in but don't just dump it because you
10
have to be careful about where that data came from.
11
DR. CHESNEY: Dr. Gorman?
12
DR. GORMAN: It strikes me that
for the
13
last six years with this group we have talked a lot
14
about the label. I have had the
opportunity to go
15
back and read old drug labels and they are very
16
brief, a page, maybe two pages.
The label
17
continues to try to struggle under its present
18
format to encompass new realities.
I am impressed,
19 and
I both dislike this and find it very valuable,
20
that in the era we now live in there are documents
21
that present executive summaries with embedded
22
links to information for those who want to pursue
137
1
more information.
2
I wonder--I love proposing work for other
3
people so let me do this really carefully--is there
4 an effort at FDA to create a new labeling
5
structure? You did that with OTC
medicines and I
6
think with great success. I think
the new drug fax
7
label is a major step in the right direction.
8
Could not a similar design be done for prescription
9
medicines with embedded electronic links, and the
10
official label stop being the piece of paper in the
11
document and start being an electronic form while
12
there is an official executive summary that
13
continues to go out with the product?
14
DR. D. MURPHY: Yes, there has
been years
15 of
work on this. You have to notify
industry that
16
they have to submit things electronically. You
17
have a whole process. There is a
deadline of when
18
they have to be doing that. One
of our problems
19
since I have been at FDA, since 1998, is that we
20
have been struggling with the fact that FDA doesn't
21
have available the labels. We
have been trying not
22
only to have them available to the public but
138
1
electronically available so you can link in these
2
ways. Yesterday I met with a
group that has
3
literally been working on this for years now. You
4
would think it would be simple.
It is not. As I
5
said, one of the first steps is getting things
6
electronically. The second thing
is getting it
7
maintained, updated, etc.
8
There is a new group that now has a
9
business plan associated with it so that we are
10
hoping that we actually will be able to have this
11
resolved, and there has been a lot of activity and
12
attention to this, I guess is what I am trying to
13
say. Everybody who is in FDA
fundamentally agrees.
14
This is our product. This is our
work. We need to
15
make it current and available and linkable and
16
searchable. From our perspective,
we also would
17
like to be able to search our labels so we can go
18 and
say I want to relabel that it has QT
19
prolongation as an adverse event.
We are trying to
20
accomplish more than just having a scanned-in label
21 up
on the web. That is what I am trying to
say.
22
There also is, and has been for a number
139
1 of
years a concerted effort to simplify and change
2 our
label, and there have been public announcements
3 of
that and feedback on that, making these labels
4
more user-friendly and that also is one of those
5
continuing works in progress. It
is very near but,
6 God
knows, I have been terrible at predicting.
7
So, yes, and I think that actually one of
8 the
things that I was thinking about because of
9
this tension, this dichotomy, you know, is you may
10 get
5 positive studies but you may have 10 others.
11 The
label cannot become a repository of negative
12
studies. But we need to be able
to find a way to
13
transmit the fact that there is information
14
available. What I think the
challenge for the
15
committee is how can we go forward with finding a
16 way
to put a statement in the label that there has
17
been data collected and how to get to that data, at
18 a
minimum--at a minimum. This whole
process of
19
having some sort of linkage would really make it
20
much easier.
21
DR. CHESNEY: I like your comment
that the
22
label can't become a repository of negative
140
1
studies. That is a good way of
phrasing it. Other
2
comments on this issue? Dr.
Hudak?
3
DR. HUDAK: I would say in a way
it has to
4 be
because those negative studies don't make it in
5 the
literature where they can be otherwise
6
accessed.
7
DR. D. MURPHY: There has to be a
way of
8
making the point of the negative studies and where
9
they are available, I think, and that they have
10
been done. That is what I think
you are saying.
11
DR. CHESNEY: Dr. Gorman?
12
DR. GORMAN: I think that is an
issue we
13
have been struggling with, that is, how do we
14
remove the veil that seems to be present for what
15 has
been done and is important clinically but isn't
16 out
there? One of the many goals of
pediatric
17
exclusivity and the process was to try to get
18
studies done in children and have that material
19
disseminated. If when it comes up
for
20
reauthorization, this continues to be an ongoing
21
problem of a continued veil of information, that
22
there is negative information out there, negative
141
1
either in terms of effectiveness or negative in the
2
fact that there are significant safety issues that
3 are
not presented, then that will have to be
4
readdressed in the legislative process.
5
DR. D. MURPHY: I think that the
positive
6
part of this is that, as you have noticed, actually
7 the
clinical review is up and it is for a
8
non-approval action. I don't
think you, guys,
9
realize what a watershed event this is.
This
10
information is otherwise not available except for
11
pediatrics now. So, it is getting
out there. One
12 of
the problems, as you know, is we have to
13
re-notify industry, as we explained last time, and
14 until we re-notify them we cannot put
information
15
up. That is now happening.
16
There is another potential problem.
It
17 may
or may not play out, but I do want to say that
18 the
agency agrees that one of the intentions of the
19
legislation was to try to make this information
20
available and to put what we would see as quality
21
information into the label. But
we are in a
22
situation where if we had a non-approval or an
142
1
approval it had to go up publicly and that is
2
happening. So, I think that is a
major watershed
3
event that is occurring for pediatrics.
4
DR. CHESNEY: Dr. Sachs, did you
want to
5 comment?
6
DR. SACHS: I just wanted to say
that this
7 was
a drug that was not approved and you can access
8
this information.
9
DR. CHESNEY: All right, I think
we will
10
move ahead then. We will have
more anticipated
11
discussion that includes venlafaxine when we talk
12
about the class this afternoon.
13
I want to clarify one issue. I
thought
14
that since question 1 had its separate bracket that
15 it
was for everything we discussed but I understand
16 it
was just for fentanyl. So, that has been
taken
17
care of.
18
Our next issue is the open public hearing,
19 and
before I read this two-page statement let me
20 ask
if there is anybody who wanted to present at
21 the
open public hearing this morning.
22
[No response]
143
1
No? Thank you very much. Our next
2
speaker is Dr. Murphy, who is going to give us a
3
pediatric update.
4 Pediatric Update
5
DR. D. MURPHY: I don't have any
slides
6 for
you. Actually, this is a ruse. I am going to
7
give you a very short update and then I hope to
8
indicate to you how much we have appreciated the
9
work of this committee.
10
Who is it that said "the best of times and
11 the
worst of times?" A dissolution has
led to an
12
evolution. By that, I mean that
you had better
13
watch out what you wish for. We
have long wanted
14
there to be a full pediatric advisory committee
15
which Congress has seen to do, to provide the
16
agency with, which is about the only way we were
17
going to get it because there are certain other
18
laws regulating how many advisory committees we can
19
have. So, Congress stepped in and
mandated that
20
there will be a full pediatric advisory committee,
21
which should help a lot with transparency so that
22
when we are having a meeting to talk about
144
1
pediatrics, it won't come out under infectious
2
diseases, and it was gracious of that committee to
3
chair you and to grow this subcommittee but it has
4
been very misleading to the public.
5
So, the good news is we have a new full
6
pediatric advisory committee, and it is charged
7
with a fair number of substantial activities which
8 you
have been told about previously, such as the
9
reporting of the post-exclusivity safety and
10
adverse events, such as the ethical issues and any
11
activity involving pediatrics within the agency
12
across all centers. So, there
will be the
13
construct of a new pediatric full advisory
14
committee.
15
The legislation also clearly tells us some
16 of
the representation that we need to have on that
17
committee, and we are working on that.
That
18
committee will be administered out of the Office of
19 the
Commissioner's Office and Tom Perez and others
20
have really done yeoman's work for us.
Jan
21
Johannssen, are you back there?
Would you like to
22
raise your hand so they will see that we now have a
145
1 new
exec. sec. for this committee and Jan is in
2
charge of making all of this happen.
As I stated,
3
this is also going to work across centers for
4 issues that may be coming up.
5
So, we have a new committee but with that
6 one
has to dissolve the old committee and this is
7
your last meeting as a pediatric advisory
8
subcommittee. It is really
sad. You know, we have
9
developed such an enormous database--I guess is the
10 way
to put it--of information with you that I wish
11 we
could just roll everybody from this committee
12
over to the new committee but I have been told that
13 is
not possible, and we did try to call and explain
14
that to everybody.
15
I want to take one more moment and just
16
quickly remind you of the work that you have done,
17 and
that is, we began this process thinking it was
18
going to be a typical sort of scientifically-based
19
activity and, clearly, it immediately became
20
evident that we would have to address the ethical
21
issues. This committee has
struggled with many
22
ethical issues, and you have advised the agency on
146
1 how
to approach trial design. When the
patient
2
doesn't have the disease, is that ethical? How can
3
that be done or can it be done at all?
4
You have advised us on placebo-controlled
5
trials in children and you have advised us on how
6 to
conduct research in a vulnerable pediatric
7
population. That advice has
resulted in consensus
8
statements that are now on the web, which I think
9 are
very helpful in answering questions that people
10 may
have because they were very thoughtful
11
discussions with a range of opinions and have been
12
referred to a number of times.
13
The scientific issues that you have dealt
14
with, in addition to the adverse event reporting
15
which you have a marathon day on today, you have
16
done on numerous occasions. You
can see some of
17 the
important scientific issues that have arisen
18
during this process, again, looking at the SSRIs,
19
looking a Duragesic patches, the neonatal
20
withdrawal, a number of events that you have all
21
asked for additional information on I think have
22
been important in helping us move this area
147
1
forward.
2
Another big milestone for this committee
3 and
an important one was the whole discussion of
4
therapies or interventions for infants who are
5 jaundiced. I think that was a very important
6
discussion and will continue to be important. I
7
think that this committee contributed very
8
significantly in the agency's assessment of how to
9
proceed in that arena of developing interventions
10 for
neonates who have hyperbilirubinemia.
11
You also dealt with issues of should we
12
even develop a product for children and I think a
13
very important contribution was to say no, such as
14 the
development of certain sleep products--don't
15
issue a written request for this.
This is not a
16
public health need we want to advocate.
17
Other big issues that you have
18
addressed--long-term follow-up, and this is an
19
ongoing problem. You didn't solve
it but you
20
helped us work at it, as we will continue to work
21 at
this because it is a very complicated process.
22
Also, I don't want to forget the topical products.
148
1
Things that one applies to the skin are not as
2
innocent as may always seem and this committee has
3
dealt with adrenal access suppression and potential
4
long-term carcinogenic effects and
5
immunosuppression of some topical products.
6
That is a lot. That is sort of on
top of
7
your ongoing activities and learning about all your
8 new
tasks that keep getting assigned to you with
9 the
new legislation.
10
Now, what I have to do today is say
11
goodbye to all of you because I guess the technical
12
legal term is that we have to declare you
13
dissolved.
14
[Laughter]
15
I have asked Raya McCree, who is our
16
administrative person who really runs the Office of
17
Counter-terrorism and Pediatric Development. She
18 is
why we get through every day, and she also
19
rescued your presentations today.
You would think
20
this would be simple to get but it is one of those
21
cartoons where the kid goes every which way, the
22
process that getting these plaque presentations
149
1
took. So, Raya, if you would come
up here?
2
I would like to start out with presenting
3 a
plaque and a certificate to Joan Chesney who, as
4 we
know, is the chairman of this committee and has
5
been the chairman of the subcommittee and who is a
6
professor of pediatrics at the University of
7
Tennessee. Joan, I tried to think
of how to say
8
this, what you have done has been so important, you
9
have helped bring this committee a level of
10
credibility within a scientific organization. I am
11
sure you, in academic medicine know, pediatrics is
12
always fighting for its academic recognition. It
13 was
really important that this committee be
14
perceived as a good science-based committee, and I
15
think Joan--all of you have--and Joan's leadership
16 has
been very important. She not only helped
bring
17
together this group and made sure that you all had
18
your say. She didn't try--and I
have seen chairmen
19 do
this--to intimidate people on the committee; not
20 let
them speak when a chairman didn't particularly
21
agree with their opinion. I think
she has been
22
very important in making sure that the committee
150
1 had
a say, whether they agreed with her or not, and
2 has
brought forth that consensus, and then helped
3 to
synthesize it. She has been very helpful
in
4
helping us synthesize what we think the committee
5
said.
6
Joan, if you would come on up here, I
7
would like to present you--I will ask each one of
8 you
actually to come up and I will give you a
9
little token of our appreciation.
Joan, thank you
10
very much.
11
[Applause]
12
Be careful where you put these.
They will
13
knock somebody down if they fall down.
Joan's
14
indicates that she was chair of the advisory
15
committee.
16
Judith O'Fallon has been our statistician.
17 One
thing that has been wonderful about Judith is
18
that she takes the statistical talk and makes it
19
applicable to the clinical. The
way she has been
20
able to condense the questions has been wonderful
21 and
much appreciated. As somebody said, we
talk
22
about therapeutic options and I don't think anybody
151
1 at
the FDA will forget the extra quivers that you
2
said we needed. Thank you very
much.
3
[Applause]
4
Let's see, who is next? Mimi isn't here.
5
Steve Ebert. After all this time,
I am still
6
mispronouncing your name. Steve
has been our
7
consumer representative, who goes through a
8
particular process to get to this place.
It is an
9
independent parallel process. We
wanted to thank
10 him
very much because I think your contributions
11
have been very thoughtful and have been the type of
12
comments that we would hope somebody in your
13
position would contribute.
14
[Applause]
15
Bob Nelson, Dr. Nelson. Dr.
Nelson,
16
professor of pediatrics, as you know, Department of
17
Anesthesia and Critical Care Medicine at Children's
18
Hospital in Philadelphia where he also serves on an
19
IRB. I went through that long
title because I
20
think it is important to know that one of the roles
21
that we hoped Dr. Nelson would play would be as our
22
point person for ethical issues and, boy, has he
152
1
done that! He has been invaluable
and has also
2
helped us identify other people to assist in the
3
more extensive discussions that we have had, and
4 has
been a critical person in the development of
5
this committee. Bob, thank you so
much.
6
[Applause]
7
Victor Santana, Dr. Santana is the
8
associate professor in hematology and oncology at
9 St.
Jude's. He has been our alternate chair
at
10
times; has been very helpful in helping us with
11
this whole issue of oncologic development where the
12
process is different, and has brought that
13
expertise to this committee. He
is also on the
14
Pediatric Oncology Subcommittee.
So, this has been
15 an
important liaison that we have had and we really
16
appreciate your time and effort.
Thank you very
17
much.
18
[Applause]
19
Dr. Danford is associate professor of
20
pediatrics and pediatric cardiology at the
21
University of Nebraska Medical Center and has been
22 our
cardiac expert. I know we haven't had
specific
153
1
drug issues in your area but we finally got one for
2 Dr.
Danford. In the last meeting on cardiac
3
imaging he did an outstanding job of synthesizing
4
that entire technical day and it has been very
5
useful to us and I really want to recognize that
6
specific effort, besides your overall efforts on
7 the
committee. Thank you very much.
8
[Applause]
9
Dr. Fink. Is he not here? Oh, shoot!
We
10
always count on him to give us a comment nobody
11
else would have thought of.
12
[Laughter]
13
Dr. Fuchs, Dr. Susan Fuchs. Dr.
Fuchs is
14 our
emergency medicine person. Actually, Dr.
15
Fuchs, you are one of the people that we haven't
16 had
a real product for but today we were counting
17 on
you to be able to provide some specific input as
18 far
as Duragesic is concerned, and we appreciate
19
your overall contributions very much.
Thank you
20
very much.
21
[Applause]
22
Dr. Gorman, general pediatrician in
154
1
Ellicott City, chair of the Committee on Drugs for
2 the
American Academy of Pediatrics and--how can I
3 say
it?--I am amazed at this man, I really am.
How
4 he
does this, continues to practice, stays up to
5
date, provides really insightful comments, is
6
chairing the Committee on Drugs at the Academy--he
7
just puts us to shame and I just want to thank you
8 for
your tremendous contributions.
9
[Applause]
10
Dr. Luban, who is the Vice Chair in the
11
Department of Laboratory Medicine, Director of
12
Transfusion Medicine and Quality Assurance for
13
Children's National Medical Center in Washington,
14 and
is our hematology and lab expert on the
15
committee. Gosh knows, the
diagnosis depends on
16 the
correctness and validity of the laboratory and
17 we
have been counting on her and she has provided
18
that type of expertise and helped to us and we want
19 to
thank you very much.
20
[Applause]
21
Dr. Sam Maldonado, Dr. Maldonado, we can't
22
give you a little thing--we don't give gifts to
155
1
industry.
2
DR. MALDONADO: I understand.
3
DR. D. MURPHY: But we can
recognize the
4
tremendous effort that you have provided by giving
5 us
your perspective, and you know we have relied on
6 you
many times during the conduct of these
7
committees to provide us that perspective and
8
input. We thank you very much and
you can come
9
back and give your carrot talk. I
just love your
10
carrot and stick talk.
11 [Applause]
12
Tom, I did recognize your efforts while
13 you
were out of the room. I want to make
sure that
14 you
knew that.
15
DR. PEREZ: Well, thank you, and I
would
16
like to recognize you for doing what you are doing
17
because, believe it or not, the bureaucracy gets in
18 the
way of doing things of this nature and it takes
19 a
little bit of money, clout and neither of those I
20
have.
21
[Laughter]
22
DR. D. MURPHY: Dr. Hudak, somehow
your
156
1
certificate isn't here. Dr.
Hudak, as you know, is
2 a
professor at the University of Florida
3
Jacksonville in neonatology. Dr.
Hudak has not
4
only contributed to the arena of information on
5
neonatology for drug development but, as I reminded
6 him
when I spoke to him the other day, he has had
7 the
joy of working specifically on the proton pump
8
inhibitor drug development program which continues
9
also to be in the process. I
wanted to thank you
10
very much and I am sorry we don't have your
11
certificate. We will get it to
you.
12
One last announcement is that I am
13 dissolving
myself too from being the Office
14
Director for the Office of Counter-terrorism and
15
Pediatric Drug Development as of September.
16
Somebody asked me was it not too much because I was
17 not
only doing counter-terrorism and pediatrics but
18 I
also was doing part-time in the Office of
19
Pediatric Therapeutics within the Office of the
20
Commissioner, and it got to be too much.
So, I am
21
going to go full-time to the Office of Pediatric
22
Therapeutics in September. So, I
will be seeing
157
1
many of you again, or some of you again, I hope.
2 But
no longer will I be with the Office of
3
Counter-terrorism and Pediatric Drug Development.
4
Dr. Shirley Murphy, who is a division
5
director for pediatrics, has done such an
6
outstanding job bringing together so many wonderful
7
people and getting this information to you, she is
8
going to continue to be here. And
Dr. Rosemary
9
Roberts--I had hoped she would be here but she said
10 if
I don't make it, they know what I look
11
like--will be the Acting Office Director. So, you
12 are
in very good hands anyway that you look at it.
13
Again, thank you all very much for your
14
participation.
15
[Applause]
16
DR. CHESNEY: I am going to take
the
17
chair's prerogative and add to the agenda. I just
18
wanted to say in our state of dissolution--
19
[Laughter]
20
--there are some people we would like to
21
thank. I am looking at the list
here and I hope I
22
don't forget anybody but, first of all, we have to
158
1
thank all the people that developed the legislation
2
that allowed us to be here at all.
So, I think
3
that involves people in the back of the room. It
4
involves mainly the Academy of Pediatrics but also
5 pediatric
department chairs, just a whole host of
6
people that even had the concept that children had
7 to
be recognized in terms of drug use.
8
I would like to thank Elaine Vining, in
9 the
back, and particularly Richard Gorman. I
think
10
they have done an amazing amount of
11
behind-the-scenes activity speaking in front of
12
Congress. In fact, Richard, you
did have dinner
13
with the President. Is that not
right?
14
DR. GORMAN: No, that is not
right.
15
DR. CHESNEY: Elaine may
have. Elaine,
16 why
don't you stand up? I don't know that
17
everybody in the room knows Elaine but she is the
18
legislative lobbyist--is that the correct
19
term?--for the American Academy of Pediatrics and
20 she
is really the one that has negotiated with all
21 the
congressional aides that work with the
22
senators. I had the opportunity
two weeks ago to
159
1 go
to the Hill to do some lobbying for the first
2
time and these legislative aides are really key,
3 and
Elaine has worked very, very closely with them
4 for
years now getting all these different laws
5
passed. So, I am so glad you are
here,
6
representing what the Academy does.
7
[Applause]
8
Richard, I don't know if you want to say
9
anything further about the Academy and the
10
Committee on Drugs.
11
DR. GORMAN: Never give up a
chance to
12
talk! I think this has been an
issue for the
13
Committee on Drugs for at least 35 years where it
14 has
ben written down, and Ralph Coffman, who is not
15 in
the room today, and Chet Berlin and Bob Ward, my
16 previous committee chairs, have carried this
torch
17 and
just passed it to me to, luckily, run the last
18 100
yards to get this legislation passed.
But the
19
Academy has been organizationally, systematically
20 and
bureaucratically involved in this effort and I
21
just happened to be the face at the end of this
22
process. As we know, we are not
at the end of this
160
1
process as refinements on our initial efforts
2
continue to be made.
3
DR. CHESNEY: Thank you. I also wanted to
4
particularly thank all the members of the staff who
5
have really made our job easy. We
really do 0.001
6
percent of the work when we sit here on the
7
committee because they have done all the work
8
behind the scenes. They have
selected what it is
9 we
are going to talk about--and who knows what they
10
don't give us to talk about. But
I have always
11
been assured that they don't bring the easy things
12 to
the committee so when we sit and struggle, I
13
think that is often correct.
14
But many, many people--and I will try to
15
recognize the few that my memory will allow me to
16
pull out--but Rosemary Roberts just walked in.
17
Stand up. I think everybody knows
Rosemary, but
18 she
has been almost as key to this effort from the
19
beginning as Dianne has. Shirley
Murphy, obviously
20
Susan Cummins. You don't know
that there are
21
always phone calls behind the scenes and Susan and
22
Shirley are on those. Rosemary
Addy, who most
161
1
recently has been extremely helpful in all of this,
2 and, Solomon, I feel like you have become one
of
3
this group of staff because you present to us so
4
often and represent so many of the issues. Tom has
5
been a wonderful executive secretary.
I think we
6 are
all befuddled by everything that goes on in the
7
FDA, but Tom is one of the people who tells us what
8 we
can and can't say and tells me when I can and
9
can't announce lunch and some very fundamental
10
things like that.
11
[Laughter]
12 But Tom has been just enormously
helpful
13 and
gets our emails to us on time, gets us our
14
reservations and gets us our limousines to the
15
airport, which he will do today in spite of the
16
ongoing events.
17
Then, all the medical officers who have
18
presented to us--I can't tell you how impressive
19 and
what an inspiration it is how clearly you
20
present; your slides are perfect.
They are always
21
readable. They are succinct. They are right to
22 the
point. I don't know who rehearses behind
the
162
1
scenes all the time but I assume it is Susan and
2
Shirley and Solomon, and lots of other people. But
3 we
really respect that you respect our time and
4
make it so much easier for us.
5
I think Bill Rodriguez and Don Madison are
6
both in the room, and Don Weis, but they have also
7
been very helpful to this whole process.
8 I am only going to say thank you to
Dianne
9
because you have been key. As you
can tell, I am
10 an
intuitive person and I don't handle this kind of
11
dissolution very well.
12
[Applause]
13
Anyway, thank you Dianne. You
have been
14
everything to this committee.
Thank you.
15
[Applause]
16
I have discussed this with a few members
17 of
the committee and I wanted to bring it again to
18 the
committee's attention. I mentioned it to
19
Dianne yesterday when we were at a very interesting
20
meeting of which I will just give you a
21
two-sentence summary, the Food and Nutrition
22
Committee has very serendipitously discovered the
163
1
presence of furan, which has some very remote
2
similarities to dioxin. It is
used to dissolve
3
resins, to prepare lacquers in a variety of
4
industries, but they have discovered very small
5
concentrations of it, on the order of parts per
6
billion, in a number of foods, primarily those that
7
have been canned and prepared.
Interestingly,
8
because the issue came up with the pediatric
9
formulation of apple juice, they looked at other
10
pediatric foods and it is in formulae and it has
11
been in a number of pediatric baby food which is
12
prepared in bottles by heating.
13
They have been extremely diligent about
14
putting this on the web. It has
been out there
15
since May 7 if you want to look and find all the
16
details. They are still very
busily trying to look
17 at
other foodstuffs. They are working very
closely
18
with the folk in Canada. They
have an extremely
19
sensitive mass spec assay now which has allowed
20
them to detect this. Of course,
nobody knows if it
21
means anything at all. But it is
out there now and
22
they are working very hard, and Dianne and Susan
164
1
were kind enough to ask me and Dr. Gorman, who
2
wasn't able to come, if we would go and listen and
3
comment. Dianne was also
there. We represented
4 the
committee in telling them that we would like
5
them to look for the presence of furan in a variety
6 of
situations, including the fetus, the
7
mother-fetal diet, the newborn infant who may have
8
extremely permeable guts, and look at whether this
9
furan is concentrated in specific tissues, look at
10
fetal and infant animal models, and so on and so
11 on.
12
I won't elaborate any further, except to
13 say
that that should all be up on the website and
14
that will be evolving. But in the
process I had an
15
opportunity to talk to Dianne and I told her what
16
many of us have felt, which is that the issues that
17 are
covered on this committee are so important and
18 so
interesting and generally not available to 99
19
percent of those caring for children just because
20
most of us don't go to the Federal Register on a
21
regular basis or go to the FDA website even though
22 we
have a vested interest in it.
165
1
So, one thought we had was that this
2
committee provide a synopsis of the events of each
3 of
its meetings to be published in potentially a
4
pediatric journal. Pediatrics
comes to mind right
5 away because it is the official spokes item
for the
6
Academy and generally one that is read by all those
7 who
care for children. This is at the moment
a
8
total hypothetical construct because the editors of
9
Pediatrics may say they don't want to have anything
10
that is not pure science and heavily peer reviewed,
11 but
that was the suggestion because issues like the
12
whole bilirubin issue I think are just fascinating
13 and
they are just not out there.
14 I have talked to my colleagues and I
have
15
told them about it; most of them don't know it.
16 So,
that was the suggestion and I would be very
17
interested in comments from the committee and the
18 FDA
and anybody else. The thought might be
that
19
somebody on the committee would write a brief
20
summary and perhaps it would be the person who
21
specialized in that particular subject or area, or
22
perhaps it would be the chair, which I can say
166
1
since I don't even know whether I will be on the
2
next committee. Then the FDA
would review it to be
3
sure that there was nothing that had been
4
accidentally included which is still confidential,
5 and
then submitted to the journal. So, I
would be
6
interested in comments or suggestions, other places
7 to
publish it--New York Times, Wall Street Journal.
8
Anyway, if you have comments, please let me know or
9 let Dianne or Susan or Shirley or anybody else
10
know. Shirley?
11
DR. S. MURPHY: I would just like
to say
12
that we have been discussing internally about how
13 to
get information, how to disseminate information,
14 and
I totally agree with you that it is just not
15 out
there. It reaches sometimes the
newspapers if
16 it
is really controversial but I think a systematic
17 way
of having a regular column and reporting in
18
Pediatrics, and I think your idea of sharing the
19
responsibility, and then we would be happy to fact
20
check it because the slides are publicly available
21 on
the web, it is all in the public domain, what is
22
discussed here, unless it is a closed session. So,
167
1 I
think it is not too onerous a job and we would be
2
happy to pitch in an help with that.
3
DR. CHESNEY: Skip?
4
DR. NELSON: Two comments, I think
if the
5
idea was to have an ongoing mechanism by which
6
information could get out to pediatricians, that
7
probably wouldn't be Pediatrics as a venue but
8
might be something like AP news where there could
9 be
an interest in more timely and less sort of
10
academic discussions. Part of the
problem with
11
this is who is going to write the first draft. I
12
mean, there are some practical things.
But if, in
13
fact, that was done one of the questions would be
14 to
what extent it could be a broad sort of
15
reflection on pediatric drug development--where has
16 it
been; where is it going, with a focus on the
17
committee but not just simply a historical basing
18 of
the topics but also stepping back and looking at
19
some of the broader process issues that we bring
20 up;
labeling issues that we have discussed; and
21
those kinds of things. If we did
that, it would
22
probably have to be more of a product of the
168
1
individuals on the committee and not of the
2
committee nor of the FDA because I presume there
3 are
some things that people in the FDA couldn't in
4
fact say.
5
DR. D. MURPHY: I think there are
a
6
variety of ways to approach this.
One possibility
7 is
just this synthesis of the discussion.
At least
8 in
one option here it would be limited to the facts
9
that were presented and the discussion, and it
10
would be synthesized--these were the issues; these
11
were the pros and cons; this is what the committee
12
advised; this is what might be happening. So
13
pediatricians, family practice people who take care
14 of
children would know that this is you, out there.
15
The broader topic I think is always
16
something that is an option for anybody on the
17
committee who can use this information because it
18 was
publicly presented. But I think what
Joan was
19
talking about was trying to identify maybe not
20
every meeting but those scientific issues that have
21
come up. You all are a panel of
experts that were
22
brought together; you think about it; and when you
169
1
think that it is important that somehow it be
2
synthesized and made more available.
3
DR. CHESNEY: One other group I
forgot to
4
thank is the committee itself. I
think this has
5
been a wonderful group and we have enjoyed each
6
other's company when we were allowed to talk to
7
each other. Thank you all for
making the little
8 bit
that I have had to do so much easier.
Tom, do
9 we
have permission to eat now? Why don't we
plan
10 to
reconvene no later than 12:30 so that we can
11
continue to move things ahead in terms of traffic?
12
Thank you.
13
[Whereupon, the proceedings were recessed
14 for
lunch, to reconvene at 12:30 p.m.]
170
1
A F T E R N O O N P R O C E E D I
N G S
2
DR. CHESNEY: I think we are ready
to
3
start. There were two people that
came in after we
4 did
the formal introductions this morning so I
5
wondered if they could both introduce themselves.
6 Dr.
Cragan and Dr. Luban.
7
DR. CRAGAN: I am Jan Cragan. I am a
8
pediatrician with the Division of Birth Defects and
9
Developmental Disabilities at CDC.
10
DR. CHESNEY: Thank you. Dr. Luban?
11
DR. LUBAN: Naomi Luban, pediatric
12
hematologist, Children's Hospital National Medical
13
Center in Washington, D.C.
14
DR. CHESNEY: Thank you. Now Dr. Iyasu is
15
going to--my apologies. As I told
you, Tom keeps
16 us
in line. He has to read a second meeting
17
statement before we have the next session. Thank
18
you.
19 Meeting Statement
20
DR. PEREZ: Thank you and good
afternoon.
21 The
following announcement addresses the issue of
22
conflict of interest with respect to the update on
171
1
neonatal withdrawal syndrome and congenital eye
2
malformations reported in infants whose mothers'
3
used an SSRI during pregnancy and is made part of
4 the
record to preclude even the appearance of such
5 at
this meeting.
6 Based on the agenda, it has been
7
determined that the topics of today's meeting are
8
issues of broad applicability and there are no
9
products being approved at this meeting.
Unlike
10
issues before a committee in which a particular
11
product is discussed, issues of broader
12
applicability involve many industrial sponsors and
13
academic institutions. All
special government
14
employees have been screened for their financial
15
interests as they may apply to the general topic at
16
hand. Because there has been
reported interest in
17
pharmaceutical companies, the Food and Drug
18
Administration has granted general matters waivers
19 to
the special government employees who required a
20 waiver under a waiver under Title 18 U.S. Code
21
Section 208 which permits them to participate in
22
today's discussion.
172
1
A copy of the waiver statement may be
2
obtained by submitting a written request to the
3
agency's Freedom of Information Office, Room 12A-30
4 of
the Parklawn Building.
5
Because general topics impact so many
6
entities, it is not prudent to recite all potential
7
conflicts of interest as they apply to each member,
8
consultant and guest speaker. FDA
acknowledges
9
that there may be potential conflicts of interest
10
but, because of the general nature of the
11
discussion before the committee, the potential
12
conflicts are mitigated.
13
With respect to FDA's invited industry
14
representative, we would like to disclose that Dr.
15
Samuel Maldonado is participating in this meeting
16 as
an industry representative, acting on behalf of
17
regulated industry. Dr. Maldonado
is employed by
18
Johnson & Johnson.
19
In the event that the discussions involve
20 any
other products or firms not already on the
21
agenda for which an FDA participant has a financial
22
interest, the participants are aware of the need to
173
1
exclude themselves from such involvement and their
2
exclusion will be noted for the record.
With
3
respect to all other participants, we ask in the
4
interest of fairness that they address any current
5 or
previous financial involvement with any firm
6
whose product they may wish to comment upon. Thank
7
you.
8 DR. CHESNEY: Thank you.
Dr. Wisner,
9
could you introduce yourself, please?
10
DR. WISNER: My name, is Kathy
Wisner and
11 I
am from the University of Pittsburgh. My
work
12
involves studies of depression and its treatment in
13
childbearing aged women.
14
DR. CHESNEY: Thank you. What department
15 are
you in there?
16
DR. WISNER: I have academic
appointments
17
primarily in psychiatry, but secondary appointments
18 in
OB-GYN and epidemiology.
19
DR. CHESNEY: Thank you. Dr. Iyasu?
20
DR. IYASU: It is my pleasure to
introduce
21 the
first speaker for this session, which is an
22
update on neonatal withdrawal syndrome.
Kate
174
1
Phelan is a pharmacist and works at the FDA. She
2 has
spent six years as a drug information
3
specialist at the United States Pharmacopeia before
4
coming to the FDA. In her current
position she is
5 a
safety evaluator in the Office of Drug Safety.
6 She
has been with FDA since 1999.
7
Update on Neonatal Withdrawal Syndrome
8
MS. PHELAN: Hi. My name is Kate Phelan.
9 I
am a pharmacist. I work as a safety
evaluator in
10 the
Office of Drug Safety.
11
In November of 2001 I completed a review
12 of
reports of neonatal withdrawal syndrome of
13
serotonin uptake inhibitors. I
will present that
14 review
to you today. First, I will give a brief
15
overview of the FDA Adverse Event Reporting System,
16 or
AERS, so you will understand the context and the
17
source of the neonatal withdrawal syndrome cases
18
that I reviewed. Second, I will
describe the
19
process of evaluating an adverse event.
Third, I
20
will present my review of neonatal withdrawal
21
syndrome after in utero exposure to SRI drugs.
22
Finally, I will give a few conclusions.
175
1
The FDA's database of adverse events
2
reported for drug and biological products is known
3 as
AERS, which stands for Adverse Event Reporting
4
System. Adverse event reports
come from healthcare
5
professionals, consumers, medical literature and
6
postmarketing trials. Healthcare
professionals and
7
consumers report to manufacturers and, through
8
MedWatch, they report directly to the FDA.
9
Reporting by healthcare professionals is
10
voluntary. However, drug
manufacturers are
11
required to send adverse event reports that they
12
receive to the FDA in various time frames based on
13 the
severity and expectedness of the event.
14
Expectedness is determined by drug labeling.
15
There are some limitations to AERS data.
16
Some limitations pertinent to the issue of neonatal
17
withdrawal syndrome are that the reporting is
18
voluntary and, therefore, adverse events are
19
under-reported. The FDA does not
have drug usage
20
data for use during pregnancy.
For these reasons,
21 we
cannot calculate true incidence rates using
22
these data.
176
1
Many reports lack information, especially
2
about other drugs that may have been used by the
3
mother. Also, most reports do not
specify what
4
steps were taken to eliminate other possible cause
5 for
the signs that are seen in the neonate.
6
Reporting biases affect adverse event reporting.
7 For
example, media attention, such as Paxil has
8
received in recent years, can stimulate uneven
9
reporting between drugs. Also,
the length of time
10 a
drug has been marketed affects adverse event
11
reporting. Reporting bias can
invalidate
12
comparisons of drugs that are made based on the
13
numbers of reports. Therefore,
AERS data can
14
suggest but it cannot confirm that a drug caused an
15
adverse event or that drugs differ in relatedness
16 to
the adverse event.
17
So what good is AERS? AERS is
invaluable
18 in
helping to discover previously unknown adverse
19
drug events, especially adverse events that occur
20 too
rarely to be seen in clinical trials or that
21
occur in populations that are excluded from
22
clinical trials such as pregnany women.
AERS data
177
1
must be supported by further investigation. Safety
2
evaluators obtain follow-up information from
3
reporters of important cases if possible and we
4
review the medical literature.
Also, FDA new drug
5
review divisions may revisit previously submitted
6
drug trial data or even request additional study by
7 a
drug sponsor. So, attempts are made to
obtain
8
data from numerous sources in determining
9
association between the reported adverse event and
10 a
suspect drug.
11
Each safety evaluator in the Office of
12
Drug Safety monitors a fixed group of drugs for
13
adverse events that are possibly related to the
14
drug and are unexpected or of greater severity,
15
frequency or specificity than is described in drug
16
labeling. Safety evaluators may
contact reporters
17 for
additional information and we search AERS and
18 the
medical literature for similar reports, as I
19
mentioned.
20
Each report is evaluated for relatedness
21 to
drug and included or excluded from the case
22
series using case definition criteria developed by
178
1 the
Office of Drug Safety or by the safety
2
evaluator. Case definitions are
used to provide
3
consistent characterization of the adverse event
4 and
to facilitate retrieval of clinically relevant
5
cases. Findings and
recommendations of the Office
6 of
Drug Safety are sent to the new drug review
7
divisions for their consideration.
8
Now that you have a general understanding
9 of
the Office of Drug Safety's reviews, I will
10
present my review of neonatal withdrawal syndrome
11
with SSRI drugs. The drugs that I
led are
12
citalopram, fluoxetine, fluvoxamine, paroxetine,
13
sertraline and venlafaxine.
Collectively, I am
14
referring to these drugs as serotonin reuptake
15
inhibitors or SRIs. As you know,
the first 5 drugs
16
selectively inhibit serotonin reuptake and
17
venlafaxine inhibits both serotonin and
18
norepinephrine uptake. Citalopram
was not approved
19 in
the U.S. at the time this review was completed.
20
Because adult discontinuation syndrome is
21 a
known effect of these drugs, when reports of
22
neonatal withdrawal syndrome appeared in AERS it
179
1 was
logical to believe that there might be an
2
association between the reported signs in the
3
neonate and the abrupt discontinuation of the SRI
4
that occurred at birth.
5
Reports in AERS are coded using the medDRA
6
terminology. MedDRA is a
hierarchical dictionary
7
designed for use in drug regulation.
In fact,
8
MedDRA stands for Medical Dictionary for Regulatory
9
Activities. I began with a review
of AERS cases
10
with the MedDRA code drug withdrawal syndrome,
11
neonatal. This review showed
predominantly
12
neurological, neuromuscular and autonomic effects
13 so
I broadened my AERS search accordingly.
14
Ultimately I did 3 AERS searches, focusing
15 on
neurological, neuromuscular and autonomic
16
events. In the first search I
used MedDRA terms
17
specific to neonates. In the
second search I used
18
general MedDRA terms but I restricted the search to
19
cases in which the patient was reported as age 0-3
20
months. The third search was
performed because
21
complications of maternal exposure to therapeutic
22
drugs are sometimes reported and coded in AERS as
180
1
though the mother were the patient so if I had
2 searched restricted by age, I would not have
3
retrieved those reports.
4
I also searched PubMed for related studies
5 and
cases. All cases retrieved from PubMed
were
6
also in AERS and will be covered in this talk. The
7 few studies available at that time will be
8
mentioned by Dr. Levin who will speak after me.
9
In deciding whether to include each case
10 as
neonatal withdrawal syndrome possibly associated
11
with SRI, I applied these criteria.
These criteria
12
were adapted from the article "Serotonin Reuptake
13
Inhibitor Discontinuation Syndrome: A Hypothetical
14
Definition," by Schatzberg et. al. that appeared in
15 the
Journal of Clinical Psychiatry in 1997.
Please
16
note that the case definitions that we apply to
17
AERS data evaluation are not synonymous with
18
diagnostic criteria. Our case
definitions must be
19
useful in the setting of incomplete data.
20
The case should have all 4 of the
21
following characteristics: First, the mother had to
22 be
taking an SRI up to the birth. Cases
were
181
1
excluded if the SRI was discontinued before the
2 birth.
3
Second, the observed signs should not be
4
attributable to factors other than the discontinued
5
administration of the SRI. Many
cases were
6
excluded because the mother was also taking a
7
benzodiazepine which could cause withdrawal in the
8
neonate.
9
Third, the signs of withdrawal should not
10 be
present at birth but should appear with some
11
delay after birth. It is possible
for withdrawal
12 to
be seen at birth depending on the timing of the
13
mother's last dose and the half-life of the drug
14 but
I applied this criterion in an attempt to
15
distinguish withdrawal from serotonin toxicity in
16 the
neonate.
17
Fourth, the sign should resolve.
Part of
18 the
hypothetical definition of SSRI withdrawal in
19
Schatzberg is that withdrawal syndrome is a
20
transient phenomenon. In a few
cases the adverse
21
event was persisting months or years after birth.
22
These cases were excluded.
182
1
Finally, most cases were reported by
2
healthcare professionals. I did
not want to
3
question the clinical judgment of the healthcare
4
professional who had witnessed the event. So, if
5 the
reporter called the adverse event suspected or
6
diagnosed SRI withdrawal and the information in the
7
case did not contradict either the first or the
8
second criterion that I have here, then the case
9 was
included in the case series. In many of
the
10
cases included on this basis the adverse event was
11
present at birth or was persisting at the time the
12
case was reported.
13
My AERS search retrieved the number of
14
cases that appears in the column headed "2001." I
15
reviewed the cases and applied the case definition
16
that I just described. The number
of cases that
17 met
the case definition of neonatal withdrawal
18
syndrome possibly related to the SRI appears in the
19
middle column, headed "met definition." Numbers of
20
cases received by FDA since the November, 2001
21
review of this issue appears in the final column,
22
headed "2001-4." I need
to stress that the counts
183
1 in
that column are raw case counts, the cases that
2 had
no evaluation, and I have included it merely to
3
illustrate that we are still receiving cases.
4 A total of 57 cases met the case
5
definition. In 47 of these cases suspected or
6
diagnosed withdrawal syndrome was reported.
7
Thirty-seven additional cases were excluded because
8 the
adverse event was present at birth. This
9
contributed to the new drug review division
10
decision not to distinguish withdrawal from
11
toxicity that Dr. Levin will discuss.
12
As an example to show you why so many
13
cases were excluded from the case series, I will
14
present fluoxetine. Fifty-six
unduplicated cases
15
were retrieved for fluoxetine and 52 of these cases
16
were excluded for the reason shown.
I will
17
elaborate on only 2 of these reasons.
In 8 cases
18 the
reported adverse event was not consistent with
19 the
characteristics of the other withdrawal cases.
20
These cases included 4 congenital anomalies, 3
21
adverse events that occurred during breast feeding,
22 and
1 report of dehydration. Also, 1 case was
184
1
excluded because fluoxetine administration to the
2
neonate did not relieve symptoms.
Although this
3 was
not a criterion of the case definition, it is a
4
characteristic of withdrawal.
Administration of a
5
similar drug should relieve symptoms.
6
In the neonatal withdrawal case series
7
there were 56 pregnancies and 1 twin birth. The
8
mother's age was unknown in most cases.
The
9
diagnoses for maternal SRI use was depression in
10 the
majority of cases that included this
11
information. There were several
diagnoses reported
12 in
one case each that I did not include here.
SRI
13
dosage was within labeled recommendations except
14 for
one venlafaxine case in which the mother was
15
taking 450 mg/day. She was taking
venlafaxine
16
tablets which have a maximum recommended dose of
17 375
mg/day for severe depression. Some cases
18
include drug use that may have been confounding and
19
perhaps, in retrospect, I should have excluded some
20 of
these cases. These were occasional
alcohol, in
21 4
cases; cigarettes, in 7 cases; and marijuana, in
22 2
cases. However, these cases were
distributed
185
1
among the SRIs so they should not greatly affect
2 the
data.
3
Neonates were premature in 5 of 35 cases
4
that included length of gestation.
These were 1
5
fluoxetine and 4 paroxetine cases.
There were 25
6
males and 17 females. Birth
weights averaged 3.04
7 kg
in the 28 cases that included birth weight
8
Apgar scores averaged 7-9 at 1, 5 and 10 minutes.
9
On this slide the drugs are
listed in
10
order of increasing half-life.
Time from birth to
11
onset of the adverse event and duration of the
12
adverse event are presented as median times if
13
there were 3 or more cases that contained this
14
information. The reported times
to onset and the
15
duration of signs actually covered rather broad
16
ranges. However, the median times
to onset
17
somewhat follow half-life. Onset
and resolution of
18
signs may be difficult to pinpoint clinically and
19
there are few cases here so we can't really draw
20
conclusions from this.
21
These are the adverse event terms that
22
were reported in more than one case.
They are
186
1
grouped by body system and presented by decreasing
2
number of mentions. The profile
is similar for all
3
SRIs, with nervous system and neuromuscular
4
excitation most frequently reported.
Feeding and
5
breathing difficulties and temperature
6
dysregulation were also reported.
Additionally, a
7
number of breathing difficulties were reported in
8 one
case each, including apnea episodes, gasping,
9
shallow respiration and hypoventilation.
10
A comparison with reported signs and
11
symptoms of discontinuation syndrome for
12
venlafaxine and SSRI class labeling show some terms
13 in
common with the neonatal reports. These
are
14
irritability and agitation. Most
of the other
15
terms in the class labeling are subjective and
16
would not be observable in a neonate.
17
More than half of the cases reported some
18
treatment for withdrawal, most commonly increased
19 hospital
stay. Regarding outcome, the case
20
definition for accepting cases as neonatal
21
withdrawal possibly related to the SRI specified
22
resolution of signs unless the reporter said SRI
187
1
withdrawal or was diagnosed or suspected. So, most
2 of
the cases did resolve.
3
In conclusion, there are possible cases of
4
neonatal withdrawal reported for all of the SRIs
5
approved at the time of this review.
They reported
6
similar signs in the neonates.
Thus, the AERS data
7
support the occurrence of neonatal withdrawal as a
8
class effect of the SSRI drugs.
9
The most frequently reported signs of
10
neonatal withdrawal are excitatory nervous and
11
neuromuscular effects. Breathing,
feeding and
12
thermal regulation difficulties have also been
13
reported. Neonates exhibiting
signs of SRI
14
withdrawal may require supportive treatment.
15 Therefore,
healthcare professionals should be made
16
aware that adverse events may occur soon after
17
birth in neonates exposed to SRI drugs in utero at
18 the
end of pregnancy.
19
The purpose of my review was to examine
20 SRI
withdrawal. However, some of the cases
that I
21
excluded from my case series, particularly those
22
excluded because the adverse event was present at
188
1
birth, suggest that SRI toxicity may also occur in
2
neonates exposed to these drugs in utero. Dr.
3
Levin will discuss that issue further.
Thank you.
4
DR. CHESNEY: Should we hold
comments and
5
questions until the other two speakers?
Are there
6 any
technical questions that anybody has for Dr.
7
Phelan?
8
MS. PHELAN: I am not a
doctor. That is
9 why
I said I was a pharmacist. I don't want
any
10
false expectations!
11
DR. CHESNEY: I call pharmacists
doctors
12
also. Next speaker?
13
MS. PHELAN: Dr. Robert Levin is
the next
14
speaker. Dr. Levin is a medical
reviewer in the
15
Psychiatry Section of the Division of
16
Neuropharmacological Drug Products.
Prior to
17
coming to FDA, he was with the NIMH where he worked
18 as
a health policy analyst in the Office of the
19
Director and as an NIMH staff fellow in the
20
Geriatric Psychiatry Branch.
Before working at
21
NIH, Dr. Levin had practiced in clinical
22
psychiatry.
189
1
DR. LEVIN: I will be talking
about a
2
recent FDA class labeling initiative regarding
3
SSRIs and SNRIs. In particular we
are focusing
4
today, of course, on the neonatal adverse events.
5
With that initiative we are also discussing and
6
proposing class labeling for adult discontinuation
7
symptoms which we will discuss a bit today in
8
comparison and contrast to neonatal symptoms.
9
Here is an example of one of the little
10
boys and girls we will be discussing.
11
These are the particular drugs that we
12
will be discussing as well. They
are all marketed
13
SSRIs and one marketed SNRI, venlafaxine as well.
14
Here are the objectives. One is
to
15
present highlights of the proposed class labeling
16
that we have for both precautions sections,
17
pregnancy and also dosage and administration.
18
Also, I would like to provide a rationale for our
19
decision to propose such class labeling.
As Kate
20
suggested, within the topic of providing a
21
rationale for the labeling, I would like to
22
emphasize that the neonatal adverse events that we
190
1
will discuss appear to be consistent with either
2
neonatal withdrawal from SSRIs, SNRIs or toxicity,
3 or
perhaps both in some cases.
4 These are the sources of information
that
5 led
us to our decision to propose class labeling.
6
Kate mentioned and detailed the usefulness and
7
limitations of the AERS data system.
Subsequently
8 I
will discuss the benefits and limitations of the
9
other three sources I have listed here.
10
This is some of the verbatim language,
11
proposed language in our precautions section.
12
There are two important points in the first bullet.
13 One
is that all the SSRIs and SNRIs have been
14
implicated or associated with the adverse events at
15 the
time of our analysis. The other major
point
16
under bullet number one is that the adverse events
17 to
be discussed have only been reported in
18
association with third trimester exposure to the
19
drugs, not the first or second trimester.
20
As Kate mentioned, the most severe
21
complications and treatments required have been
22
prolonged hospitalization, admission to special
191
1
care nurseries, respiratory support including
2
ventilation and CPAP, tube feeding as well as use
3 of
anticonvulsants, IV fluids, and in some cases,
4 just a handful of cases, clinicians have
decided to
5 use
antiserotonergic drugs such as thorazine.
6
Also, clinicians have used propranolol.
7
Apparently, they made the claim that there was
8
improvement in the symptoms but it is hard to tell.
9 It
is hard to interpret with those few cases.
10
Whereas symptoms may arise immediately
11
upon delivery, they can also arise anywhere from a
12 day
and a half to five days. It seems that
the
13
most typical time for presentation of these signs
14 or
symptoms is roughly several hours to a day and a
15
half. Beyond a day and a half it
seems to be rare
16
that these events arise.
17
This is a list of the most commonly
18
reported neonatal signs associated with maternal
19 use
of SSRI and SNRI during pregnancy. You
can
20
read those. As Kate suggested
also, we can roughly
21
categorize these in several clusters.
One is
22
feeding difficulty. Another is
respiratory
192
1
distress/autonomic instability.
Also, there are
2
cases in which signs are consistent with
3
temperature instability, as well as abnormal tone,
4
both hypotonia and hypertonia; tremor and
5
jitteriness and the non-specific sign of "constant
6
crying" or "increased crying." Also, sleep
7
disturbance is a very common sign reported in these
8
cases.
9
One of the more important points in the
10
precautions section in pregnancy is that, as I
11
mentioned, the signs reported look to be consistent
12
with either SSRI or SNRI discontinuation symptoms,
13 so
analogous to the adult symptoms, or direct toxic
14
effects of the drugs in question.
15
In the more severe cases these neonatal
16
signs resemble or are consistent with serotonin
17
syndrome, quite a severe form of serotonin
18
toxicity. We refer to that
warning section which
19
contains language about serotonin syndrome.
20
This may be a bit controversial but we
21
have included this based on some evidence,
22
admittedly not on data for controlled studies.
193
1
There is some suggestion that when treating a woman
2
with SSRI or SNRI one might decrease the risk of
3
both SSRI withdrawal and toxicity by carefully
4
tapering the drug roughly 10-14 days before the
5
expected due date and in the case of fluoxetine
6
perhaps abruptly discontinuing the drug at about 14
7
days. That is why we made the
suggestion included
8 in
that dosage and administration section.
9
These are the six terms, all of them have
10
been used and reported for what appear to be
11
somewhat identical syndromes, meaning the signs and
12
symptoms we just discussed that are in the labeling
13 and
that we have been talking about and that Kate
14 has
talked about. I will go over one in
particular
15
because it is probably one of the terms that is
16
least familiar to most of us.
Poor neonatal
17
adaptation is defined by Chambers et al. as
18
tachypnea/respiratory distress, oxygen desaturation
19 upon feeding, hypoglycemia, poor tone, weak or
20
absent cry. That is the extent I
think of the
21
consensus definition of poor neonatal adaptation.
22
Maybe there will be other investigators who will
194
1 use
a slightly different definition but those are
2
typically the signs and symptoms that are included
3
under that definition.
4
The other terms listed have various levels
5 of
definition as far as consensus goes but SRI
6
withdrawal, as Kate mentioned, does have a
7
hypothetical definition as per Chambers et al.
8
which we will discuss subsequently.
We will talk
9
about the other syndromes in a few minutes too.
10 This is Schatzberg. This paper is the
11
result of an expert panel that was convened for two
12
basic reasons: The participants
wanted to decide
13
upon a hypothetical definition of adult SSRI/SNRI
14
discontinuation syndrome, and they also wanted to
15
identify particular symptoms involved, cluster of
16
symptoms. You can see the six
clusters that they
17
have agreed upon. I think in
general it is fair to
18 say
this is well accepted by clinicians and
19 investigators
for definition of adult SSRI/SNRI
20
withdrawal.
21
These are very common reports. It
is
22
quite common for patients to report dizziness upon
195
1 discontinuation
or light-headedness. GI
2
disturbance is quite common, as are reports of
3
flu-like syndromes. It is quite
common for
4
patients to report "electric shock" sensations or
5
"my brain is shorting out" or "my head is shorting
6
out." It is very common also
to have sleep
7
disturbance and neuropsychiatric symptoms that are
8
listed on this slide.
9
One of the main points of the
10
neuropsychiatric symptoms is that, of course, they
11 can
resemble the very disease patients are being
12
treated for but there are also "new" symptoms that
13 are
not identical to a patient's previous symptoms.
14 If
one does have those signs and symptoms, it is
15
more suggestive that this may be a discontinuation
16 or
withdrawal syndrome rather than a recurrence of
17 the
illness. That has practical implications
for
18 how
to treat and interpret the symptoms.
19
This is a brief list, a well-accepted list
20 of
toxicity symptoms in adults. Kate
suggested
21
this too. There are largely CNS
effects,
22
neuromuscular effects and GI disturbance.
196
1
A more severe for of serotonin toxicity is
2
serotonin syndrome. This can be
life-threatening,
3 and
these are the three main clusters that are
4
involved. Notice that the more
severe symptoms
5
include convulsions, disorientation, cognitive
6
impairment, abnormal muscular tone as in the case
7 of
the infants, and serious complications such as
8
autonomic and temperature instability.
One can
9
note the similarities of these symptoms to some of
10 the
neonatal cases reported. Again, those
neonatal
11
cases that overlap with these symptoms are the more
12
severe cases. It is likely that
cases of SSRI and
13
SNRI withdrawal toxicity are probably
14
under-reported, as are many adverse events, and as
15 a
result of the various biases that we see the most
16
severe cases are reported. So,
even though the
17
labeling reports the most severe, we include also
18
some typical symptoms.
Admittedly, we have
19
purposely included the more severe symptoms.
20 You can look at this slide and my
point
21
here is to try to make the case that what
22
investigators are reporting to be neonatal
197
1
withdrawal is, in fact, consistent and analogous to
2
adult withdrawal. As Kate
mentioned, it is quite
3
difficult to elicit symptoms in a neonate. We must
4
rely on signs. But in the ones
that I have listed
5
here there does seem to be an overlap, in the first
6 bullet, between the neonatal withdrawal
syndrome
7
report and the adult withdrawal symptoms. The
8
timing of onset of symptoms is also important to
9
consider and may help us make an interpretation of
10
whether the syndrome is withdrawal versus toxicity.
11
Time to resolution also might help but that is a
12
little more difficult to interpret.
13
This is an analogous slide. here
I am
14
trying to make the case that in some cases neonatal
15
toxicity of SSRIs and SNRIs is consistent with and
16
perhaps identical in some cases to adult toxicity.
17 In
my opinion, there is more overlap in many cases
18
with adult toxicity than with neonatal withdrawal
19
but, again, both cases probably exist.
20
What is especially suggestive are several
21
things, both the quality and severity of the
22
symptoms and the treatment required, as well as the
198
1
immediate onset of symptoms in many cases that Kate
2
referred to as well. The cases
that seem
3
suggestive of toxicity also appear to have a longer
4
duration. The other suggestive
piece of
5
information that is rarely available, but in some
6
case reports clinicians have obtained serum levels
7 of
the drug and the active metabolites to try to
8
correlate symptoms and resolution with drug levels
9 and
the decrease of drug levels. That has
been
10
somewhat successful but, admittedly, it is just a
11
handful of cases and one study which we will review
12
uses that approach.
13
Actually, it is this study. This
is a
14
study by Laine et al. It is a
prospective study,
15 not
randomized, with matched controls. The
point
16 was
to prospectively assess the possible
17
association between SSRI/SNRI use during pregnancy
18 and
subsequent neonatal adverse events that we have
19
discussed, in the short term, meaning 0-4 days,
20
which we will discuss.
21
The subjects included were women who
22
either had depression or panic disorder.
There
199
1
were matched controls who were not receiving these
2
drugs. The two drugs that the
women had been
3
treated with before--in other words, these were not
4
randomized women but had been treated with
5
fluoxetine or citalopram, hopefully, by other
6
clinicians, and were included in the study and they
7
must have been using one of the two drugs
8
throughout pregnancy up until delivery to be
9
included in the study.
10
In yellow I have highlighted two of the
11
important points about the study.
One of the
12
benefits of the study compared to others is that
13 the
investigators used specific outcome
measures
14
that were quite helpful in making an assessment of
15
whether or not the drug exposure was related to the
16
subsequent symptoms. Of course,
they elicited
17
spontaneous adverse events. They
looked closely,
18 in
a serial fashion, at both maternal and neonatal
19
drug levels and active metabolite levels and they
20
looked at monoamine levels, including serotonin as
21
well as their active metabolites.
22
Also quite helpful was their use of the
200
1
specific 7-item assessment looking in particular
2 for
potential signs of toxicity that have been well
3
accepted in adults. Their scale
is based on 2
4
validated scales by authors who had studied
5
serotonin toxicity in adults.
6
Those are the 7 items that they monitored
7
prospectively. They found that
the 3 most common
8
adverse events among the 7 were tremor,
9
restlessness and rigidity. One
important finding
10 was
that in the group treated with SSRIs throughout
11
pregnancy, compared to the control group, had a
12
4-fold increase in serotonergic symptom score and
13
severity during days 1-4, from birth to day 4.
14
They also compared groups at day 14 and day 28 but
15 did
not find a significant difference at those 2
16
points.
17
The mean neonatal drug levels
were in the
18
usual adult range, the "normal" range of adults.
19
There may have been a few in the abnormally high
20
range but generally the levels were within the
21
normal adult range. They also
reported that
22
symptom resolution correlated with decreasing serum
201
1
SSRI drug level.
2
Another interesting finding was that the
3
SSRI group had a mean lower cord 5-HIAA which is a
4
metabolite of serotonin and purportedly suggests a
5
higher serum of CSN serotonin activity and the
6
serotonergic symptom score correlated inversely
7
with that measure.
8
In this slide I want to make the point
9
that if one looks at the green cubes, they
10
represent the numerous factors that are involved in
11
pregnancy, in the normal physiology of pregnancy as
12
well as perturbations of the physiology of
13
pregnancy. I am referring to
drugs such as SSRIs,
14
SNRIs, other psychotropic drugs, drugs such as
15
alcohol and other drugs of misuse, vitamins,
16
nutrients--all those obviously have an effect on
17 the
outcome of pregnancy and we must consider the
18
numerous variables when trying to interpret these
19
neonatal adverse events that we are talking about.
20
In the cases that Kate has discussed and
21
that I am referring to, I think it is fair to say
22
that the majority of the cases had confounding
202
1
variables, either depression itself--and that is
2 one
of the most important points to focus
3
on--depression itself clearly has associated
4
adverse events that are similar to the adverse
5
events that we may attribute to SSRI/SNRI
6
withdrawal or discontinuation.
For example, babies
7
born to mothers who are not treated for depression
8 but
who are clearly depressed can have jitteriness,
9 low
birth weight. They are described as
being hard
10 to
soothe frequently. So, the signs do
overlap
11
with the symptoms we are talking about in relation
12 to
SSRI exposure. Also, there are clearly
numerous
13 factors that we don't know of in the case
reports,
14
which are limited in the AERS system.
15
Another important point in trying to sort
16 out
to what extent other drugs in this class are
17
similar or different is that we really don't know
18
what the denominator is. We don't
know what the
19
actual quantitative use of these drugs in pregnancy
20
is. We also don't know the
background rates of the
21
adverse events in pregnancy or other conditions.
22 So,
these are huge problems in making certain
203
1
determinations. I mentioned the
limitations of the
2
data.
3
Another difficulty in interpreting these
4
neonatal adverse events is, of course, the limited
5
repertoire of neonatal behaviors.
We can't elicit
6
symptoms per se and the signs that they exhibit are
7
within a fairly tight range so it makes it more
8
difficult, of course, than making an interpretation
9 in
adults with adult adverse events.
10
One of the other problems with
11
interpreting whether or not, for example, these
12
drugs have a causal relationship to neonatal
13
adverse events, either the drug effect or the
14
discontinuation, is that many of the SSRI/SNRI
15
symptoms of neonatal withdrawal or toxicity have an
16
overlap. The very symptoms that
are reported for
17
withdrawal such as jitteriness, for example, or
18
tremor or increased tone are also reported for
19
purported neonatal toxicity.
20
Despite the uncertainty that we discussed,
21 we
feel that there is a strong association between
22 the
use of these drugs with neonatal adverse events
204
1
that these should be listed in labeling.
Again, we
2
emphasize that this is only associated with third
3
trimester use of these drugs.
4
To repeat one of the points, we feel that
5 the
adverse events can be consistent with the
6
SSRI/SNRI withdrawal or toxicity and perhaps both
7 in
an individual case. In fact, there were
several
8
cases that were suggestive of a neonate having
9
toxicity several days or perhaps a week later going
10
through withdrawal so that is theoretically
11
possible.
12
Several other reasons for deciding to
13
place this language in labeling is that, at least
14 in
the cases reported, many of the neonates
15
required serious specialized care such as
16
hospitalization, ventilation, etc., the types of
17
treatments we have mentioned. Of
course, because
18 of
this, clinicians need to be aware of the
19
potential for development of these adverse events
20 in
neonates who had been exposed in utero to these
21
drugs. It may be possible, and
Dr. Wisner may
22
discuss this, that there may be prevention
205
1
strategies that are practical and effective. We
2
also need to consider diagnosis, meaning, making
3
differential diagnosis between withdrawal and
4
toxicity, or neither. Of course,
in many cases
5
these symptoms may have nothing to do with the
6
drug. We can't make definitive
attributions.
7
On the last slide I want to emphasize
8
that, of course, it is very important to treat
9
depression during pregnancy.
There is extreme
10
morbidity of depression and everything that applies
11 to
a man or woman, pregnancy or not, in depression
12
applies to women during pregnancy--suicide, severe
13
dysfunction, social dysfunction, poor weight gain,
14
malnutrition which, of course, impacts the
15
development of the neonate.
16
In contrast to previous years during which
17
many authors reported that pregnancy "protected"
18
women against mood disorders or recurrence of mood
19
disorders, it is becoming more clear that the
20
prevalence of depression during pregnancy is quite
21
high, as high as 10-16 percent.
With more
22
information that is available as time goes on,
206
1
physicians and the patient can weigh potential
2
risks and benefits to the mother and neonate when
3
deciding whether to treat depression or not or
4
other psychiatric symptoms.
5
That is another important point, that we
6 are
not just talking about depression. These
7
drugs, of course, are used for anxiety disorders
8
such as panic disorder and PTSD and
9
obsessive-compulsive disorder so it is a larger
10
population than I was actually referring to.
11
Also, it is possible that the clinician
12
might reduce the risk of neonatal exposure to these
13
drugs by tapering near term and that they might
14
reduce the risk of recurrence of depression or
15
postpartum depression by promptly restarting the
16
drug in some cases upon delivery in the delivery
17
room. That is one potential
strategy. Of course,
18 we
do not have a consensus about interpretation and
19
management of these complicated neonatal adverse
20
events and ideally controlled trials are needed in
21
this important field.
22
The last point--of course, it is hard for
207
1
many of us to imagine, including myself, that one
2 can
conduct truly randomized, controlled studies in
3
pregnant women but it is possible to consider the
4
ethics of not treating, not knowing what is
5
happening in these studies. It
would be
6
interesting to see what might happen, particularly
7
whether or not investigators might be able to
8
conduct randomized, controlled trials.
9
Finally--I thought the last slide was
10
final; this is the last slide and I want to point
11 out
the status of the proposed class labeling, and
12
this is for both labeling in pregnancy and labeling
13 in
precautions in adults and for dosage
14
administration. Firstly, all the
drugs have
15
incorporated the proposed labeling.
Those listed
16 in
the first bullet have included the language.
17 The
sponsor of fluoxetine has verbally accepted the
18
class labeling and currently our Division is in
19
discussion with the sponsor of sertaline about
20
whether or not they will consider incorporating the
21
class labeling. Thank you very
much.
22
DR. CHESNEY: Thank you. Any technical
208
1
questions? Dr. Gorman?
2
DR. GORMAN: In accepting this
class
3
labeling, do the sponsors have the opportunity to
4
modify it in any way or is it a whole or none, up
5 and
down quote?
6
DR. LEVIN: They have the chance
to
7
attempt to do so.
8 [Laughter]
9
No, seriously, we had discussions about
10
that. Of course, as you might
guess, especially
11
with drugs that have a longer half-life, companies
12
might argue that qualitatively and quantitatively
13 these
adverse events are different but, in fact,
14
that is probably not true from the data available.
15 So,
for practical reasons, probably each company
16 did
request making modifications but in the end
17
they accepted the verbatim language that we
18
proposed.
19
DR. CHESNEY: Dr. Ebert?
20
DR. EBERT: I hope this is a
technical
21
question, but does the AERS database enable the FDA
22 to
do any long-term follow-up on these children?
209
1 You
have the immediate postnatal adverse events but
2 are
you able to follow-up these individuals two or
3
three years later to identify the long-term
4
effects?
5
DR. LEVIN: I think one answer is that it
6 is
extremely difficult based on the fact that these
7 are
spontaneous reports and voluntary reports.
It
8
would be great if we had that. It
is very hard
9
under the current system. There are
companies,
10
maybe one company I can think of that is
11
prospectively monitoring women who are using an
12
antidepressant during pregnancy.
That seems to be
13 a
more productive strategy. At this point,
14
although what you are suggesting would be ideal, I
15 am
not really sure to what extent one can request
16
further follow-up unless there are serious adverse
17
events. If it is a serious
adverse event, defined
18 by
regulatory language, then the companies are
19
obliged to give follow-up reports.
But the typical
20
reports describe, as Kate mentioned, the type of
21
symptoms and signs, the timing of onset, a few of
22 the
obstetric factors and co-morbidities, some
210
1
concomitant meds, but my recollection is that it is
2
fairly rare for those reports to have included the
3
duration of the adverse event or the time to
4
resolution.
5
DR. D. MURPHY: Just to reinforce
that,
6
AERS is not set up for long term.
Also, it would
7 be
difficult to sort of imagine how someone would
8
make that connection later on to a therapy given
9
earlier so you really would need to set up some
10
sort of prospective study.
11
DR. CHESNEY: Dr. O'Fallon?
12
DR. O'FALLON: It seems to me it
is
13
crucial to be able to distinguish between
14
withdrawal or discontinuation versus toxicity
15
because you have to treat them totally differently.
16
Correct?
17
DR. LEVIN: Right, yes.
18
DR. O'FALLON: So, I am looking at
your
19
list and I don't see how you could possibly, just
20 by
looking at these descriptions, tell. Is
there
21 any
way you can? Here is a person who has
this
22
problem, can you distinguish which one it is? Is
211
1
there something you can do?
2
DR. LEVIN: Yes, you are
right. Exactly.
3
That is one of the major points.
It is extremely
4
difficult in some cases primarily because of the
5
relative lack of information as you are saying,
6
also there is clearly an overlap in the wording for
7 withdrawal and toxicity.
8
DR. O'FALLON: Yes.
9
DR. LEVIN: In my mind, and of
course I
10
acknowledge that people can disagree completely,
11 but
I think it is the severity of the symptoms.
12
DR. O'FALLON: The severity?
13
DR. LEVIN: The severity is one
point.
14 One
reason I mention that is it is comparing and
15
contrasting to adult syndromes.
Typically, in the
16
adult syndromes with withdrawal they can be quite
17
distressing. In adults they are
usually mild to
18
moderate and transient but in some cases they can
19 be
quite distressing and temporally disabling,
20
meaning, people are not be able to take care of
21
their families for days or miss work for several
22
days. But it is quite rare. Usually they are mild
212
1 and
transient. Most likely there are
neonatal
2
cases that are withdrawal that aren't reported. In
3
personal communications clinicians have suggested
4
that the most common scenario if neonates have
5
these type of symptoms, they have things such as
6
feeding difficulty and increased crying which
7
doesn't require specialized care and resolves
8
fairly quickly.
9
But, yes, you made several important
10
points. There is an overlap in
the symptoms.
11
Another way to answer your question is that I think
12
getting serial drug levels would be very helpful.
13 It
has been done in several cases. I think
it is
14 one
of the most important pieces of information
15
given the confusion and uncertainty about these
16
symptoms.
17
People have also given sort of treatment/
18
diagnosis. In other words, I
remember only two or
19
three cases in which a clinician decided or thought
20 it
was probably withdrawal syndrome and they gave
21 the
neonate the very drug that they may have been
22
withdrawing from. I remember two
cases. In one
213
1
case they reported that the infant became better, I
2
don't know in what time frame. In
the other case
3 it
got worse. The symptoms were exacerbated.
4
There was a handful, three cases in
5
which--this is interesting actually, there were
6
three cases in which the clinician clearly
7
diagnosed the infant with having withdrawal
8
syndrome and he decided to give the drug thorazine
9
which is known to have antiserotonergic properties.
10
Even though we can't make attribution, the
11
chronology was such that within minutes to hours
12 the
infant was "remarkably" better.
We don't
13
really know what that means but it is interesting
14
that he chose to use the drug while still using the
15
term withdrawal. Beta blockers
also may be
16
helpful.
17
DR. O'FALLON: It just seems to me
that we
18
can't even deal with this very well until we have a
19
good idea of which problem it is.
20
DR. LEVIN: Exactly. That is true.
21
DR. O'FALLON: Do you think you
will have
22 an
opportunity to explore that further as we get to
214
1 the
questions? It is an issue.
2
DR. CHESNEY: Dr. Danford has a
technical
3
question.
4
DR. DANFORD: Well, I wonder if
there are
5
observable fetal effects that we ought to be
6
looking for to help make this distinction. Is
7
there an impact on the baby's biophysical profile?
8 Is
there observable jitteriness, abnormal
9
movements, that sort of thing that, if we just were
10 to
look in an organized fashion for those among
11
fetuses of pregnant ladies on these medicines we
12
would either find them or not--
13
DR. LEVIN: Right.
14
DR. DANFORD: --and were we to
find them,
15 we
would think that toxicity might be in effect.
16
And, if were to find them only after delivery
17
perhaps that would be withdrawal.
18
DR. LEVIN: Right. Exactly.
That is why
19 we
are considering that and perhaps beginning
20
studies to look at that with ultrasound, especially
21
with ultrasound, to look for potential
22
abnormalities of movement. I
haven't read anything
215
1 as
far as results. There may be some, I
just don't
2
recall seeing results of any studies, even
3
preliminary studies looking at that but that is a
4
critical question to ask and to answer.
It would
5 be
extremely helpful. That would be an
excellent
6
piece of information to have in sorting out whether
7
this might be toxicity or withdrawal.
8
DR. CHESNEY: I think those will
all come
9 up
when we try to answer the questions. I
guess
10 you
are going to introduce Dr. Wisner.
11
DR. LEVIN: Yes, I would like to
introduce
12 Dr.
Wisner. It is my pleasure to introduce
her and
13 I
am very glad that she is here. Dr.
Wisner is the
14
Director of the Women's Behavioral Health CARE, a
15
specialized treatment research program for
16
childbearing women at the University of Pittsburgh.
17
Parenthetically, I was a resident in psychiatry
18
there and had the great pleasure and privilege to
19
learn from Dr. Wisner so that is another reason why
20 I
am especially happy to see here. Dr.
Wisner
21
conducts several NIMH-funded studies involving
22
pregnant women and postpartum women with mood
216
1
disorders. She is trained in a
number of fields in
2
adult and child psychiatry, as well as pediatrics.
3 She
has done a postdoctoral program in
4
epidemiology. She has academic
appointments in
5
psychiatry, obstetrics, gynecology and
6
epidemiology.
7
DR. WISNER: Thank you, Bob, for
that very
8
nice introduction, and it is a great pleasure to be
9
here and I thank you for the invitation.
Again, it
10 is
a real pleasure to be here and I am thankful for
11 the
opportunity to address you.
12
I have several goals for the talk this
13
afternoon. The first is to
discuss an approach to
14
making treatment choices for pregnant women who are
15
depressed. The second is to think
about how to
16
conceptualize the diagnosis of the effects that we
17
have been talking about. In other
words, how do we
18
think about whether what the neonate is
19
experiencing is acute side effects or what has been
20
called toxicity or, in fact, is a withdrawal
21
syndrome from the same medications?
Finally, I
22
would like to tell you about a study that I am
217
1
doing now, an RO1 that is funded by NIMH in which
2 we
are actually trying to address some of these
3
issues.
4
Well, how big of a problem is this, that
5 is,
depression and other disorders that require
6
treatment with medications during pregnancy? In
7
fact, it is a major public health problem. Bob had
8 a
rate of about 10-16 percent of women who
9
experience depression in pregnancy.
In fact, that
10
fits with the kind of rates that we see in
11
childbearing age women. If we
look at the rate of
12
depression across ages in women compared to men,
13
about twice as many women have depression during
14
their childbearing years as do women [sic] and, in
15
fact, somewhat unfortunately, it is right in the
16
childbearing age time that women experience this
17
devastating illness.
18
Given that many women are going to have
19
this disorder during their childbearing years, how
20 do
you deal with the fact that at least the
21
pharmacologic therapy is a chronic treatment for
22
this illness and, in fact, women want to conceive
218
1
while they take this medication?
Our American
2
Psychiatric Association put together a committee to
3
look at these issues several years ago and a number
4 of
papers resulted. One of them is
referenced in
5
which we defined what kinds of issues docs would
6
need to think about in talking to women who are
7
contemplating pregnancy if they are depressed or
8
they are already taking an antidepressant
9
medication.
10
This is a somewhat complicated slide but I
11 am
just going to break it down into components.
12 The
first area is what are the responsibilities of
13 the
physician. Of course, talking to
patients
14
about what depression is is incredibly important
15
because many patients feel like the depression is
16
like having a bad day, or they have a lay person
17
definition and, unfortunately, the word depression
18 is
used colloquially--"I had a fight with my boss;
19 I'm
depressed." This major depression
that we are
20
talking about is a clinical diagnosis called major
21
depression and it is a dysregulation illness in
22
which the physiologic functions of the patient are
219
1
affected. We will talk a bit more
about that a
2
little bit later.
3
But just the criteria of this disorder are
4
very weird. Here is a medical
illness where
5
dysregulation of mood, ability to enjoy life--those
6
things are affected but the dysregulation is
7
confounded. You can either have
too little sleep,
8 for
some patients an hour of sleep a night; or some
9
patients sleep 23 hours. Those
are both
10
dysregulated sleep that count as part of the
11
diagnosis. Another example is
agitation. You can
12
have excess motor activity and not be able to sit
13
down, be very agitated, or be so slowed down you
14 can
barely move. Again, it is indicative of
not
15
just something that is "I feel sad emotion" but
16
this is a whole body dysregulation illness.
17
In talking to patients, what I typically
18 do
is discuss what treatments are available for
19
major depression, and there are many.
Then I talk
20 to
her about what specifically might be appropriate
21
with respect to her clinical history and then,
22
secondly, how that might be modified because she is
220
1
either pregnant or she wants to become pregnant.
2 We
have many options for depression in pregnancy in
3
terms of treatment. Many patients
are already
4
taking effective medications.
Psychotherapy has
5
certainly been studied as a treatment for
6
depression in pregnancy. Due to
some of the issues
7 we
are talking about in this very meeting, my group
8 has
begun to pilot light therapy for treatment of
9
depression in pregnancy. We also
talk to patients
10
about the risks of no treatment during pregnancy,
11
which I think is a very poor option.
12
We are also then obligated to talk to
13
patients about what are the outcomes if she accepts
14 a
particular form of treatment during pregnancy,
15 and
what are the outcomes for her depression if she
16
doesn't accept treatment or wants to consider
17 moving
to a different treatment which may or may
18 not
be effective for her. Our discussion
today
19
really focuses on this final area of neonatal
20
toxicity. In this paper we meant
to designate the
21
kind of broad construct that Bob talked about.
22
That is, negative symptoms that occur in the
221
1
post-birth period for those neonates.
2
But there is another issue here that I
3
don't want to exclude from the discussion, and that
4 is
the idea of behavioral teratogenicity.
That is,
5 of
course, the idea that these potent central
6
nervous system acting agents when brain, as
7
vulnerable as the neonatal brain, is exposed
8
through pregnancy and perhaps there may be effects
9
that occur that manifest later on in life. That is
10
often talked about as later on in life, like way
11
down the line. A hypothetical
example might be
12
that a child might be at higher risk for learning
13
disabilities as a school age child.
But there is a
14
very real question of when behavioral
15
teratogenicity occurs, meaning that there is no
16
time point so that some of the effects that we see
17 may
be really due to this particular kind of
18
mechanism as opposed to either withdrawal or acute
19
side effects or toxicity. That is
another issue
20
that hasn't been explored. I
think that that
21
question is inherent in some of the questions here
22
which are how long does this thing, whatever we
222
1
call it, last. Because it will
help us define the
2
mechanism.
3
Treatment of depression in pregnancy is
4
important. The outcomes for
untreated depressed
5
women in pregnancy are not good.
Unfortunately,
6
every paper that has been cited in this meeting
7
today has not uncoupled the occurrence of the
8
illness, that is depression, from the drugs used to
9
treat it. That is like saying we
want to study a
10
hypoglycemic agent as an exposure in pregnancy but
11 we
are not going to control or look at the blood
12
sugars of the pregnant women, and that is our major
13
problem with this field.
14
An interesting area is what kinds of
15
treatments do women select in pregnancy.
There is
16 a
common belief that because women are pregnant
17
they might want psychotherapy or light therapy but,
18 in
fact, in my research program many of the women,
19
particularly those who get very good responses from
20
antidepressants, are very interested in continuing
21
those medications in pregnancy.
22
These large blocks are just there to show
223
1
that the decision is really a dynamic one and the
2
choice of providing, say, a medication treatment in
3
pregnancy means that we have decided that the
4
benefit of that is greater than the risk for that
5
patient. But if, in 4-6 weeks,
that medication
6
does not produce an antidepressant effect or
7
sustain an antidepressant effect, then that
8
decision-making process has to be reconsidered.
9
This committee that I spoke about, the
10
American Psychiatric Association committee, wrote
11
this first paper in which we reviewed the
12
prospective data for antidepressant use in
13
pregnancy. Although I am not
going to go into that
14 in
detail because I want to focus on poor neonatal
15
adaptation and neonatal effects that are the topic
16 of
this meeting, one issue that I think is
17
important is that because these agents are not
18
major morphological teratogens there has been over
19 the
last several years a relative comfort about
20
their use in pregnancy. So, there
is a much larger
21
population of mothers being exposed to these
22
agents, and I think we are seeing these kinds of
224
1
outcomes, like neonatal toxicity, that are becoming
2
more frequent, in fact, because of the increased
3
use.
4
Tina Chambers' article which Bob
5
mentioned, I think is a very important article
6
because the agent studied was fluoxetine. About a
7
third of patients have this poor neonatal
8
adaptation and 24 percent of her patients were
9
admitted to special care nurseries.
Because that
10 is
a prospective study specifically of fluoxetine
11 at
least it gives us a rate in which the
12
denominator is known.
13
Well, I want to focus on treating maternal
14
depression and the importance of uncoupling that
15
factor in these data sets because there are papers
16
that show that maternal depression and anxiety
17
increase the odds ratio or the risk of multiple bad
18
things in pregnancy, like preeclampsia, and also
19
that there are investigators, particularly in
20
England, who have looked at uterine artery
21
resistance in the face of depression and anxiety.
22
These factors have been related to growth
225
1
restriction in fetuses as well as preeclampsia.
2 So,
again, depression itself can create negative
3
outcomes, and how to uncouple the disease-produced
4
negative effects from the medication is incredibly
5 important.
6
We also know that maternal stress and
7
certain anxiety disorders and mood disorders result
8 in
dysregulation of the HPA access and that, in
9
fact, that has ramifications for the fetus as well
10 and
has effects on fetal ability to respond to
11
stress.
12
The question was asked before about
13
ultrasound and in utero behavioral studies of
14
fetuses to look at this issue. In
fact, I have an
15 MT
at Brown, named Amy Salisbury who is working
16
with me and Gianne DePitro, whom you know, who has
17
done these in utero studies. They
are doing
18
parallel studies of fetuses with the same three
19
groups that I will talk about in my study. So,
20
those kinds of investigations are being performed
21
right now.
22
We also know that even before an infant is
226
1
born to a depressed mom it interacts with that
2
depressed mom. Those infants have
been seen by
3
nursery care staff to be irritable, difficult to
4
console. So, these same kinds of
behavioral
5
effects that we have been talking about as due to
6
medication also occur because infants are born to
7
moms who have this dysregulation disorder we call
8
depression. Again about
depression, the point that
9 I
want to emphasize is that depression is this
10
physiological dysregulation but it really is
11
probably a variable. That is, the
presence of
12
depression that really brings with it a whole
13
multitude of factors that contribute to poor
14
outcomes for pregnancy if it is left untreated.
15
We talked about appetite changes and food
16
choice changes that occur.
Certainly, the ability
17 to
comply with prenatal plans, such as vitamins and
18
other prescribed treatments in pregnancy are less.
19
Women can be irritable. They can
have isolation
20 and
alienation of psychosocial relationships right
21 at
a time when it is natural for families to begin
22 to
think about being parents, to begin to bond
227
1
together. Many women who are
depressed also use
2
other drugs and smoke, which create confounds, and
3
many women who are depressed elect not to breast
4
feed which then deprives the infant of that
5
particular favored choice of feeding.
6
This is a model from Dawn Misra that I
7
like to use to think about this whole group of
8
factors that relate to outcomes for mothers and
9
babies no matter what the disorder is.
The way she
10
conceptualized it is very relevant to this
11
discussion in that depression is an illness with
12
genetic factors. It runs in
families, like most
13
disorders. Physical environments
affect it,
14
including light. Where you live
and the amount of
15
light affects your risk for depression, and social
16 environments affect it. So, if you are in a
17
wonderfully comfortable neighborhood versus a
18
neighborhood in which there are drive-by shootings
19
every day that makes a big difference.
Those are
20
factors that are more distal risk factors in terms
21 of
distal from the pregnancy, but they shape the
22
biological and behavioral responses that the woman
228
1
brings to conception. So, those
factors which
2
increase the risk for depression are what she
3
brings along with her to the pregnancy.
Again, you
4
have the physiologic dysregulation, the HPA access
5
dysregulation and other difficulties that she
6
brings to the pregnancy state with her.
7
What we are trying to do as healthcare
8
professionals at these intervention points is say
9 all
right, we know there are these whole groups of
10
variables that come with a mother who has major
11
depression. How can we deal with
that so we
12
maximize the outcome for both the mother and the
13
baby? And, how can we do that not
only in a
14
short-term way but in a long-term way?
How do we
15 get
the best result? Because we know that we
would
16
rather not use pharmacotherapies for these
17
depressed women but leaving them untreated is not
18
particularly good either. I think
this particular
19 mom
sums it up the very best when she says,
20
"believe me, mommy's mood stabilizing drugs are not
21
something that anybody wants mommy to just say no
22
to."
229
1
[Laughter]
2
I have many patients who really feel like
3 they
are in this particular situation.
4
Well, let's look at some specific issues
5
related to use of SSRIs during pregnancy,
6
especially the final part of pregnancy, and the
7
risk of neonatal complications.
We have talked
8
about several papers, especially Laine's paper,
9
which have shown this increase in the risk of
10
difficulties in the neonatal period related to
11
fluoxetine. Now, fluoxetine is
unusual among the
12 SRI
medications in that it has an incredibly long
13
half-life and it has a metabolite that is equally
14
active with an even longer half-life.
So, it is
15
distinct in that pharmacologic way which may make
16 it
distinct in the way it behaves in neonates as
17
well.
18
Paroxetine, or Paxil, has been most
19
commonly identified in case reports, but there is a
20
recent article in one of the pediatrics journals in
21
which the investigators sought to replicate the
22
finding that paroxetine was the SSRI that was
230
1
particularly problematic and was unable to do so.
2
Paroxetine is unusual as well in that it is not
3
only a serotonergic antidepressant, it is the only
4 one
that has significant anticholinergic effects as
5
well so that one could imagine having cholinergic
6
overdrive in addition to the serotonergic mediated
7
effects in newborns. And, we have
less data on the
8
other three agents, sertaline, citalopram and
9
fluvoxamine, so we sort of make inferences based on
10 the
pharmacology of those agents.
11
One issue is certainly placental passage.
12
Vicky Hendrick has looked at this particular
13 problem and shown that these agents have lower
14
placental passage, and these agents have higher.
15 So,
one would expect that agents with greater
16
access to the fetal compartment might have more
17
effects as well. Again, there is
a look to could
18 we
think about, or is there enough evidence to
19
suggest that certain agents might present less
20
distribution into the fetal compartment and,
21
therefore, might be less problematic.
So, that is
22
again another area of investigation.
231
1
The other issue is the variability in
2
fetuses and moms in general so that why some kids
3
have major difficulties and other kids don't
4
becomes probably related to individual variability
5
differences. So, one of my
concerns is not so much
6
about the full-term babies lately, but we have had
7
some premature babies born where you have all the
8
sequelae of prematurity in addition to a
9
significant amount of drug on board in those
10
patients.
11
There are also additional exposures that
12
might complicate the baby's ability to metabolize
13 the
drugs with which it is born. The overall
14
health and nutrition of the newborn are major
15
factors. There are also genetic
issues so that we
16 all
know that our and our baby's ability to
17
metabolize drugs really depends on the ability of
18
hepatic enzymes to metabolize them and there are
19
poor metabolizers as well as rapid metabolizers
20
distributed in the population.
The activity of the
21
particular enzymes within the fetus and within the
22
parent are important as well.
Some people can
232
1
break down serotonin rapidly; others really can't.
2
Then, there is the final issue of how available are
3
serotonin precursors. So, there
are a number of
4
factors that really go into this decision about
5
what is really happening in the neonatal period and
6 how
do we understand it.
7
This point has been made by Bob very well.
8 If
we look at poor neonatal adaptation defined by
9
Tina Chambers and Carey Laine's paper which looks
10 at
serotonin over-stimulation, and you say, well,
11
these are the symptoms here, these are the symptoms
12
there, and these are the overlapping symptoms, you
13 are
really struck with the sense that we have
14
different groups defining different things, and how
15 we
can really put them together is somewhat
16
problematic.
17
Bob mentioned something that I think is
18
very important, and that is that Carey Laine's
19
paper really suggests that the babies born to women
20 who
take antidepressants through the final
21
trimester, if they are going to experience
22
something at birth based on those high serum
233
1
levels, the suggestion in that paper is that it
2
truly is serotonin effects. So,
what this group is
3
saying is that essentially there are side effects
4 and
in the severest form you have serotonin
5
syndrome. Those babies are
essentially born with
6 an
adult level of the drug on board so these are
7
acute side effects and, in fact, maybe they have
8
those in utero, and we will find out with some of
9 the
fetal studies.
10
As the cord is cut though, the
source of
11 the
drug is not there so there is an abrupt
12
discontinuation but since babies don't metabolize
13
these drugs particularly well there is a rate at
14
which those drugs come down in the baby's body and
15 it
is possible, in fact, to have both these acute
16
effects. It is possible to have
what are really
17
more consistent with withdrawal effects down the
18
line. The other possibility we
have entertained
19
with paroxetine is that, depending on the receptor
20
occupation, it may be possible to have one or both
21 of
those syndromes related to cholinergic receptors
22
versus serotonin receptors. So,
it is a very
234
1
complicated picture.
2
Bob and others have already talked about
3
these domains of symptoms that are really affected
4 in
neonates exposed to SSRIs. The other
point that
5 I
think is interesting is the repertoire of babies
6 to
tell us that they are not particularly
7
comfortable. In fact, there are
likely to be
8
symptoms and signs that overlap across time that
9 can
be indicative of either of those. So,
the time
10
course is particularly critical.
11
So, the questions that I think we need to
12
understand are what are the symptoms that
13
characterize these syndromes?
What is the
14
incidence? Because we really
don't know that. The
15
data from Tina Chambers' paper is probably the best
16 and
it is for one agent. Is this withdrawal
or
17
intoxication or some form of neurobehavioral
18
teratology? Are they all equally
likely to cause
19
it? The answer to that I think we
can say pretty
20
confidently is no. Because the
risk of not
21
treating depression in women typically outweigh the
22
risks, what can we do to prevent these things,
235
1
minimize them or even treat them so that when we
2 get
back to that model I showed you we are really
3
maximizing the short- and long-term outcomes for
4
moms and babies?
5
What our group has done is to take Loretta
6
Finnigan's wonderful scale that was designed to
7
look at withdrawal from drugs of abuse, and we
8
integrated these symptoms and signs that have been
9
described here into the scale to try to understand
10
what is happening. I will show
you that scale in a
11
minute.
12
In this particular investigation that I
13
have under way now what we are doing is picking up
14
women before week 20 of pregnancy and studying the
15
moms and their babies out to month 24 postpartum.
16 We
have exposures week by week in this study.
So,
17
originally they come in; they have 20 weeks
18
retrospective exposure history, and by exposure I
19
mean drug, not only SSRI but anything else they
20
have taken, and we think of depression as an
21
absolutely separate exposure. So,
on the exposure
22
chart you will have criteria for major depression
236
1 or
depression scores because we get continuous
2
measures as well.
3
We are after three groups, although we
4
have ended up with a fourth group here as well.
5 The
three groups are pregnant women with depression
6 who
refuse medication. You cannot say they
can't
7
have other therapies and many of our patients do
8 but
many of the women don't respond. So, it
is
9
positive depression, no drug group.
The second
10
group is women who are not depressed because
11
probably they are taking an antidepressant so it is
12
negative depression, positive drug.
We have a
13
normal control group. And, not in
the original
14
design but certainly as a part of life, we have
15
patients who are partial responders.
You know,
16
they are a little bit better but they are exposed
17 to
both the drug and some level of depression as
18
well. We are studying these four
groups.
19
At week 36 of gestation, what I do is I
20
talk to them about a choice they have, whether they
21 are
going to continue the drug right through the
22 end
of pregnancy, or whether they would like to
237
1
consider tapering the drug 2 weeks before the EDC.
2 If
it is fluoxetine we just discontinue it, again,
3
because of the long half-life.
4
At this point I can tell you that over
5
half, probably close to two-thirds of the women we
6
offer this option and give a careful risk/benefit
7
discussion to elect to stay on their medication
8
through the end of pregnancy, and their reasoning
9
typically is every time I go down on the dose or I
10
stop this medication I get sick very quickly and I
11
don't want to go into labor and delivery like that.
12 The
women who can say, gee, it is a couple of
13
months off and when I taper my drug or discontinue
14 it
before I get symptomatic again--those women are
15
willing to do this alternative strategy but it has
16
been very intriguing to see under what
17
circumstances they choose this strategy.
18
What we then do is monitor weekly with a
19
continuous depression measure through whenever they
20
give birth, and we are looking at a number of
21
outcomes at birth and at two weeks and beyond to
22
compare across the two groups.
238
1
Here are some questions that we are
2
struggling with. Does this taper
regimen--let's
3 say
they decide to go off the drug in the latter
4
part of pregnancy--we don't abruptly discontinue
5 it,
we taper down--does that affect the near-term
6
fetus in utero? We make the
assumption that if we
7 are
withdrawing the drug, that slow withdrawal is
8
better than the abrupt discontinuation of cutting
9 the
cord at birth, but the kinds of studies that
10
were mentioned about fetal well being are critical
11 in
that context.
12 Does the baby or infants who are born
of
13
mothers who taper their drug in the latter part of
14
pregnancy compare to unexposed moms?
I mean, does
15 it
really work?
16
Do mothers become symptomatic during the
17 taper
phase? By and large, they don't and that
18
probably has to do with the fact that they are
19
choosing based on their history of how long it took
20
them to get sick and, you know, it takes a while
21
before women respond to antidepressants.
It takes
22 2-4
weeks. We are trying to take advantage
of that
239
1
time frame with withdrawing the medication to try
2 to
get more of the drug out of the fetal
3
compartment before the baby is born.
Does that
4
work? We are finding out.
5
Does restarting the medication at birth
6
prevent recurrence of the episode?
I can tell you
7
that, by and large, it does. The
baby comes out;
8 mom
goes back to her room; the drug goes right
9
in--you know, no delay in getting the drug in.
10
An intriguing question is that small
11
amounts of all these drugs occur in breast milk.
12
Does breast feeding provide some partial protection
13
against at least the component that we think may be
14
withdrawal?
15
In our study the raters are totally blind
16 to
not only the status of the baby in terms of
17
exposure but to the study hypotheses.
So, the
18
raters for the birth assessments and 2-week
19
assessments are totally blind. We
do maternal
20
serum and cord blood antidepressant levels. We
21
also do cortisol and other hormone levels as well.
22 We
do a mother and baby breast feeding level at
240
1
week 3. We do cry analysis at
birth and 2 weeks; a
2
pediatric neuro exam at 2 weeks.
It is an exam
3
that was given to us by Lynne Singer who works with
4
addicted moms in Cleveland; and we do the modified
5
Finnigan scale.
6
I had a heck of a time trying to figure
7 out
how to put this document on Power Point but I
8
finally figured it out last night.
Essentially,
9
what we have done with Dr. Finnigan's scale is to
10 say
here are the items for her scale that we don't
11
think are relevant to these syndromes.
Here are
12 the
items that seem to overlap with her particular
13
scale. As you look down, they are
pretty much the
14
symptoms that we have been talking about. These
15 are
additions that didn't occur in her scale that
16 we
wanted to assess. So, this scale is now
17
integrated into our study that I told you about.
18 In
fact, Amy Salisbury, at Brown, who is doing the
19
other study I told you about, is having these as
20
well.
21
Well, the point has already been made that
22 we
really have to understand how to diagnose this
241
1
because the treatments are exactly opposite. If
2 you
think, because of the high level that a baby
3
might have a birth that it is a serotonin toxicity
4
that is side effects, which is what I think the
5
majority of these kids have, when it is an early
6
presentation, then it is toxicity and what might
7 you
do?
8
Our main interventions have been parental
9
education and cognitive strategies.
These are moms
10 who
are very prone to feel guilty. You know,
"what
11 did
I do to my baby?" So, we really do
a kind of a
12
therapy to help them understand what is happening
13 and
that it is transient. Certainly, the
strategy
14 I
mentioned in terms of an attempt to taper in the
15
final part of pregnancy is an option, but we tend
16 to
be very conservative and we have a very good
17
pediatrician that talks to the moms about kangaroo
18
care and swaddling, and do more behavioral
19
management strategies.
20
What if, though, a baby had very severe
21
symptoms? In fact, the case that
was described by
22
Manna et al., which is one of the
first ones in
242
1
Cleveland, was actually a baby born to one of the
2
moms that I treated and that baby was really quite
3
ill. Might we think about an
antiserotonergic drug
4
like cyproheptadine? There you
always get into the
5
issue of we don't know what kind of dose to use so
6
some sort of dose-ranging safety and efficacy study
7
would be appropriate. What if we
give too much?
8 Do
we then give back the agent? Those kinds
of
9
studies are important to think about and we have
10
begun working with our neonatal pharmacologists to
11
think about those as well.
12
Well, what if it is withdrawal?
Again,
13
does lactation provide some potential prevention
14 against
withdrawal? Again, the kind of
15
conservative management strategies that we have
16
already talked about may make sense.
Bob mentioned
17
that if you really think it is withdrawal, for
18
adults you give a dose of the medication they are
19
withdrawing from and they really feel better fairly
20
rapidly. Is that the case for
these babies as
21
well, and might we think about a dose-ranging
22
safety and efficacy study to define a model so that
243
1 if
they are a certain number of days postpartum and
2
they are having withdrawal symptoms you give a dose
3 and
then taper it in some prescribed way?
That is
4 all
work that needs to be done.
5
Let me finish with this thought, that
6
mental health truly is fundamental to health, and
7 how
to package this so we get the best result for
8 the
mom and baby who are clearly not independent is
9
critically important. Thank you.
10
DR. CHESNEY: Any technical
questions for
11 Dr.
Wisner? Dr. Maldonado?
12
DR. MALDONADO: Excuse my
ignorance, I
13
just have a couple of concepts that I want
14
clarification for. These concepts
are new to me,
15
behavioral teratogenicity. Is
there biological
16
evidence for that, or any animal models?
If there
17 is,
what kind of hypotheses do you think need to be
18
tested in clinical trials to answer that question?
19 The other is the symptoms of
depression
20 you
said are equal to physiological dysregulation.
21 Are
there biological markers that can be used as
22
surrogates to test where the dysregulation is or
244
1
whether those biological markers actually may be
2
good markers to use to see response?
3
DR. WISNER: How long do I
have? Those
4 are
really good questions. Let me deal with
the
5 biological
markers issue first. There is a lot of
6
interest in working particularly at HPA access
7
regulation changes in patients with depression.
8 The
majority of patients, particularly those with
9
what we call typical depression, have high levels
10 of
cortisol and they have accentuation of the HPA
11
access products. There are
patients, those
12
particularly with post-traumatic stress disorder,
13 who
have high proactivation of that access.
There
14 are
some interesting differences diagnostically in
15 how
those axes are affected.
16
Secondarily, we have studies that look at,
17
say, osteoporosis in depressed women, which tends
18 to
be higher. The extension of that is,
well, in
19 women who have depression with HPA access
20
difficulties, are there changes in pregnancy that
21 we
need to know about? In the National
Children's
22
Study, I was in the stress and pregnancy work group
245
1 and
talking about looking at cortisol, CRH and
2
other measures in pregnancy were important and, of
3
course, there are papers which have shown that CRH
4
levels may actually be somewhat predictive of
5 premature birth. So, there is an attempt to look
6 at
some of the changes that we know occur in
7
depression and bring it into a much broader
8
construct of medicine and say, well, what does that
9
really mean? One of those is what
would
10
potentially be the effects for pregnancy.
11
The other question was about behavioral
12
teratogenicity. There certainly
are studies. The
13
ones that I have looked at more recently are
14
studies on long-term effects of fluoxetine during
15
pregnancy in rats and long-lasting changes that
16
occur that result in behavioral problems, but they
17 are
not manifested until a later point in
18
development, or the point in development when they
19 occur
is delayed or made earlier. So, the
issue is
20
that as these potent central nervous system agents
21
occur in the fetal brain, changes happen that we
22
might not see directly at birth but we might see,
246
1
say, at age 7 as development unfolds.
2
The problem that I was trying to identify
3 is
an interesting concept. There certainly
are
4
animal data, and I am more familiar with the animal
5
data on this to support it. But
then how far back
6 in
time do you go to say that is the mechanism?
7 Or,
what if the exposure was at birth, whatever it
8 was
happened at one month or two weeks, how would
9 we
distinguish something that is the result of that
10
mechanism from either withdrawal or acute side
11
effects? That is the point I was
trying to raise.
12
DR. CHESNEY: Thank you. I think that
13
last point you made is something that has been
14 puzzling
me and I think that is one of the very
15
subtle aspects of this whole issue that we are
16
going to be wrestling with.
17
DR. WISNER: In the study that I
am doing,
18
although I focused just on the birth and 2-month
19
effects because we are looking at this neonatal
20
issue, it is embedded in a study in which we are
21
also doing a minor physical anomalies assessment
22
because Tina Chambers' paper found higher minor
247
1
anomalies in the fluoxetine-exposed kids, as well
2 as
far as major anomalies and overall developmental
3
progress as well. So, it is
couched in a study
4
that goes out to 24 months.
5
DR. CHESNEY: Any other technical
6
questions for Dr. Wisner? Naomi,
you had one. Go
7
ahead, Dr. Luban.
8
DR. LUBAN: I am just
curious. The only
9
articles that I could find that actually quantified
10 the
drugs were in a very, very small case report.
11 Is
there a broader-based data set that has looked
12 at
the differences in clinical manifestations
13
apropos of drug level actually measured in the cord
14 or
in a newborn infant?
15
DR. WISNER: Carey Laine's paper
that Bob
16
mentioned is really the best paper because they
17
have not only levels of drug and metabolite but
18
levels of serotonin metabolites as well.
They also
19
scanned the babies' brains to show that there were
20 no
structural abnormalities. But as far as
a paper
21
which really needs to be done in which the cord and
22
then potentially serum levels have been tracked
248
1
across time and related to symptoms, that has not
2
been done to my knowledge.
3
DR. LUBAN: Thank you.
4
DR. CHESNEY: Yes, Dr. Sachs?
5
DR. SACHS: I was just curious
about two
6
things. One thing that struck me
is that I know
7 for
lithium, for example, there is a lot of
8
variation in the way the drug is metabolized right
9
around delivery. It kind of
occurs to me that that
10
might be the case here and I am curious if your
11
study is going to look at that.
12
DR. WISNER: That is such an
interesting
13
point. With respect to
antidepressants and how
14
their dose and metabolism might change across
15
pregnancy, there is one paper that is published
16
that looks at serum levels across pregnancy and
17
antidepressant dose. It is a
paper that my group
18
published about tricyclics in '93.
That is pretty
19
bad. Essentially, what we showed
was that there is
20 an
increase across pregnancy, particularly starting
21
with the second half of pregnancy and then in the
22
third trimester the oral dose required to achieve
249
1 the
same serum level was an average of 1.6 times as
2
high. Others have described that
but not looked at
3
serum levels for SSRIs.
4
In this study, in fact, we have serum
5
levels and cortisol hormones--all kinds of stuff,
6 at
weeks 20, 30 and 36 across pregnancy.
So, we
7 are
looking at that issue. We have, again,
very
8
careful mapping of depressive symptoms.
My major
9
interest is in sorting out what are the things on
10
these scales that happen with depression with no
11 drug, and what are the things that happen with
drug
12 but
no depression, and what is the mush in between.
13 We
call that fourth group affectionately that I
14
defined our mush group because they are probably
15
going to give us that answer.
16
DR. SACHS: And you mentioned that
you are
17
doing I guess questionnaires about substance abuse
18 and
things like that. Are you actually doing
drug
19
screens, alcohol levels?
20
DR. WISNER: Yes, at the 20-week
intake we
21 do
a drug screen and exclude any women with any
22
substances of abuse. In fact,
that has been very
250
1
interesting. The number of
positive drug screens
2
from women who declare absolutely that they never
3
took anything, those women are excluded.
You still
4
can't exclude everybody based on a drug screen and
5
some of our women consume what I think are
6
unhealthy doses of alcohol after they are in the
7
study. We keep them in but we
continue to track
8
that. But my guess is we will
have to analyze
9
those patients separately.
10
DR. CHESNEY: Dr. Gorman?
11
DR. GORMAN: Of particular interest
to me
12 was
the longer-term follow-up to 24 months.
Will
13
there be any objective non-maternal, non-physician
14
office evaluation of those babies?
15
DR. WISNER: Objective? Well, let me tell
16 you
what we are doing and you can tell me if it
17
fits into your categorization. At
18 and 24 months
18 we
were very interested in more subtle behaviors
19
like task persistence. So, we
have our mastery
20
motivation model that Kay Jennings developed that
21 has
to do with the toddler's ability to attend to a
22
prescribed task. There are timed
measures in that.
251
1 It
is sustained attention, propensity to be
2
activated to continue to solve a task.
It is that
3
kind of measure. It is a measure
that is affected
4 by
maternal depression so, again, we are interested
5 in
that in the four groups. Across the
postpartum
6
period for all time points we have an appropriate
7
measure. The Bailey scales. We do the full
8
scales. One of our neonatal
psychologists does the
9
Bailey scales across the postpartum period for kids
10 as
well. We have standard pediatric exams
at all
11
points. Is that what you
meant? What are you
12
thinking of?
13
DR. GORMAN: No, those are
commonly
14
accepted and I think perhaps the state-of-the-art
15
evaluations, sometimes some of the global
16
impression scales that I have very little faith in,
17
except I think they actually do work.
When Kennedy
18
Kreeger asked the mothers in the waiting room to
19
give a developmental age for the children and then,
20
after they did a 72-hour exam, they were within a
21
month or two of being correct.
So, I was looking
22 for
day care providers or child care centers or
252
1
kindergarten teachers--I know you are not going out
2
quite that far--are they different?
Or, what do
3 you
think about these kids?
4
DR. WISNER: You mean collection
of data
5
about the kid that is as uncontaminated by maternal
6
report as possible.
7
DR. GORMAN: Correct.
8
DR. WISNER: No. We have CBCL at age two,
9
which is again a maternal report.
We have a number
10 of
measures of maternal function, like maternal
11
role function, maternal role gratification and
12
completion of immunization rates in the first year
13
that are more functional measures for the mom, but
14 no
totally independent--I mean, even the Bailey's
15
would not be totally independent although it gets
16
closer than some of the other things you are
17
talking about.
18
But in the resubmission and competing
19
continuation of this grant, we certainly are going
20 to
propose to go out to school age kids because
21
that is really important.
22
DR. GORMAN: It is just that in
this
253
1
particular population the contamination with
2
disease diagnosis or potential disease diagnosis
3
makes the data even harder to interpret for those
4
outside the field.
5
DR. WISNER: There are a couple of
things
6
there. It is something that we
can at least look
7 at
across the four groups because we have the
8
occurrence of depression and drug all the way from
9 pregnancy out to that 24-month time. So, we will
10
have women with trait depression, that is, they
11
have had it but they are well, commenting on these
12
measures; women who are actively depressed,
13
commenting; women who are normal controls and that
14
mush group. What I think you are
getting at is
15
what is the validity of material that is
16
observational about an infant or toddler if it is
17
reported by someone who is depressed, whether it is
18
state, that is right now, or whether it is
19
potentially trait. So, it is more
of a validity
20
issue.
21
I think you are right, the way to really
22 get
at that--I am cringing because it is hard to
254
1 do,
but the way to get at that is what we think
2
about, say, blind observer ratings.
Now, even a
3
teacher isn't though because that teacher, knowing
4 the
parent, is going to be to some extent affected.
5 It
is a little more clean but still the validity
6
issue is important. I mean, if
you ask a teacher
7 of
a five year-old to fill out a CBCL, that teacher
8
knows the family. I mean, it is a
little more
9
non-biased. The ability to
comment on that child
10
related to a class of 30 is probably more what we
11 are
after.
12
DR. CHESNEY: I have a suggestion
for you,
13 to
hire Dr. Gorman as your consultant for your
14
study!
15
DR. WISNER: Fabulous!
16 Discussion of Questions 2 and
3
17
DR. CHESNEY: Thank you very, very
much
18 for
a rigorous drilling here. I think we
need to
19
move on to the questions. Dr.
Iyasu is going to
20
post those for us and maybe get us started on the
21
first one.
22
DR. IYASU: We have two questions
for you,
255
1 as
usual, and we have subparts to those questions.
2 The
first question has to do with how we
3
disseminate the information, the new label
4
information to the public and prescribers. The
5
second question deals with additional research that
6
could eliminate some of the issues on neonatal
7
toxicity and withdrawal.
8
I will read the first question:
The FDA
9 is
proceeding with class labeling about neonatal
10
toxicity/withdrawal syndrome related to in utero
11
exposure to SSRI/SNRIs.
Considering the
12
risk/benefit of SSRI/SNRIs use in pregnancy with
13
depression versus the risk/benefit to the
14
fetus/newborn, how should this new information on
15 the
label be disseminated to child health
16
practitioners and the public?
17
For your comments, here are
the options
18
that we have listed. Please
discuss the following
19
options: No further action is
necessary. Label
20
change is adequate.
21
A "Dear Healthcare Professional" letter.
22
Prescriber or healthcare
professional
256
1
education through professional groups.
2
The last option is a public health
3
advisory. After you have
discussed this I will
4
read the next question.
5
DR. CHESNEY: Could you just
elaborate on
6 the
public health advisory? What would that
7
involve?
8
DR. IYASU: Well, that would
involve
9
issuing a public health advisory.
That means
10
really an explanation of what the label change is
11 and
why we are doing it. It is usually
issued by
12 FDA
and includes information about the rationale,
13 the
new information and is disseminated to the
14
public and also put on the website, and also there
15 is
a paper that goes out. So, it is really
a
16
high-level dissemination so that everybody knows
17
about this new label information.
18
DR. CHESNEY: Thank you. So, we should
19
proceed with question number two and then you will
20
come back with question number three.
The issue is
21
that the FDA is moving ahead with class labeling.
22
That is a given. They are asking
us for
257
1
information as to how the fact that the label is
2
going to be changed should be disseminated to child
3
health practitioners and the public, and they have
4
given us four potential options.
Dr. Nelson?
5
DR. NELSON: In trying to
formulate an
6
answer to which approach is best, I would start by
7
framing it as a question of informed consent. What
8
strikes me about this area is, as compared to a
9
label which gives you data, you have a complex
10
balancing within the decision-making of the
11
pregnant woman between risks to herself, risks to
12 the
fetus and risks to the newborn. I think
there
13 has
been a lot in the ethical literature about that
14 in
other areas.
15
So, in framing it as informed consent,
16
then the question would be which of those actions
17
would be most effective in providing information
18
that could be useful within the informed consent
19
process. I would be concerned if
that were seen
20
simply as providing information to the healthcare
21
professional. Looking at the
existing label with
22
non-teratogenic effects and looking at the
258
1
pregnancy, it simply says tell your doctor if you
2 get
pregnant. Then, under the
non-teratogenic
3
effects it talks about what the physician should
4
think about. But there is really
nothing in here
5 about
the risks of untreated depression in
6
pregnancy. I mean, there is
nothing in here, as
7
opposed to the articles, and there was some
8
discussion of that, but nothing that I think you
9
could give to a pregnant woman to say here is
10
something that can help you and, in fact, if it
11
helps here it probably helps the health
12
professional think through this complex
13
risk/benefit decision.
14
So, the question I would ask is could one
15
develop information for the patient, much as
16
Duragesic had, that could go through the kind of
17
decision-making issues that would have to be
18
addressed? That would be a very
complex document.
19 But
I am not sure any of these four actions that
20 are
proposed actually would really get at the
21
informed consent question which I think is at the
22
heart of this.
259
1
DR. D. MURPHY: In a way, I think
what you
2 are
telling us--and I would ask you all to comment
3 on
this, is that the information that we are
4
putting in the label-- because, again, our labels
5
have to try to at least raise this issue--is that
6
that is not adequate for people to make a
7
prescribing decision. So, you are
proposing--and I
8 am
not quite sure whether you are saying it is not
9
adequate for the physician or mostly for the
10
patient--and you are proposing that we have in
11 addition
a patient insert on this issue that would
12
have more information that would allow the patient
13 and
the physician to have a more detailed
14
discussion. Is that correct?
15
DR. NELSON: Yes, I think it is
correct.
16 I
think some of the comments that were made about
17
women making decisions based on their response to
18
coming off medication and whether they get sick
19
quickly or get sick slowly, you are not going to
20 put
that in the label. It can't be put in
the
21
label. So, how you give people
information to do
22
that kind of balancing is the question.
You can
260
1
certainly have the risks of the non-teratogenic
2
effects in here but I can't imagine a sponsor
3
wanting you to put in the risk of untreated
4
depression in the label--
5
DR. D. MURPHY: Yes.
6
DR. NELSON: --for an
antidepressant.
7
DR. D. MURPHY: Sandy Kweder, from
the
8
pregnancy labeling group is back in the audience.
9
Sandy, would you like to make any comments on this
10
area, and then I would like to go back to the
11
Division and see what the Division might have to
12 say
too.
13
DR. KWEDER: Good afternoon. One of the
14
things that we are in the process of is trying to
15
revise the regulations for how drugs are labeled
16 for
use in pregnancy and lactation. One of
our
17
goals in that is to try and frame risk information.
18
What I mean by that is try to include in labeling
19 any
information that would be relevant to take into
20
account when considering the risk to the extent
21
possible. In the version that we
are working on of
22 a
new regulation, one of the things that we will be
261
1
asking companies to do in labeling is, to the
2
extent possible, to include some information about
3 the
risk of the illness in pregnancy, of not
4
treating the illness in pregnancy.
5
We have done this in several cases
6
already. Even though we are quite
a while away
7
from a new regulation, we have been trying to
8
incorporate that to the extent we can.
A couple of
9
examples where we have done it have been in drugs
10 to
treat and prevent malaria. Because the
risk of
11
malaria in pregnancy to the mother and fetus is
12
extremely high and grave, we have incorporated that
13
juxtaposed to any risk information.
We have done
14 it
recently for some asthma medications.
The risk
15 of
untreated asthma in pregnancy is discussed.
16
So, nothing is perfect and, you know, the
17
unfortunate thing is sometimes we don't have data,
18
although in this case I think there are some and it
19
certainly could be done. One of
the things that we
20
know about this section of the label, unlike most,
21 is
that doctors read it. It is also often
the only
22
thing that they read and take into account when
262
1
considering whether to prescribe a medicine in
2
pregnancy. We also know that patients
read this
3
section of the label. Pregnant
women are a
4
population that is very savvy and they look stuff
5
up. One of the first things they
find when they
6
look things up when they are pregnant is the label.
7
So, even though the
information, as you
8
said, is for the prescriber, and the label itself
9 is
not necessarily the tool through which to
10
communicate information to the patient, we have to
11
take into account that they will read it and we
12
need to take care in how we frame things in the
13
label because it is likely to reach both prescriber
14 and
patient. Is that what you were looking
for,
15
Dianne?
16
DR. D. MURPHY: Yes. I think we always
17 have to deal with that balance. Bob, did you want
18 to
say anything more about where the Division is?
19
DR. LEVIN: Sure. Dr. Nelson, I think one
20
thing you are suggesting, and if this is the case I
21
agree, is that the labeling currently doesn't
22
address the risk/benefit as fully as one might like
263
1 and
focuses more, obviously, on potential adverse
2
events than it does on potential benefits of
3
treatment.
4
Also, in general my sense is--at least in
5 our
Division we talk about this--that in labeling
6 we
try to stay away from micro-managing, dictating
7 or
strongly suggesting treatment. Even
though in
8
some cases we do, obviously, in dose
9
administration, my sense is that people try to stay
10
away from giving real definitive recommendations on
11
exact treatment. So, it might
relate to what you
12 are
saying. But I agree that it would be
ideal to
13
have something in the labeling that more carefully
14
details the risks and benefits of treatment or not
15
treating.
16
DR. CHESNEY: I think what we have
heard
17
from Dr. Kweder is that the agency already has
18
experience in terms of putting information into
19
label situations in which there is a very high risk
20 to
both the mother and the infant of doing one
21
thing or another. So, they have
had that kind of
22
experience and we would assume it would be carried
264
1
over into this area. Dr. Wisner,
you had your hand
2 up?
3
DR. WISNER: I guess the way I
think about
4
this is the way that was mentioned, which is what
5 do
we want people to do? And, what we want
them to
6 do
is recognize that things that happen to neonates
7
born when moms take these drugs have to be
8
considered in the context of that risk/benefit
9
decision, which is more of an education issue.
10
The thing that makes me a little uneasy is
11
that what to do is so unclear. I
make the choice
12 to
offer the option to taper but as a researcher I
13 sit
here and say but I am cleaning my data about
14 the
outcomes for the babies about whether that
15
intervention actually works, and I am cleaning the
16
data about depression scores in the moms. So, I
17
would like to have more to say to them, other than
18 be
aware.
19
Just as an aside, when Solomon called me
20
about this and Sandy too, I actually put on a
21
couple of graduate students to clean that data
22
because I understand now the importance of getting
265
1 it
out into the literature. But it seems to
me I
2
would like to have more meat in terms of telling
3
them what to do once I get to the risk/benefit
4
decision.
5
DR. CHESNEY: Dr. Nelson?
6
DR. NELSON: Just a follow-up
comment on
7 the
labeling experience that you have already had,
8 I
am thinking of it from the sponsor's point of
9
view, and it is pretty clear that if you have
10
malaria, in fact, listing the risks of untreated
11
malaria drives individuals to realize the
12
importance of getting treatment.
That is very
13
different than sharing the ambiguity about
14
something that is so extensive in the population
15
that you are then on it when you get pregnant and,
16 in
fact, given the variability in the diagnosis of
17
depression, in many ways what you are trying to do
18 is
encourage people not to take the medication.
19 So,
I could imagine the discussion around the label
20
would be framed very differently in depression than
21 it
would be perhaps in malaria and the other
22
conditions. It sounds like you
are going in the
266
1
direction that I encourage, but whether or not you
2
would get there in this case, based on the other
3
ones, I think is an open question.
4
DR. D. MURPHY: Bob will have to
help us
5
carry the message back.
6
DR. CHESNEY: Another consultant,
along
7
with Dr. Gorman, for Dr. Wisner.
I think what we
8 are
all groping with is what you just mentioned,
9
which is that we don't know what these
10
manifestations represent and, therefore, we don't
11
really know what to do about it, and I don't know
12
that we can--in fact, I am sure can't solve that
13
today, but I think what the FDA is asking us is
14
what level of anxiety should we have, should they
15 have in terms of how to at least let people
know
16
that this is a recognized phenomenon, even if we
17 are
not exactly sure what to do about it--if that
18 is
correct, I think that is where you wanted the
19
focus to be.
20
Any comments about that? Do we
want to go
21 to
the equivalent of a public health advisory or
22
just let the process of label change move ahead, or
267
1
something in between? Dr. Hudak?
2
DR. HUDAK: Well, again, I
struggle with
3
exactly what all this information means for the
4
baby. I mean, what we have heard
so far I think is
5
that some subset of babies have what appears to be
6 a
transient period of symptoms. We have no
idea
7
whether or not there are later-term persistent
8
effects, and studies certainly need to be done on
9
that.
10
I would say that even if you do get
11
studies at two years of age that show that there is
12 not
an apparent effect, that doesn't guarantee that
13
there is not an important long-term effect because
14 in
babies what we have been finding out is that we
15
often have neurodevelopmental follow-up at a year
16 or
two years of age where it shows no difference
17
between the two groups, whatever they are, but by
18 the
time you get to school age and look at function
19
there are very significant things that are present
20
that impact how those children can be taught.
21
I think here it is very difficult, without
22 any
data, to sort of have a huge public health
268
1
advisory. On the other hand, I
would say that the
2
information--I mean, this is relatively new and I
3
don't think widely available information to target
4
obstetricians and family practitioners who deliver
5
mothers and those professionals that take care of
6
newborns should know.
7
One of the important things--this is a
8
trivial thing but one of the important things is
9 you
would think that a pediatrician who took care
10 of
a baby whose mother was treated with one of
11
these drugs would know that the mother was treated
12
with one of these drugs, but I will guarantee you
13
that that doesn't happen. That is
a shocking thing
14 but
in the hospital environment we have been
15
working for years with medical record systems and
16
obstetricians, and so forth, to let us know a
17
simple thing, that is, is the mother Group E strep
18
positive and, if so, did she get antibiotics and
19 how
long before she delivered did she get them
20
because it impacts how we evaluate that baby and
21
take care of that baby. And, we
have just gotten
22 to
the point where we are successful but we do not
269
1 get
prenatals; we do not get any information in the
2
neonatal record in the hospital as to what
3
medications the mother is on necessarily. That may
4 be
known but it is not available easily to the
5
people taking care of the babies.
I imagine in the
6
office setting, Dr. Gorman, when you see a baby for
7 the
first time that information is even more
8
closeted.
9
So, I think that one of the things in the
10
advisory needs to be communication, that if mothers
11 are
on treatment for these things, rather than
12
making it, you know, something that should be
13
hidden, it should be something that is accessible
14 and
made known to the people who are taking care of
15 the
infants.
16
DR. CHESNEY: Can I ask a very
pointed
17
question? What is the downside of
a public health
18
advisory? I wonder if an upside
wouldn't be in
19
alerting everybody that this is a concern and much
20
more research is needed. Would
that, thereby,
21
stimulate granting agencies to recognize that this
22 is
a very pronounced problem at this point in time
270
1
that we need to address? Would
that be an upside?
2
What are some recent examples that we could perhaps
3
compare this to? If we have to
come down on some
4
side or another, I guess I would come down on that
5
just to get the discussion started.
What are the
6
pros and cons of a public health advisory for
7
something like this?
8
DR. D. MURPHY: We had all this at
one
9
time. For the health advisory I
think in this
10
situation the positive would be, yes, you would get
11 it
out to a large number of people. But the
very
12
potential downside is exactly what the committee
13 has
been discussing, which is what are we telling
14 you
to do? Not that we tell you at FDA what
the
15
practice of medicine is. That is
not it. But do
16 we
have enough data to even tell you anything
17
beyond the fact that this occurs?
18
Now, one could argue that that is a
19
sufficient message but if you send out too many
20
messages you lose the effect of the messages. So,
21 I
think that saying, very well articulated, this is
22
what we know about it, you know, we know that it
271
1
occurs in certain situations; we can tell you what
2
these are. From the FDA's point
of view, it is,
3
again, informing the physician who is prescribing
4
this medicine so it is back to Dr. Nelson's point.
5 You
know, what else can we tell you about how to
6
prescribe it or not prescribe it?
That needs to be
7 really
put into some sort of context.
8
So, I think if you are going to do the
9
advisory you have to be able to come up with a
10
context that would allow people to make those
11
risk/benefit assessments. I mean,
has it always
12
been true for every one of our advisories? No. As
13 you
know from some of our early SSRIs, we were
14
criticized for some of the advice we gave there,
15
which was just be aware. But we
though it was
16
important enough, there was enough concern that we
17
went out--you know, we are still struggling with
18 how
much information we don't have but we thought
19 it
was important to get it out.
20
So, that is sort of what we are asking the
21
committee. With this limited
information should we
22 do
any of these other things at this time?
Sandy
272
1 has
worked at a lot of these with us.
2
DR. KWEDER: Good afternoon again. Yes,
3 we
have done a number of these and we try to be
4
judicious in selecting simply because you can only
5 do
so many of these before people stop listening.
6
Also, when we do issue them, usually it is because
7
there is something that people can do.
Some of the
8
more recent ones that come to mind are risks of a
9
particular drug that are new and that are
10
potentially serious and immediate that would
11
require stopping a medicine. Or,
we have done them
12
when a drug is being withdrawn from the market and
13 we
expect that clinicians need to know right away
14
that there is a serious safety issue.
15
In the few cases where we have issued them
16
when there is not something like that, as someone
17 who
takes a lot of the press calls, people are very
18
confused when we don't have "and, therefore, you
19
should do this." Both the
professional groups and
20 the
lay public don't really understand why we do
21
that. So, these things do have
their pros and
22
cons.
273
1
One of the things that we have learned is
2
that it is very frustrating for practicing
3
clinicians and professional groups when a public
4
health advisory comes out and they aren't aware of
5 the
data. They understand it when it is
something
6
that is really critical with, you know, a major
7
public health issue that is immediate but when
8
there are nuances and there are data behind it that
9 may
be complicated they are frustrated when FDA
10
comes out with something and they haven't had an
11
opportunity to digest the data that underlie it and
12
prepare themselves in their practice for what may
13 end
up being a deluge.
14
DR. CHESNEY: Just for the sake of
15
argument, I was interested in the materials and the
16 one
thing that can be done is to observe these
17
infants for a longer period of time for some of
18
these findings. They said, for
example, now with
19
discharge within 24-48 hours maybe these infants
20
need to be observed for a longer period of time. I
21
realize that is trivial compared with what we would
22
like to tell them to do but, again, just for the
274
1
sake or argument, there is something that could be
2
done which is watch for these children because you
3
might see some difficulty eating or all these
4
different things. Dr. Hudak?
5
DR. HUDAK: No, I think that is a
good
6
point and that is why it is necessary for the
7
physicians who treat the baby to know those things.
8 The
good news is that the trend is in the opposite
9
direction now, that mothers are staying not 24
10
hours but more like 48 minimum, which is a good
11
thing. The other thing that this
would do is that
12
even if the baby is okay at 48 hours, it would
13
encourage the baby to be seen in early follow-up
14
which would mean one or two days after discharge
15
rather than two weeks, which is typical in many
16
practices, especially the non-nursed baby.
17
DR. CHESNEY: I am going to stop talking
18
right after this, but if we go back one
19
step--professional education, I feel like that is a
20
given. We have to do that through
a whole variety
21 of
different societies and so on. But our
next
22
alternative is the "dear healthcare professional"
275
1
letter which is the sponsor's responsibility. And,
2 I
will stop talking and get some other input.
Dr.
3 Gorman?
4
DR. GORMAN: I think one of those
has
5
already arrived in my mail box.
Being the good
6
doctor that I am, I haven't read it yet.
It is
7
sitting on the pile of unopened mail but it does
8 say
a large pharmaceutical company and it says
9
"open immediately, dear doctor" letter. So, there
10 may
be one of those already out there.
11
I am going to take the side of the
12
obstetricians and psychiatrists for a moment and
13 say
that we are talking about a neonatal withdrawal
14
syndrome but this decision has the potential for
15
major negative impact on the mother.
In the
16
present state of information where we have what we
17
presume to be an acute withdrawal phenomenon, I
18
think the label is adequate as it is because if we
19 try
to change practice for obstetricians and
20
psychiatrists that have negative adverse events on
21 the
moms when we have just an acute withdrawal
22
syndrome for babies, pediatricians and
276
1
neonatologists should be able to handle an acute
2
withdrawal syndrome.
3
Having said that, my background and my
4
spotty career or checkered career or mosaic career,
5
depending on which way you want to think about it,
6 is
lead. Lead is how I got interested in
this
7
whole field and there is an area where there is
8
obviously an initial incident and then a long-term
9
devastating neurological outcome.
Like lead, this
10 is
so commonplace today, if 10 percent of pregnant
11
women are going to be on these medications we will
12
have a really hard time teasing this out if it
13
doesn't get teased out earlier, meaning in 2004,
14 '05
or '06 rather than in 2024 or '25 or '26.
15
DR. CHESNEY: Dr. Nelson?
16
DR. NELSON: I have a question but
it
17
would help me then frame how you would target the
18
professional education. There are
two options. I
19
would assume that most of the overlap between
20
depression and pregnancy are women who are on
21
antidepressants becoming pregnant, as opposed to
22
pregnant women getting depressed.
277
1
DR. WISNER: That is probably true
but
2
there are, in fact, many women who have a first
3
episode of depression in pregnancy.
So, there are
4
both subgroups.
5 DR. NELSON: And if they have that in
6
pregnancy, are there differences in terms of
7
response to antidepressants?
8
DR. WISNER: It has never been
studied
9
systematically but from our long-term experience,
10 no. Differences in side effects but not efficacy.
11
DR. NELSON: Because depending, I
guess,
12 on
which approach you take when thinking about it,
13 I
mean, I am not sure I would advocate a "dear
14
health professional" letter because if I got it, I
15
mean, I would kind of look at it and say okay.
16
But, you know, what would you say to someone who is
17
prescribing antidepressants when they would counsel
18 a
woman, should she be thinking about becoming
19 pregnant
as one set of questions, and then what to
20 do
if she becomes pregnant. And then a
whole other
21 set
of questions is then depression during
22
pregnancy and the kinds of decision-making that
278
1
would be different and would be approached
2
differently and would be complex in
3
short-term/long-term issues.
4
So, I think, by default, you would end up
5 in
the third because the healthcare professional
6
letter which a sponsor sends out I can't imagine
7
could go into the kind of detail that you would
8
need to tease out those issues, in particular since
9
most of them would end up on the second page and
10
people wouldn't read past the first paragraph or
11
two. So, I think by default you
end up in three.
12 So,
the question is, is there a way you can
13
stimulate the third in a way that is productive?
14
My own bias is that I think a public
15
health advisory--it doesn't sound like there is
16
enough concrete information to where sending that
17 out
wouldn't send up an alarm and everybody says,
18
well, what do I do about it? And,
the answer is we
19
don't know. That would strike me
as crying wolf in
20 a
way that would undercut that process.
21
DR. CHESNEY: I hear what you are
saying.
22 On
the other hand, SSRIs are such a hot button item
279
1
now. Other comments? Dr. O'Fallon?
2
DR. O'FALLON: We really have two
patients
3
here. We have the mother and we
have the babe. We
4
don't really know at this point how damaging this
5
toxicity or withdrawal is. We
truly don't have the
6
data. So, this may be a horrible
problem about to
7
explode, I mean, down the line five years from now
8 or
it may not.
9
DR. WISNER: There are long-term
follow-up
10
studies of kids that have been exposed during
11
pregnancy, particularly to fluoxetine and the
12
tricyclics. By and large, the
development, at
13
least on fairly global but standard measures, has
14
been indistinguishable between the groups. So, I
15
don't think we are looking at something that is
16
going to blow up and be very bad down the line at
17
least on those major impacts.
There is still a
18
dis-ease about some more subtle, perhaps those
19
neurobehavioral things that we are talking about,
20 but
I don't honestly think it is a major horrible
21
thing.
22
DR. CHESNEY: I think that is
critical
280
1
information. I don't believe I knew
that. So,
2
there have been extensive long-term follow-up
3
studies of infants exposed to SSRIs in utero and
4
those children are now without clear complication
5 or
problem?
6
DR. WISNER: That is correct.
7 DR. CHESNEY: Dr. O'Fallon?
8
DR. O'FALLON: The point here is
we have
9 all
been talking about giving the information out
10 to
the doctors and, you know, it is important.
But
11 you
also have to give it to the mother. I
mean,
12 the
mothers have to have this information given to
13
them the best that they can have.
You know, I
14
don't know what our options are but that letter to
15 the
patient we saw earlier today, maybe something
16
along those lines, or maybe the FDA could have a
17
pregnancy website where they could keep the latest
18
information about issues pertinent to pregnancy so
19 if
a woman gets pregnant and wants to do her
20
homework she could go look up as much information
21 and
possibly that would at least help her.
Because
22 if
she is depending on one of these doctors that is
281
1 so
busy, and I understand because I work with them,
2 they have never read it. They never read that
3
advisory and she is not going to get the
4
information either. There should
be a way to get
5
that information to the public directly.
6
DR. D. MURPHY: When you say patient
7
letter, you are talking about the patient insert
8
that we are proposing for the label?
9
DR. O'FALLON: Right.
10
DR. D. MURPHY: That would say in
more
11
detail to the mother about depression and treatment
12 and
just making her aware of the fact that this a
13
risk. We know that there is a
risk to the infant;
14 we
can say that as far as acute, manageable
15
toxicity or adverse event. The
question I think we
16 are
struggling with is--and that may be fine but
17
then is there anything else to say?
At this point
18 I
am hearing even though the Division is proposing
19
that we have in there that your physician may want
20 to
taper your medication, we would then have to
21
actually frame that in a way that would be more
22
balanced about the limitations of information.
282
1
DR. O'FALLON: That is right.
2
DR. D. MURPHY: I mean, the point
of a
3
patient letter is that it gives you the ability to
4 say
more to balance it instead of just saying we
5
don't know. There is that
opportunity.
6
DR. O'FALLON: And like she just
said, if
7
they have some information--it may not be the
8
highest quality because it may be voluntary
9
information and all that, but if there is something
10
there that says, "hey, look, we haven't seen all
11
these long-term things" the mother could say,
12
"well, you know, it won't be so awful for my baby
13 if
I stay on my medication"--that type of thing.
14
Give them the information so that they can make an
15
informed decision like he was talking about in
16
terms of an informed consent.
17
DR. CHESNEY: Dr. Wisner?
18
DR. WISNER: I agree very much
with what
19 you
just said, but what I worry about is exactly
20
what I saw in a case recently. I
did a
21
consultation on a patient who was very pregnant,
22 who
came in because she said, "you know, I went on
283
1 the
web and I saw all this terrible stuff that
2
happens to newborns if the mom is taking an
3 antidepressant
so I stopped my drug a month ago
4
because what if I deliver early and maybe my baby
5
will get those terrible things?"
We had to
6
hospitalize her because she was really quite
7
depressed and suicidal and she went into labor,
8
delivered and had to be transferred down to the
9
maternity hospital. I have no
question that all of
10
that was way worse than to continue the drug.
11
So, if the information is in a context in
12
which she is helped to value the traces, that makes
13
sense, but delivered, you know, in a situation
14
where, in fact, her treating physician wasn't very
15
aware of the issues and she didn't have a lot of
16
confidence, I just worry about the meta-message.
17
When this organization says something it can pack a
18 big
wallop and that meta-message may lead to more
19
negative outcomes than we hope, or the kind of
20
negative outcomes that we don't want to happen.
21
DR. O'FALLON: But maybe they are
going to
22 go
look at the website where it doesn't have any of
284
1 the
nuances in it. So, I am suggesting that
the
2 FDA
try to create a balanced message and point to
3 the
different issues so that they would have a
4
halfway chance of knowing what they are doing.
5
DR. WISNER: Except that my point
was
6
somewhat different. That is, if
it is an FDA
7
message there is a meta-message that is separate
8
from what the content says that packs a bigger
9
wallop even if it is tempered.
That was more my
10
point.
11
DR. D. MURPHY: If FDA sends out a
public
12
health advisory, you are right, there is a big
13
meta-message and you have to read a lot to be able
14 to
overcome that meta-message, which is the point
15 of
not sending out a whole lot of them because we
16
want you to pay attention when we send them out,
17
versus the other proposal I am hearing.
I am just
18
trying to make sure I am getting your perspective
19 on
this, versus the patient insert, which is
20
another way "FDA says..."
Are you still concerned
21
about that meta-message that we would have a
22
patient insert that went through this issue of
285
1
mothers on antidepressants and that there are these
2
syndromes, but we don't know what the long-term
3 effects
are? We have some evidence at least in
4
certain situations that there aren't that we know
5 of
at this point or have been able to identify, and
6
that this is a decision you need to balance against
7 the
importance of maintaining your health during
8
this process. I mean, you are
concerned that even
9 for
the additional--because something is going to
10 go
in the label because we have this information
11 and
we have to tell people.
12
I think what is being brought forth is
13
that there is a concern that that alone may not be
14
balanced enough, and is there another way to
15
balance it without making it worse, and is the
16
patient insert that way versus--I think I am
17
hearing we don't want to send out an FDA notice but
18 is
there another way?
19
DR. CHESNEY: What I am hearing
are three
20
things. One is that there will be
a label change.
21 The
second is that we should educate. I
think that
22 is a given.
Then, the third, which is the issue
286
1 now
is whether, Dr. Wisner, you have concerns that
2
just by putting in a patient insert or an enhanced
3
patient insert would frighten women, without any
4
black boxes or anything, just a more informational
5
patient insert?
6
DR. WISNER: Actually, I was
responding
7
more and I think agreeing with the negative
8
feelings about the advisory as a major message.
9 You
know, I guess as you were talking what I kept
10
thinking about is what was raised before, that
11
women will find out about this if they are
12
industrious anyway, and if we take the meta-message
13
liability away, which I think would happen with the
14
advisory, and it is a balanced presentation, that
15
sounds more reasonable because at least it is a
16
trusted source. If the directive
is to say here is
17
information we think you need as you consider your
18
choices, that makes sense to me.
19
DR. CHESNEY: In the interest of
moving on
20
because we still have another question, is there
21
anybody on the committee that would not agree with
22 an
enriched patient insert, enhanced education of
287
1
physicians and then the label change that is in
2
process? Does everybody agree to
that? Dr.
3
Gorman?
4
DR. GORMAN: The only part of the
enhanced
5
label that I saw was in the dosing and
6
administration and I still have reservations about
7
where it prescribes a course of action for
8
physicians which is to taper these doses. I think
9
that is a leap of faith that we don't have
10
information for.
11
It says when treating a pregnant woman
12
with these drugs carefully consider potential risk
13 and
benefit. Then, physicians may consider
14
altering or revising or rethinking these treatments
15
during pregnancy as information becomes available
16
that will give the physician alternatives. The
17
only alternative they put in here is tapering and
18
then it gets that imprimatur of that is the way to
19
go. I don't think we have any
data to say that
20
that is the way to go. That is
the only
21
reservation I have about the proposed label change.
22
DR. CHESNEY: Thank you. Dr. Iyasu, do
288
1 you
want to give us question number three?
2
DR. IYASU: The second question
deals with
3
research. I will just read
it. BPCA does not
4
provide a mechanism for issuing a written request
5 to study drug therapies for pregnant
women. There
6 are
no population-based estimates of SSRI or SNRI
7
exposure data in pregnant women and there are no
8
systematically collected data on neonatal outcomes
9 in
infants exposed to these drugs.
Furthermore,
10
determining causality for neonatal reactions is
11
challenging as the role of drug discontinuation,
12
direct toxicity (example serotonin syndrome) and/or
13
other drug/substance exposure during pregnancy is
14
often unclear.
15
Is there a need for further research to
16
evaluate and characterize the neonatal effects of
17 in
utero exposure to SSRI/SNRIs? If your
answer is
18
yes, in your discussion of research options, please
19
discuss feasibility and potential sponsors for each
20
option. I think you have answered
the first
21
question already.
22
DR. CHESNEY: Does anybody feel
that we
289
1
know everything that we need to know?
2
[Laughter]
3
Thank you. Moving on then?
4
DR. IYASU: Here are the options
that we
5
have for you for discussion, and think about
6
feasibility and potential sponsors from these
7
approaches.
8
DR. CHESNEY: And we are always
interested
9 in
more options.
10
DR. IYASU: The first option is to
11
continue evaluating/monitoring postmarketing
12
adverse event reports, like what we do with the
13
Office of Drug Safety.
14
Conduct some population-based prospective
15
study of pregnancy exposed to antidepressants and
16
assess neonatal outcomes.
17
Another option is to conduct a
18
retrospective study of neonatal withdrawal syndrome
19 or
serotonin toxicity.
20
The last option is conducting a
21
randomized, controlled trial of treatment of
22
maternal depression. As a subpart
to this
290
1
question, if yes, what research questions should be
2
addressed by the trial?
3
DR. CHESNEY: Probably we can
already
4
x-out the first option because you will continue to
5
monitor postmarketing adverse events.
6
DR. D. MURPHY: I think one of the
things
7
that that was meant to try to say is that there is
8 a
difference between the routine monitoring or
9
saying we are going to follow up, and we want to
10
make sure that when we tell you that we are just
11
going to do routine monitoring that means that we
12 are
going to then be bringing it back again.
Of
13
course, like many of the things today we really
14 didn't
think we needed to bring back to you but
15
this is really that question in a way, do you want
16 us
to just to continue versus just have the routine
17
path of reporting process?
18
DR. CHESNEY: I understand. In other
19
words, with an emphasis on further research, are we
20
content with that or do we want you to continue to
21
look at this internally and bring us additional
22
feedback each meeting?
291
1
DR. D. MURPHY: It could be simply
that we
2
will continue to look at it a year from now and if
3 we
don't see anything we won't come back to you if
4 we
don't have any new information, but clearly if
5 we
did, we would. I mean, you would be
putting
6
this as a task for us, meaning Pediatrics, to
7
follow-up with you versus we are not following up.
8 If
we report to you 1-year post-exclusivity and
9
there is nothing there, we do no longer follow-up
10
with you.
11
DR. CHESNEY: It is a form of
ongoing
12
research for you all, if you will, and you are not
13
required to do that; we would be asking you to do
14
that. Comments? Dr. Nelson?
15
DR. NELSON: I guess I have
three. The
16
first is that I would be interested in Sam's
17
comments on whether a registry requirement for
18
industry for women who become pregnant or are
19
pregnant and they get placed on the medications,
20 whether that would facilitate data collection
and
21
also meet some of the problems behind the voluntary
22
system of adverse events.
292
1
The second two I think would be fine,
2
although I would modify the retrospective and talk
3
about case control, and I would advocate that
4
something like that ought to be NIH funded as
5
opposed to industry funded.
6
The reason for that is that I get nervous,
7
particularly when I look at number four and think
8
about the impact of an industry-sponsored
9
trial--which I presume you could only get them to
10 do
if there was money on the table to be earned by
11
doing it--in a setting where there is under-funding
12 of
basic care for mental health and the potential
13 for
undue influence on women even going into the
14
trial, and the complexities of even designing an
15
ethical trial under those circumstances.
I think
16 it
could be done but I would prefer it then to be
17
done without the sort of recruitment drive that
18
industry-funded research creates even if that
19
recruitment is carried out appropriately. I mean,
20
mental health is so under-funded the undue
21
influence to then go into that trial, even if their
22 own
risk assessment independently might not to be
293
1 on
antidepressants, I think could be potentially
2
large.
3
DR. CHESNEY: Thank you.
4
DR. MALDONADO: I think actually
the FDA
5 has
experience with registries. When I was
at the
6
agency in antivirals we asked companies to do that
7 and
I remember that one of the first registries was
8 for
acyclovir in pregnancy and also AZT.
9
Unfortunately, those registries--please correct me
10 if
I am wrong--never yielded any of the goals that
11
they were created for. I am sure
that Sandy knows
12
that very well. So, basically,
over the years they
13
have not been very good at giving data.
They
14
basically give extemporaneous reports to an 800
15
number by clinicians or by women exposed and they
16
didn't yield the results that we wanted.
17
DR. CHESNEY: Thank you. Dr. Cragan?
18
DR. CRAGAN: I have actually been
on the
19
scientific advisory committees of four of those
20
registries that are sponsored by industry. They
21 are
very similar in methods but the usual one is
22
that the outcomes are obtained.
There is active
294
1
attempt to contact whoever reported the pregnancy
2
originally to the registry, and most of the times
3
that is the obstetrician; occasionally it is a
4
pharmacist or a neurologist or some other
5
specialist. They are contacted to
find out the
6
outcome of the infant. So, it is
somewhat whatever
7 the
obstetrician knows about the outcome or if they
8
take the extra step to contact the pediatrician and
9
find out.
10
But I think in this kind of behavioral
11
type of symptoms, withdrawal versus toxicity and
12
such, you really need to get to hard, objective
13
data by the person who is caring for the child and
14 not
just the first day in a delivery room.
So,
15
that method I think is not well suited for this
16
type of outcome.
17
Now, there is one registry, the
18
anti-epileptic registry, that is multiple company
19
sponsored and they give a grant to someone in a
20
university setting who actually administered the
21
registry, and the mother enrolls herself and they
22 get
informed consent to contact the pediatrician
295
1 and
attempt to get copies of hospital records, and
2
such. So, it is possible; there
is one registry
3
that does that better but I think this kind of
4
outcome is not well suited to that design in
5
general.
6
DR. CHESNEY: Could you comment on
what
7
design would be good, specifically with respect to
8
bullet number two?
9
DR. CRAGAN: What comes to mind is the
10
National Children's Study which is a
11
population-based enrollment, a longitudinal study
12 of
children and they do have work group on drugs
13 and
they have a group on newborn outcomes I think.
14 I
would make sure that these kinds of issues will
15 be
covered in what they are addressing. I
think
16
they will be automatically but presumably SSRIs and
17
antidepressants in general are a common enough
18
exposure in the population that you really may be
19
able to get some good data from that.
20
The other things that come to mind that
21
have an existing structure that you might be able
22 to
tap on--one is the teratology information
296
1
services, the one in California particularly that
2
Christina Chambers heads or is part of.
They are
3 set
up to interview mothers about exposures.
They
4
have several studies where they follow infants out
5 to
a year. They have physicians who travel
around
6 in
California to examine infants. I think
with
7
some funding and some support they would probably
8 be
able to take that on for a longer term.
It
9
depends on how many you have.
10
Our division funds a number of state-based
11
population-based birth defect surveillance programs
12 and
those are geared toward malformations.
But
13
some of those, the ones in California, the one in
14
Texas, do have abstractors that go out to hospitals
15 and
look for abstract information about children
16
with specific conditions. Again,
with some extra
17
support or funding some of those might be able to
18
broaden those to look for symptoms noted in the
19
newborn that then could be followed up to look for
20
exposure. Those are the thoughts
that come to
21
mind.
22
DR. CHESNEY: Other suggestions or
297
1
comments? Dr. O'Fallon?
2
DR. O'FALLON: You know, I think
that
3
retrospective studies are the ones that can get
4
done the fastest and probably provide the most bang
5 for
the buck, at the beginning anyway. But
the
6
problem with them is they do depend on data that
7 was
or was not recorded so you have interesting
8
biases that show up, but they still are the
9
greatest bang for the buck and, at least in the
10
beginning, give us some information.
It would
11
probably have to be validated through something
12
like a prospective study forward in time if it
13
looks like there is something going on.
14
DR. CHESNEY: I was struck in
reading the
15
articles and here today that we don't really know
16
what we are looking for in the newborn.
I mean,
17
there may be two totally different syndromes, one
18
being the behavioral teratogenicity and the other
19
being the toxicity/withdrawal. We
don't even have
20
good definitions for what we are looking for if we
21
were to go retrospective. I just
kept thinking
22
this would be so perfect for somebody that, you
298
1
know, would actually get in and examine the infants
2 and
develop some kind of scale for evaluation.
Dr.
3
Gorman, I think you had your hand up.
4
DR. GORMAN: After listening to
the AERS
5
disclaimer for the last three years, it strikes me
6
that the process we follow for identifying
7
off-label drugs for study at NIH might be a useful
8
analogy to start looking at for AERS signals that
9 are
picked up. They get a signal in
AERS. It says
10
there may be a toxicity that has previously been
11
unrecognized. You use some group
to rank them in
12
terms of their significance and find a way through
13 NIH
to fund them through whatever mechanism.
It
14
struck me as incredibly serendipitous that there is
15 an
RO1 trial going on today that is trying to
16
answer the question that we are being faced with.
17 I
would like to make that less serendipitous.
18
DR. CHESNEY: Dr. Luban?
19
DR. LUBAN: There certainly are
two
20
NIH-sponsored groups that look into pharmacologic
21
trials that PBRUs, and certainly the neonatal
22
network, which is Maternal and Child Health-funded,
299
1 is
another resource. It almost seems like
you need
2 to
get all these people to sit down in one room
3
together and talk to one another.
I would imagine
4
that between that you could get the measurements
5 that
I would be most interested in looking at
6
because the PBRUs have very extensive drug testing
7
methodologies available to them and the neonatal
8
network certainly has a broad base of diffuse
9
neonates from different socioeconomic groups that
10 are
from across the United States.
11
DR. CHESNEY: Maybe I could hazard
a
12
response to your first bullet. I
partly feel like
13 you
have done your job and we don't see you as a
14
research agency. On the other
hand, given that it
15 may
take some time for some of these other programs
16 to
get up and running, and hopefully they will
17
fairly quickly, maybe it would be useful for us to
18
have this on the agenda for a year to have you come
19
back and say here is what we have found since last
20
year and, again, I just put that on the table for
21
others on the committee to comment on.
Dr. Nelson,
22 you
always have a comment to my comments.
300
1
DR. NELSON: Well, I was thinking
that the
2
data look like it needed to be updated.
If I
3
recall, it was 2001 cleaned and then 2001 to 2004
4
uncleaned. So, at the very least,
an update of the
5
2004 data cleaned would probably make sense. So, I
6
guess I am in agreement that that might be a useful
7
thing to do and then present.
8
DR. CHESNEY: Thank you. We agree on
9
that. Does anybody disagree with
that?
10
[No response]
11
That is number one. Number two,
12
population-based prospective studies, I think we
13 all
feel that that is absolutely important and
14
essential through whatever mechanism we can come up
15
with.
16
Retrospective studies--there are some out
17
there now. They have alerted us
to the problem. I
18
don't know how much more information they could
19
give us because we don't know what to look for now,
20 let
alone what to look for retrospectively.
I am
21
just free-associating and then I will let other
22
people comment.
301
1
A randomized, controlled trial of
2 treatment
of maternal depression and what research
3
questions should be addressed--that is much more
4
difficult. Dr. Wisner?
5
DR. WISNER: If what you mean is a
6
drug-placebo, controlled study during pregnancy, it
7 wouldn't
be funded by NIH. One of the
8
possibilities is a drug-other treatment control,
9 and
there are investigators who have tried to
10
compare drug treatment versus psychotherapy and
11
have submitted such studies but for a whole number
12 of
methodological and ethical reasons they have not
13
been funded by NIMH.
14
What we are doing is a study of light
15
therapy. We have done two pilot
studies and hope
16 to
present enough evidence that it is an effective
17
treatment to do a light therapy versus drug
18
treatment when then could potentially give us the
19
chance to look at drug versus another active
20
treatment on the kinds of neonatal outcomes that we
21
have been talking about today.
22
DR. CHESNEY: What do we know
about
302
1
neonatal outcomes of women who had severe
2
depression before we had effective drugs? Do we
3
know anything about those neonates?
How they
4
behaved?
5
DR. WISNER: Other than being
described as
6
irritable and difficult to console--you mean
7
longer-term outcomes?
8
DR. CHESNEY: No, just the
immediate, just
9 the
effects of the maternal depression without any
10
therapy. They were irritable,
difficult to
11
console?
12
DR. WISNER: Yes, and can have
long-term
13
growth and certainly socioemotional difficulties.
14 In
Lynne Singer's data set in which she looked at
15
women who had abused various substances in
16
pregnancy, the motor and more physical effects on
17
kids long term could be related to drug use, but
18 the
socioemotional development was pretty highly
19
correlated with maternal depression score, which is
20
kind of interesting. The effects
on development
21 are
pretty devastating.
22
DR. CHESNEY: Fascinating. What research
303
1
question should be addressed by the trial? Shall
2 we
continue to pursue that part of bullet number
3
four?
4
DR. D. MURPHY: I think we have
enough
5
here to work with.
6
DR. CHESNEY: I am hearing no's
all around
7
me. I hope you could hear
that. Thank you,
8
everybody for getting us through all of that. I
9
think, unless Tom is signaling me something else,
10 we
need to move on now to Dr. Iyasu's presentation
11 on an update on congenital eye malformations
in
12
infants.
13
Update on Congenital Eye Malformations in Infants
14
DR. IYASU: Good afternoon
again. This
15
will not take a very long time and you have the
16
break after that.
17
I am going to discuss congenital eye
18
malformations reported through AERS with the
19
maternal use of antidepressants during pregnancy.
20
First I would like to acknowledge Kate Phelan from
21 the
Office of Drug Safety for performing the
22
primary review of adverse event reports.
I think
304
1
Kate is still here.
2
To provide you with some background,
3
during the February, 2003 meeting of this committee
4 we
reported a case report of a potential eye
5
malformation related to the use of citalopram
6
during pregnancy. Namely, it was
a patient with
7
ptosis, eye muscle paresis and nystagmus. At that
8
time an expanded review of citalopram and several
9
other antidepressants for potential reports of eye
10
malformations were under review.
The review has
11
been completed and today's talk is an update of
12
congenital eye malformations for citalopram and its
13
enantiomer, escitalopram and several other newer
14
antidepressants.
15
In March, 2002 the WHO Upsala Monitoring
16
Center of Drug Safety published three possible
17
reports of congenital eye malformations with the
18 use
of citalopram during pregnancy. Two of
these
19
reports were congenital optic nerve hypoplasia.
20
Both were from Sweden. The third
was a report of a
21
non-specific eye malformation from Great Britain.
22 All
were exposed to citalopram during the first
305
1
trimester. The publication of
this report
2
triggered an FDA review of the AERS database.
3
First I will just give you some background
4
again about this drug. I will
discuss relevant
5
labeling for the drug products included in this
6
current review. Citalopram is
labeled as a
7
pregnancy category C drug. In rat
embryo or fetal
8
development studies teratogenic effects have been
9
reported at maternally toxic doses, and this
10
included decreased embryo or fetal growth,
11
decreased survival and increased incidence of
12
cardiovascular/skeletal defects.
However, this did
13 not
include any teratogenic effects on the eye.
14
Fluoxetine, flovoxamine, paroxetine,
15
sertraline and venlafaxine are labeled as pregnancy
16
category C drugs. No teratogenic
effects have been
17
seen with these drugs, except decreased pup
18
survival in rats. Like all
antidepressants, the
19
label also recommends use of these drug products
20
during pregnancy only if the benefit outweighs the
21
risk to the fetus.
22
The other drug that was reviewed was
306
1
bupropion which is labeled as a pregnancy category
2 B
drug. No teratogenic effects in rat
studies have
3
been reported.
4
The last medication is desipramine which
5 has
no pregnancy category on the label but does
6
carry a warning about use in pregnancy, like
7
antidepressants, and reproductive studies are
8
reported to be inconclusive.
9
Now on to the search strategy of the AERS
10
database, the Office of Drug Safety searched the
11
AERS database for reports of "eye disorders,
12
congenital" in relation to
citalopram,
13
escitalopram and the other drugs that I mentioned
14
before.
15
The AERS search results for citalopram
16
revealed that there were 5 unduplicated pediatric
17 eye
malformations. One was a U.S. case; 4
were
18
international reports. Only one
congenital optic
19
nerve hypoplasia, reported in the WHO bulletin, was
20
found in the AERS database. There
were no adverse
21
event reports for escitalopram.
22
Of the 5 reports that are in the AERS
307
1 database,
one was a congenital optic nerve
2
hypoplasia and there were other medications also
3
used concomitantly during pregnancy, cefuroxime and
4
nitrofurantoin for urinary tract infection about
5 the
fifth month of pregnancy.
6 The second case was a non-specific
eye
7
malformation, also with multiple medications were
8
used concomitantly during pregnancy.
9
The third case is the one we reported last
10
February, which was a congenital ptosis and
11
nystagmus. The report does not
indicate any
12
concomitant medication use.
13
The fourth case is bilateral retinal
14
coloboma, right hydronephrosis, respiratory
15
distress syndrome with collapsed lung.
There were
16 no
other medications except multivitamins.
17
The last case was downward deviation of
18
gaze without paralysis. In the
case report there
19
were no concomitant medications.
No other
20
neurologic or increase in pressure was noted in
21
this patient.
22
Looking at the other search results, for
308
1
bupropion there were 2 cases, one with lacrimal
2
duct obstruction and another case of eyelid
3
malformation. Fluoxetine had 2
cases, optic nerve
4
anomaly and congenital lacrimal passage anomaly.
5 For
paroxetine there were 2 reports, retinopathy
6 and
congenital cataract. For Sertraline
there were
7 3
reports. One was an eye deformity which
was
8
non-specific; an anomaly of the orbit; and then
9
lacrimal passage anomaly.
Desipramine, fluvoxamine
10 and
venlafaxine did not reveal any case reports.
11
In conclusion, these adverse event reports
12
were reviewed extensively by the Office of Drug
13
Safety and also by the review division, as well as
14 the
ophthalmology group at FDA. The
conclusion is
15
that the report of congenital eye malformations
16
does not constitute a recognizable pattern that
17
could be attributed to the use of citalopram or any
18 of
the other antidepressants during pregnancy.
19
There were too few cases to make any significant
20
attribution or association with its use during
21
pregnancy, and there were also several concomitant
22
medications.
309
1
Therefore, we will continue, as was
2
mentioned before, with monitoring of the AERS for
3 any
additional cases of eye malformations with
4
these medications.
5
DR. CHESNEY: Thank you. Any technical
6
questions for Dr. Iyasu?
7
[No response]
8 Open Public Hearing
9
Thank you very much. We do have
one
10
speaker for the open public hearing today and I do
11
have something I have to read before that. Both
12 the
Food and Drug Administration and the public
13
believe in a transparent process for information
14
gathering and decision making. To
ensure such
15
transparency at the open public hearing session of
16 the
advisory committee meeting, the FDA believes
17
that it is important to understand the context of
18 an
individual's presentation.
19
For this reason, the FDA encourages you,
20 the
open public hearing speaker, at the beginning
21 of
your written or oral statement to advise the
22
committee of any financial relationship that you
310
1 may
have with any company or any group that is
2
likely to be impacted by the topic of this meeting.
3 For
example, this financial information may include
4 a
company's or a group's payment of your travel,
5
lodging or other expenses in connection with your
6
attendance. Likewise, the FDA
encourages you at
7 the
beginning of your statement to advise the
8
committee if you do not have any such financial
9
relationships. If you choose not
to address this
10
issue of financial relationships at the beginning
11 of
your statement, it will not preclude you from
12
speaking.
13
Our speaker is Dr. Philip Sandy Zeskind,
14 who has provided us with a set of his slides
in our
15
packet. Also, Tom tells me, one
of his papers is
16 in
the blue book material we received before coming
17 to
the meeting. Dr. Zeskind?
18
DR. ZESKIND: Thank you very much. While
19 Tom
is coming up to rescue me, I am not funded by
20 any
drug companies. There is no conflict of
21
interest there. In my role as
director of
22
neurodevelopmental research in my hospital, they
311
1 are
supporting my transportation here.
2
I have a series of comments and I am going
3 to
try and whip through this pretty quickly.
After
4
hearing the discussion, some of the issues that are
5
embedded in this presentation directly address some
6 of
the questions that you are asking about what
7
kinds of questions should be asked and what kinds
8 of
research methods should be done.
9
This is in reference to an article that I
10
published in the journal Pediatrics in February of
11
2004. I will whip through the
stuff that is
12
obvious, that there is a lot of depression.
13
Depression in and of itself, as Dr. Wisner said,
14
does have debilitating effects.
Serotonin is a
15
neurotransmitter that is going to affect
16
development. In my view, it is
not whether it
17
affects development but how much and in what ways.
18
That, to me, is a given as someone who studies
19
prenatal development.
20
Unfortunately, the way these questions
21
have been answered in the past is by using measures
22
such as birth weight, gestational age and physical
312
1
anomalies. Quite honestly, in my
view, these are
2 the
measures that Sparta used to see if a baby was
3
healthy 3,000 years ago and I think that, as a
4
field of research, we have moved on way beyond that
5 yet
we are not applying it to some of these
6
important questions that we have in front of us
7
today.
8
For the study that we did, the issues that
9 are
relevant for us, especially if you are talking
10
about doing retrospective studies and big
11
population-based studies--all infants in this study
12
were full birth weight. Except
for one infant,
13
they were all full term, and no infants had any
14
physical anomalies. There were
absolutely no
15
differences between the SSRI-exposed and the
16
non-exposed infants. It is a
small sample, yet it
17 was
an intensively studied sample. If
anybody went
18
back to the medical records on these infants, they
19
would find "no effects of SSRI exposure." That is
20 the
bottom line.
21
What we found, through methods that are
22
described in the paper, is quite a list of
313
1
neurobehavioral differences in the prenatal
2
SSRI-exposed babies They showed
incredible amounts
3 of
tremulousness that was not picked up by the
4
attending physicians; increased startles or some
5
call them arousals. There was
also independent
6
motor activity besides those. The
whole
7
sleep-state architecture of the infant was totally
8
disrupted. We sat and we watched
these babies
9
sleep and recorded blinded measures of state
10
regulation in a way that we have developed over 30
11
years and others have used as part of standard
12
newborn exams. We measured
increased REM sleep;
13
rigid state organization and depressed range of
14
states. Normal babies should get
up and cry.
15
These babies are functionally, physiologically,
16
behaviorally depressed.
17
We have also worked out a way of spectrum
18
analyzing the infants' heart rate variability to
19
look at oscillations in the heart rate.
What we
20
found is that the oscillations in heart rate over
21
time are totally messed up, just to put it
22
colloquially. They are not
rhythmic. The
314
1
parasympathetic and sympathetic nervous system is
2
disrupted seriously.
3
Importantly, all these measures have been
4
previously used to detect effects of prenatal drug
5
exposure or differentiate high risk infants in the
6
past. These are not new
measures. They have been
7
used to assess cocaine-exposed infants,
8
cigarette-exposed infants, alcohol-exposed infants,
9
prenatal malnutrition, etc. They
are the same
10
kinds of behaviors.
11
We also found, as far as
12
neurodevelopmental effects, a lower gestational age
13 by
one week even within a full-term sample.
Again,
14
going back to medical records which can't use NICU
15
admissions as a measure, yet, there was seemingly
16 an
effect on gestational age.
17
As far as whether these have effects on
18
subsequent development, I want to address this
19
because this came up and Dr. Wisner answered it
20
very nicely, that is, the long-term studies of SSRI
21
exposure have focused on the equivalent of using
22
birth weight an physical anomalies of the newborn.
315
1
Doing standard IQ tests at 2 years of age is not a
2 way
of measuring long-term effects, and that is
3
what has been used, standard IQ tests and language
4
development tests. They will not
detect effects of
5
SSRI exposure. The cocaine
literature, the
6
cigarette smoking literature knows this already.
7
They have moved on from that kind of analysis. Now
8
what people look at is socioemotional regulation,
9 how
people handle emotional issues, regulation of
10
arousal, those kinds of things.
11
I will say one more thing with that, the
12
measures that we have found, these neurobehavioral
13
measures in my own work and others' work have been
14
predictive of subsequent differences in
15
development. I don't think we can
say that there
16
aren't differences at this point.
We still need,
17 of
course, better research to look at it directly.
18
I will say also that since I published
19
that article I have received a plethora of letters,
20
unsolicited emails and letters from parents saying,
21
"my God, I'm glad someone finally said this because
22 my
baby, at 2 years of age, is having these motor
316
1
tremors and my doctor says it looks normal, but
2
there's something not right about my child." I
3
have a list of emails and letters from parents.
4
Again, that is a biased sample, self-selected,
5
however, we need to throw that into the hopper
6
here.
7
I really enjoyed listening to the
8
discussion. I don't know if we
should just call
9
this withdrawal syndrome. We
heard whether this is
10
serotonin toxicity. Serotonin is
the precursor for
11
synaptic development. I don't
know how being
12
bathed in extra serotonin for nine months during
13
gestation would not have some kind of serious
14
long-term detrimental effect. It
just escapes me.
15
My conclusions are that what we have found
16 is
that prenatal exposure to SARIS during pregnancy
17
disrupts neurobehavioral development.
I think we
18
have clear evidence of that. I
don't think that
19
birth weight, pre-term birth, NICU admission and
20
physical anomalies are sufficient measures of the
21
effects, and I think that if we continue to do that
22 we
will be missing the boat. Number three,
there
317
1 may
be neurotoxic effects; it may not just be
2
withdrawal. I think number four
is obvious.
3
In conclusion, I think when we were asking
4
here what should we do about this, I think it is a
5
question of balance. As Dr.
Wisner said,
6
depression during pregnancy is a serious problem in
7 and
of itself. For me, talking to the patients
8
that I examine, SSRIs have been given out pretty
9
much like M&Ms during pregnancy.
I think it is a
10
question of balance and concern.
"Oh, you're
11
feeling a little bit down? Here,
have this," I do
12
believe that characterizes some of the
13
administration of this drug.
14
So, we don't want to throw the baby out
15
with the bath water with this, but I don't think it
16 is
safe for us to conclude, well, don't worry; it
17 is
only a transient effect. It is only
withdrawal.
18 The
baby will get over it or there are no effects
19
because the baby is full birth weight and full
20
term. I will stop there and thank
you very much.
21
DR. CHESNEY: Thank you very
much. We
22
really appreciate your perspective of many years of
318
1
having looked at this very issue that, obviously,
2
many of us are coming at from a much less detailed
3
background. So, we really
appreciate your input.
4 Are
there any questions of Dr. Zeskind? Dr.
Gorman
5 and
then Dr. Wisner.
6
DR. GORMAN: SSRIs have now been
out in
7 the
population for ten years. I assume
pregnant
8
women start taking them whether they know they are
9
pregnant or not near the beginning.
What epidemic
10 are
we seeing today, in your opinion, that has been
11
predicated on this use? And, it
doesn't have to be
12 an
epidemic of such, you know, is this why all the
13
patients in my practice use Game Boys--
14
[Laughter]
15
DR. ZESKIND: You joke--
16
DR. GORMAN: I am not joking.
17
DR. ZESKIND: We are talking about
18
emotional regulation. You know,
one of the
19
long-term effects, now that we know about prenatal
20
cigarette exposure for example, is some very nice
21
research that shows it is attention deficit. Where
22 did
that come from? I don't know what the
epidemic
319
1 is,
but I do know, as you said and it is not an
2
exaggeration; I run a child clinic at my hospital
3 and
there are a lot of children with regulation
4 disorders
that are associated with things that moms
5
take during pregnancy, including subclinical
6
effects of alcohol, cigarettes, SSRIs.
I think we
7
have a lot of children. Where is
all the
8
depression coming from that makes the headlines of
9
Newsweek magazine for bipolar disorder?
We are
10
creating children that just, by the amount of it,
11
appear as "normal" in a statistical sense.
12
That is the best answer I can give and I
13
think it is a very good question.
But I don't
14
think if we give the kid IQ tests and they appear
15
normal, then we should conclude there is no effect
16 on
develop.
17
DR. CHESNEY: Dr. Wisner?
18
DR. WISNER: Sandy, I wonder in your
19
application of the wonderful measures you have
20
developed, you have looked at the SSRI cases but,
21 as
I recall, you have a parallel literature on the
22
effects of depression as well.
Right? I mean,
320
1
moms who are depressed give birth independent of
2
SSRIs. Is there a difference on
those outcomes?
3
DR. ZESKIND: That is a great
question,
4
Kathy. From my clinical
experience, the effects of
5
depression are different than the effects of SSRI
6
exposure. The study would have,
of course,
7
benefited greatly by having an untreated depressed
8
group. I mean, that is obviously
the next question
9
that needs to be answered so I know the limitation
10 of
the study in that sense. But these do
not look
11
like infants of depressed moms; they look like a
12
different kind of issue.
13
DR. WISNER: And just to add to
that,
14
Sandy has a wonderful cry analytic procedure that
15 is
being done with RR1 so that at birth and 2 weeks
16 the
cries of all kids in those groups will go to
17
Sandy for analysis.
18
DR. ZESKIND: What we do with that
is we
19
spectrum analyze the cries. This
is something we
20
have been developing over 30 years.
By spectrum
21
analyzing the cry and looking at the 4-minute
22
frequencies you can actually tell if there is a
321
1
problem with the brain stem. I am
very sure, based
2 on
some of the stuff we have already received, in
3
collaboration with others and my own work, that
4
these babies have the kinds of cries, cry
5
thresholds and sounds that are evidence of damage.
6
DR. CHESNEY: Dr. Gorman?
7
DR. GORMAN: Just to be pesky,
threshold
8
effect or non-threshold effect dose response, and
9 is
it uniform across babies or are there babies who
10 are
spared and babies who are dramatically
11
affected?
12
DR. ZESKIND: That is another good
13
question. I can't answer
that. That is not my
14
area of expertise. I believe Dr.
Oberlander--he is
15
over on the West Coast of Canada--has been looking
16 at
differences in genetic populations with
17
different cultural groups and how they metabolize
18 the
drug. There may be differences in
metabolic
19
activity that may have an effect.
My study cannot
20
address the dose response or whether an one SSRI is
21
worse than another, that kind of stuff.
22
DR. CHESNEY: Thank you again very
much
322
1 for
taking the time to come and be with us.
2
DR. ZESKIND: Thank you for having
me.
3
DR. CHESNEY: Tom is asking
whether we
4
want to take a break, and my thought is that a
5
5-7-minute break isn't really going to impact
6
traffic and it may impact traffic to the men's and
7
ladies' room. So, if is all right
with everybody,
8 I
would like to take a 7-minute break.
Plan to be
9
back here at 3:45. At that time,
Dr. Maldonado
10
also wants to give a response to some of the issues
11
raised this morning, briefly, before we move on.
12
Thank you.
13
[Brief recess]
14
DR. CHESNEY: Dr. Maldonado had
asked if
15 he
could spend just a few minutes responding to
16 some
of the issues that were raised this morning
17
that involved pharmaceutical companies.
He has
18
promised he will be brief but they are important
19 and
I think it is important for him to enlighten
20 us.
21
DR. MALDONADO: Thank you, Dr.
Chesney. I
22
know we talked a little bit about formulations and
323
1 I
will be brief on this. That is an issue
where,
2
unfortunately, the science has not evolved as
3
rapidly as in other parts of pharmaceutical
4
aspects. The reason is that
basically there are
5 two
very good solvents for a lot of the products,
6
especially in liquid. One is
water and the other
7 is
alcohols, and there are limitations with
8
alcohols. A lot of the new drugs
are not very
9
soluble in water. So, even when I
was still at the
10
agency, I remember a sponsor trying 200
11
formulations and failing in every one of them. So,
12 the
science, unfortunately, is not very conducive
13 to
producing formulations sometimes.
14
The other thing is that we talked a little
15 bit
about negative studies. That might have
16
actually a negative connotation.
Negative studies
17 are
not necessarily unsafe; they just fail to
18
demonstrate what they thought they were going to
19
demonstrate, and in Phase 1, those are exploratory
20
studies many times. I am glad
that Dr. Murphy
21
clarified that within BPCA those studies are
22
becoming public and published, if not in journals,
324
1 at
least the BPCA requires that the FDA make them
2
public. So, there is not the veil
that there used
3 to
be. This is the first time that actually
4
industry has the incentive to send so-called
5
negative studies to the FDA. Many
of those studies
6
were never sent before because they knew they were
7 not
going to be reviewed anyway.
8
But the most important thing that I want
9 you
to consider, the committee to consider and even
10 the
people from FDA--and I have to be very careful
11
because I don't want you to perceive that I am
12
trying to create a negative impact on how you do
13
business. For example, today we
saw several drugs
14 in
which the pediatric use is very minimal, 0.1
15
percent, 0.2 percent in some of them.
So, most of
16 the
drugs are used in adults. After the
meeting in
17 the
morning I inquired of one of the reviewers how
18 is
that different, how is the adverse event profile
19
different in pediatrics than it is in adults. I
20 was
told it is the same.
21
Now, we are talking about changing the
22
labels and focusing on the pediatric part. So, I
325
1
think we shouldn't do that because we are creating
2 a
perception of liability in pediatrics. If the
3
adverse events are similar in adults, let's do it
4
because of what is happening globally.
I can tell
5
you, I am an advocate for pediatric studies in my
6
company. Otherwise, when we go
back in front of
7 the
people who hold the wallet of the company,
8
there is going to be some reluctance to approve
9
pediatric studies because they are going to be
10
perceived as being a liability.
As you see here, I
11
mean it is a no-brainer, if the companies only sell
12 0.1
percent of the drug in pediatrics or 0.2
13
percent, not even 1.0 percent--actually, I had a
14
lawyer ask me at one of the labeling meetings why
15 are
we doing this to ourselves? I said, no
we are
16 not
doing this to ourselves. We are
providing this
17
information for kids. So, this is
the connotation.
18
So, if there are particular things for
19
pediatrics, frame it on pediatrics but if it is not
20
particular to pediatrics, then let's not frame it
21 in
pediatrics. Let's say, okay, these
things--for
22
example the abuse of some of the drugs, happen
326
1
actually more in bigger absolute numbers in the
2
general adult population. So,
that way we don't
3
create a perception that there is something wrong
4
with the pediatric drug development or pediatric
5 use
of these drugs because many times that is not
6 the
case. Thank you very much for the
opportunity.
7
DR. D. MURPHY: I think, Sam, we
have to
8
find for the committee your presentation on what
9
companies think about when they go through the
10
process of trying to develop drugs for pediatrics.
11 I
think that it is a very useful process for the
12
committee to be aware of.
13
DR. CHESNEY: On the next
committee, of
14
which we may not be members, we will get to hear
15
your presentation. Dr. Shirley
Murphy is going to
16 talk
to us about the Pediatric Research Equity Act.
17 Pediatric Research Equity Act
18
DR. S. MURPHY: We have heard
today about
19
what Dr. Gorman and the American Academy of
20
Pediatrics did to really lobby to get this into
21
legislation. In your packet you
have the law. It
22 is
a long way from the law to what it actually
327
1
means and how you interpret it, and what I am going
2 to
do today is just give you a very top-line
3
overview of how we are starting to interpret this
4 law
at the FDA.
5
It really takes a whole team to interpret
6 the
law, and on this team have been Terry Kwizenzi,
7
Grace Karmuz, Rosemary Addy and Rosemary Roberts.
8 It
is evolving. It is like medicine, it
almost
9
takes a case-by-case. You look at
an application
10 and
you see if it triggers PREA. You discuss
it,
11 why
it does; whey it doesn't. So, it takes a
while
12 for
precedent to be developed, just like in BPCA
13
with the written request.
14
But I will give you a very quick overview
15 of
PREA, what it means, how it compares to the Rule
16 and
really how we are interpreting at the FDA.
The
17
Pediatric Research Equity Act has lovingly been
18
called PREA, and is known throughout the FDA and
19
through the pharmaceutical world too as PREA. So,
20 it
rapidly got a nickname.
21
It became law, as you know, on December 3,
22
2003 when it was signed by the President. The
328
1
legislation mimics the Pediatric Rule with some
2
changes. It required pediatric
studies of certain
3
drugs and biological products unless they are
4
waived or deferred. It is
retroactive to all
5
applications back to April 1, 1999, and that was
6
when applications started to be triggered by the
7
Rule. So, what happens is instead
of just starting
8 the
date it was approved, it makes sure that
9
certain applications didn't fall through the cracks
10 so
nobody got an "out of jail" free card with this.
11
PREA is not applicable to drugs with
12
orphan designations or orphan applications. That
13 is
very different, as you will see, from BPCA.
14
There is a guidance under development.
Initially
15 we
had hoped that this guidance would be available
16 to
hand out to you but it is not quite baked yet.
17 It,
very importantly, establishes the pediatric
18
advisory committee.
19
How does PREA compare to the Rule?
It is
20
actually quite similar. PREA is
legislation so it
21 is
a law. So, thank you very much, American
22
Academy of Pediatrics and everyone who worked so
329
1
hard on this. The Rule was a
regulation and, as
2 you
know, the courts enjoined its enforcement.
3
PREA does not specify meetings at
4
appropriate times, although we anticipate that the
5
guidance will give some guidance about this. The
6
Rule said you should have a pre-IND meeting and
7
discuss pediatric plans. That
wasn't in PREA. It
8 is
retroactive and it does establish the advisory
9
committee.
10
Well, PREA is the return of the stick and
11
BPCA remains the carrot. You
know, why do
12
companies do studies when there is such a small
13
percentage of the patients that are taking the
14
drug? Well, it is for the billion
dollars that you
15
make on the six months. It is not
for the 0.1
16
percent of the kids that may take that drug. And,
17
BPCA remains a very successful carrot.
18
These studies are voluntary in BPCA.
They
19
include orphan drugs and orphan indications, where
20
PREA does not include orphan drugs and orphan
21
indications. Now, BPCA is
wide. It is just huge
22
wide. It covers the entire
moiety. So, if you
330
1
have a corticosteroid, for instance, like
2
fluticasone, the written request may be for the
3
lung, the nose and the skin, all of those
4
indications. PREA is very, very
narrow. The
5
studies are limited to the drug and the indication
6
that is under development. So,
that is very, very
7
different.
8
Now, a pediatric assessment is required
9 for
applications, or applications trigger PREA when
10
there is a new ingredient. So,
say, a combination
11 of
Tylenol and a muscle relaxant wanted to add
12
caffeine in, that would trigger PREA.
A new
13
indication, say, a skin steroid wanted to go for an
14
indication of eczema; a new dosage form, something
15
goes from a liquid to a chewable, dispersable
16
tablet; a new dosing regimen goes from 4 times a
17 day
to 2 times a day; or a new route of
18 administration, it goes from an IV
administration
19 to
a patch. So, these are the things that
trigger
20
PREA and require the company to have a pediatric
21
plan and do pediatric assessments.
22
A pediatric assessment--and pediatric
331
1
assessment is probably interchangeable with
2
pediatric studies--it has to be data adequate to
3
assess the safety and effectiveness of the drug or
4 the
biologic product and data to support dosing and
5
administration for each of the relevant pediatric
6
subpopulations.
7
Just like with drug development in
8
general, effectiveness can be extrapolated from
9
adequate and well-controlled studies in adults, and
10
then can be supplemented, just like we do in BPCA
11 and
the written request. Where there are
gaps, it
12 can
be supplemented with safety and PK/PD data in
13
children. You can extrapolate
from one age group
14 to
another where appropriate. So, you might
be
15
able to extrapolate from an adolescent down to a
16
child of, say, 6 but it would be a big gap to go
17
from adolescents to neonates.
18
Now, I mentioned that there could be
19
deferrals and there could be waivers.
A deferral
20 is
granted when the drug or biologic product is
21
ready for approval in adults and the pediatric
22
studies aren't completed yet. You
cannot hold up
332
1
access of medication for adults under PREA so you
2 go
ahead and approve it and then you would have
3
then you would have the adult [sic] studies come in
4 a
year or two later. Or, the FDA believes
that
5
additional safety or effectiveness data that is
6
necessary before this drug is studied in children.
7
Some sponsors will come in very eagerly at a
8
pre-IND meeting and want to start study in children
9
when adults haven't been studied and the FDA can
10 say
wait, let's get some adult data, or it can even
11 go
up to approval and say let's get it on the
12
market a while and see what happens before we
13
subject children to it. This is
really with a new
14
molecular entity most often. But
these deferrals
15 are
tracked in a database at the FDA as Phase 4
16
commitments so they don't get lost.
17
What about waivers? Well a full
waiver,
18
meaning you don't have to do pediatric studies at
19
all, are granted when a condition doesn't occur in
20
children. Prostate cancer would
be an example of
21
that. Or, necessary studies are
impossible or
22
highly impractical, and these are probably some
333
1
cases we will have to go through to see what this
2
exactly means. Or, strong
evidence suggesting a
3
drug or biologic would be ineffective or unsafe.
4 Or,
a drug or biologic does not represent a
5
meaningful therapeutic benefit over existing
6
therapies and is not likely to be used in a
7
substantial number of pediatric patients.
8
A partial waiver can be granted when there
9 is
a subset of kids that can't be studied.
That
10
might be neonates. Or, reasonable
attempts to
11
produce a pediatric formulation necessary for that
12 age
group has failed, and that gets to what Sam was
13
saying.
14
But there is a labeling requirement.
If a
15
full or partial waiver is granted because there is
16
evidence that the drug or the biologic would
17
ineffective or unsafe, the information then has to
18 be
placed in the label.
19
These are the drugs that I was talking
20
about that are new, new applications, new
21
ingredients, new formulations, new chemical
22
entities. What about already
marketed drugs, drugs
334
1
that are already out there on the market? Can they
2 be
triggered by PREA? This had the same
3
stipulation under the Rule, but the FDA never
4
invoked it and it is a very, very long, laborious
5
process under PREA in which the FDA has to notify
6 the
company, give them a chance to come in and have
7 a
meeting. Then the FDA writes a written
request.
8
Then the sponsor declines it.
Then the written
9
request is referred to the NIH Foundation and, if
10
there is no money there, then the sponsor is
11
required to do the studies. And,
if the sponsor
12
doesn't do the studies the drug can be misbranded.
13 In
that are lots and lots of meeting periods and
14
time periods. So, it would take
over a year to go
15
through this. But, as I said, it
was never invoked
16
with the Rule but it is there as, I guess, the
17
heavy part of the stick if it is really needed.
18
PREA establishes, like Dianne talked
19
today, very importantly, a full pediatric advisory
20
committee at the Office of the Commissioner with
21
very broad responsibilities that go across foods,
22
devices and biologics and lots of issues that go
335
1
across the FDA. The advisory
committee will
2
continue to have the adverse event reporting, and
3
labeling dispute resolutions will also be heard. I
4
have to say, you know, we have gone almost to the
5
line of having to bring a labeling dispute
6
resolution to the committee but, somehow, just the
7
threat of, "well, we're going to take it to the
8
advisory committee," gets those things resolved
9
in-house so I guess it is another form of a stick.
10
Subpart D referrals that Dianne has talked to you
11
about will also be part of the advisory committee.
12
In summary, we feel that PREA and BPCA,
13
just like BPCA and the Rule, really go hand-in-hand
14 to
give us new pediatric information for labeling.
15
Thank you. Any questions?
16
DR. CHESNEY: Dr. Santana?
17
DR. SANTANA: So, under PREA the
likely
18
scenario, and I will speak from the oncology point
19 of
view, is that most drugs in oncology are not
20
developed for kids; they are developed for the
21
common adult cancers, prostate, breast and so on.
22 So,
in the developmental process of those drugs if
336
1 the
sponsor knows already that they are going to
2
develop the drug for prostate and breast, then
3
there will never be pediatric studies.
Right?
4
Because those indications are not part of what they
5
ultimately want to develop their drug for. So,
6
PREA will not help us be able to study those drugs
7
effectively in children. Am I
correct?
8
DR. S. MURPHY: Well, for that
specific
9
indication--it is indication specific.
So, if they
10 are
coming in for a prostate cancer indication,
11
yes, it wouldn't. It would
probably be waived.
12 But
if they start to broaden out into solid tumors
13
that would occur in children or hematologic tumors
14
that would occur in children, then that indication
15
would trigger PREA.
16
DR. D. MURPHY: I think, Victor,
you are
17
getting at the struggle that we are aware of, and
18 one
of the reasons that there was that forum that
19 the
Academy called a number of years ago is because
20 it
was recognized that the Rule just isn't going to
21
work as well where you don't have similar diseases
22
between adults and kids. I am not
talking about
337
1
extrapolation but just talking about, you know, you
2 can
get an indication for pneumonia for adults and
3
kids. But where you don't have
that link you are
4 not
going to be able to use this hook. That
is why
5 the
exclusivity process was reevaluated as to how
6 it
can most effectively be utilized for cancer
7
development for children. That is
why there is a
8
special guidance out on how products that are being
9
developed for cancer therapies in kids could get
10
exclusivity at a stage that is less clear as where
11
they are going to go than in other products. That
12 was
a particular focus of that guidance.
13
DR. S. MURPHY: And I think
exclusivity
14 has
worked extremely well in oncology, which we saw
15
this morning. By doing really
Phase 1 and Phase 2
16
studies, without doing Phase 3 companies do get
17
exclusivity.
18
DR. SANTANA: The exclusivity only
applies
19 to
marketed drugs. Am I correct?
20
DR. S. MURPHY: Well, it can be
planned
21
premarketing. Design the studies
and plan them--
22
DR. D. MURPHY: Right.
338
1 DR. S. MURPHY: --way before.
In fact, it
2 is
included in the forecasting for products now as
3
they come out that they are going to get
4
exclusivity and how much money they are going to
5
make.
6
DR. GORMAN: One of the hopes that
some of
7 us
expressed during the creation of PREA was that
8 as
drugs go through Phase 1 testing in adults, and
9
oncology drugs would be included in this, and they
10 are
tested for mechanism of action, if the
11
mechanism of action is shared it might become clear
12
where pharmaceutical companies might go with the
13
development of agents, and that would be an
14
opportunity to initiate pediatric studies. It may
15 not
be as effective as we hope because of the
16
limitations of PREA, which is for targeted
17
indication and it may not be there, but it will
18
alert the agency, as well as the pediatric research
19
community, that these agents are coming down the
20
pike and have potential pediatric utilization.
21
DR. CHESNEY: I think Dr. Nelson
commented
22 on
this, but I am intrigued that the companies can
339
1 do
Phase 1 and 2 studies with preparations that
2
can't be used commercially. I
realize it is very
3
hard to develop those. We have
heard that over and
4
over again. Dr. Spielberg used to
talk about that
5 all
the time. But does the company have to
be
6 actively working on trying to get the
product into
7 a
commercially usable preparation at the same time
8
that they are doing Phase 1 and 2 studies?
9
DR. S. MURPHY: Well, I think it
all
10
depends on the product. A lot of
the verbiage in
11 the
law talks about the severity of the illness,
12 the
existence of other therapies, the need
13
basically and the number of patients affected, as
14 to
how early the pediatric plan would come in and
15 be
accepted and go into effect. So, I think
it is
16
really, like I said, almost like medicine. We are
17
seeing this already because Grace and Rosemary are
18 the
repository of all the questions in the FDA
19
about PREA. They come in case by
case and we are
20
actually having to decide is this good; is this
21
bad.
22
I think the pendulum, you know, we don't
340
1
want that when things are added to drugs,
2
especially generic drugs and it triggers PREA. We
3
don't want children studied with things that are
4
unsafe, that have AERS reports coming in, that are
5
probably never used in kids. So,
we have to be
6
really careful because it is a balance.
7
DR. CHESNEY: Thank you very much,
Dr.
8
Murphy. Finally, Dr. Nelson is
going to give us an
9
overview of the Institute of Medicine report, of
10
which we have a copy although I think it hasn't
11
been published yet--"Ethical Conduct of Clinical
12
Research Involving Children."
13
Overview of Institute of Medicine Report,
14
"Ethical Conduct of Clinical Research
15 Involving Children"
16
DR. NELSON: Thank you and, given the hour
17 and
time, I am going to try and go through this
18
quickly. I think all of you have
copies of the
19
slides and I would comment the full report.
20
What I would like to do--you know, we say
21 you
shouldn't look at the trees and miss the
22
forest. In this case I want to
point out some
341
1
trees and, you know, you can look at the full
2
report for the forest. So, let me
run through this
3
fairly quickly.
4
This just lists the study committee, and
5
Dick Behrman did an admirable job chairing it.
6
The study process--you see on the slide,
7 in
terms of number of committee meetings and public
8
discussions and forums. It was
publicly released
9 on
March 25th but I don't think it is yet in a form
10 you
can actually purchase and hold in your hand at
11
this point though it is available on the Internet.
12
The issue I think fundamentally--and this
13 is
from Dick Behrman's preface--is that we are
14
trying to balance providing benefit but, yet,
15
making sure that in the process of providing
16
benefit we are not inappropriately exposing
17
children to risk. That is really
the purpose of
18 the
regulations that guide our research process, to
19 try
and balance the appropriateness of the benefit
20 and
risk that is involved.
21
The charge to the committee was
22
specifically focused on clinical research. This
342
1
lists the specific topics that you can look in the
2
report for. The only comment that
I want to make
3
here is that these topics were mandated by the Best
4
Pharmaceuticals for Children Act.
The report, if
5 it
went outside those topics, in fact, couldn't
6
include it because it would be struck.
This was
7
mandated by Congress and if the committee wanted to
8
talk about something that wasn't on that list, it
9
really couldn't put it in there.
So if you don't
10 see
it, that is why it is not there.
11
Three broad things, first, good research
12 is
important for the health of children. It
is
13
pretty obvious but we wanted to strike that theme
14 at
the beginning of the report.
15
The second is that protecting children is
16
part of human subject protection overall. There
17 has
been a fair amount of criticism and concern
18
about the overall system. To the
extent that you
19
want to protect children you need to look at that
20
overall system was our second broad theme.
21
The third broad theme was that the
22
effective implementation in terms of both
343
1
protection and, I would argue, also the conduct of
2
research is appropriate expertise.
I will come
3
back to that towards the end in some of the
4
comments. But, basically,
appropriate expertise in
5 the
design, review and conduct of such research.
I
6
don't think anybody who has been part of the
7
discussion on this committee over the last five
8
years would disagree with any one of those themes.
9
Now, the summary statement was that the
10
committee felt that the regulations were by and
11
large appropriate. But the
problem with it is that
12
there is insufficient guidance about the
13
interpretation of those regulations.
It is
14
difficult to get data about that.
I mean, you look
15 at
the Adverse Event Reporting System--try and get
16
data about the IRB system, it is worse, believe it
17 or
not. It is worse. There is nothing out
there.
18
And, there is a lot of variability in the
19
application and interpretation.
20
The feeling was you are not going to
21
reduce this variability by trying to narrow down
22 the
regulations themselves, which really means
344
1
guidance. When you think about
it, it is much like
2
PREA. You have the regulations
themselves at a
3 certain general level, then you have to have
4
guidance that implements that down to the case
5
level. You are not going to solve
the problem by
6
going back to the regulations and changing them;
7
what you need is better guidance.
8
In the spirit of that, the recommendation
9 of
the report was an attempt to look at some of
10
those areas where guidance is necessary.
I am
11
going to run through these five areas quickly, just
12
highlighting some of the trees, if you will, rather
13
than the forest.
14
One of the issues that has been discussed
15
over the years is the interpretation of minimal
16
risk. In pediatrics it is
important. If it is
17
minimal risk research you are eased of certain
18
restrictions and the balancing of risk and benefit
19 is
very different. Well, does minimal risk
refer
20 to
the normal, healthy, average child on the street
21 or
does it refer to the child with leukemia?
Every
22
single commission, including the national
345
1
commission originally and the Institute of Medicine
2
report has said this should refer basically--here
3 is
the definition, by the way, which is 45 CFR
4
46.102. I apologize, I think it
is 21 CFR 54.102
5 but
I could be mistaken.
6
The committee as well said that you should
7
interpret minimal risk in relationship to the
8
normal experiences of healthy, average, normal
9
children. So, when you are
looking at the
10
definition of minimal risk, this has an impact on
11 how
IRBs would review the research. I won't
go
12
through the rest.
13
Now, the second category, which is in the
14 FDA
50.53, which is minor increase over minimal
15
risk, basically this is to be slightly more than
16
minimal risk. That doesn't really
tell you much
17 but
where this becomes important is if you want to
18 do
this kind of research--and often single-dose PK
19
studies are approved by IRBs under this minor
20
increase over minimal risk--you then get into
21
condition because this particular category of
22
research is restricted to research where a child
346
1 has
a particular condition.
2
So, where I think this becomes important
3 for
our discussion is how do you decide whether a
4
child has a condition or not.
This is how the
5
Institute of Medicine committee approached this.
6 It
should be interpreted--and this is one of the
7
trees I want to talk about--it should be
8
interpreted as referring to a specific or a set of
9
specific physical, psychological,
10
neurodevelopmental, or social characteristics--we
11 had
a lot of discussion about that word "social."
12
Should it be in there? Should it
not? What are
13 the
issues, etc.--that an established body of
14 scientific evidence--and those who wanted it
in,
15
kept it there but the issue is what is the
16
evidence. If you want to use the
social
17
characteristic what is the evidence that that is,
18 in
fact, tied to something that would negatively
19
affect children's health and well being or increase
20
their risk of developing a health problem in the
21
future?
22
So, the emphasis here is on evidence and
347
1 on
risk of development. For example, we
talked
2
about fenfluramine study in New York and the issue
3 of
whether or not you could consider being the
4
sibling of a child who is incarcerated as a
5
condition. That is not in the
report but that was
6
part of our discussion.
7
Now, personally, in looking back at this
8
consensus statement, I think there are some
9
ambiguities in it that would merit perhaps
10
revisiting by the next edition of the committee.
11
Here susceptibility to the disease I think does
12
imply this notion of risk but it is tied to this
13
notion of benefit in a way that doesn't really
14
capture condition in the same way that the
15
Institute of Medicine reported, in my view, in the
16
same way the regulations tried to capture it. So,
17 I
think this statement that is up on the web melds
18
risk of condition and benefit together in a way
19
that is ambiguous and not as helpful as it could
20 be,
looking back now four years later. So, I
think
21 it
would make sense to revisit this particular
22
statement at some future time.
348
1
The committee does call for the need for
2 the
development of guidance and I think that is
3
fairly straightforward. I might
just say that
4
there is a potential mechanism for this.
The
5
Secretary's Advisory Committee on Human Research
6 Protections
does have a pediatric working group
7
which could be one locus for that discussion, and
8
then having guidance work its way up through
9
SACHRP, which is the Secretary's advisory
10
committee, and then work with FDA and OHRP, and I
11
think there may well be a process under way to look
12 at
that.
13
Another tree is what is called component
14
analysis of risk. For research
that offers
15
benefit, the argument here is that if you have a
16
procedure that doesn't offer benefit as part of
17
it--let's say a bone marrow aspirate where the
18
oncologist clearly says this does not offer any
19
clinical benefit to that child--within the
20
pediatric regulations you are supposed to judge the
21
appropriateness of that procedure against its risk,
22
minimal risk or minor increase over minimal risk,
349
1 as
opposed to the benefit of other things that may
2 be
in that protocol, the chemotherapy. The
risk
3 and
benefit of the interventions that offer the
4
possibility of benefit are evaluated on their own
5
merits, and then the risks of procedures that don't
6
offer the prospect of direct benefit need to be
7
restricted to minimal risk or a
minor increase
8
over minimal risk. That is called
the component
9
analysis of risk.
10
Otherwise, what you could do is take a
11
very risky non-beneficial procedure, toss it in and
12
offset it with all sorts of other benefits that are
13
totally unrelated to that procedure and that is not
14
felt to be appropriate. So, it
raises an
15
interesting question about choice of control
16
groups. This is what the
Institute of Medicine
17
report says, for research involving children and a
18
placebo control group to be approved by an IRB
19
under federal regulations, either the balance of
20
potential harms and benefits for children in the
21
placebo control arm must be as favorable as those
22 for
children receiving the active, standard
350
1
treatment--that is simply restating the conditions
2 of
50.52 which is the regulation.
3
Or, the potential harms to which children
4 in
the placebo control arm would be exposed are no
5
more than minimal or involve only a minor increase
6
over minimal risk. So, what that
is saying is if
7 you
are removing the potential benefit from that
8
placebo group, then the risk that they are exposed
9 to
because of the removal of that benefit needs to
10 fit
within the minimal risk or the minor increase
11
over minimal risk.
12
That then raises the question about
13
whether one of my favorite documents, ICH-E10,
14
control of control group, how you should interpret
15
that in light of pediatrics. We
had a discussion
16 of
this on the committee, probably in 1999, at this
17
point, and one or my favorite quotes is from Bob
18
Temple saying that he doesn't think E-10 dealt with
19
pediatrics because the E-10 did not apply to
20
children because it assumed as the ethical basis
21 for
withholding effective treatment informed
22
consent. Look back at the
transcript. I will tell
351
1 you
where it is. I use it all the time when
I talk
2
about E-10. I think it is an open
question.
3
The issue of the threshold, death or
4
irreversible morbidity--I would propose to you the
5
absence of that may not be the same as minimal risk
6 or
minor increase over minimal risk. It
probably
7
isn't. So, I think that is an
open area that would
8
have to be addressed in the ethics.
So, that is
9
another tree.
10
Parental permission and child assent--this
11 is
a drum that I think ethicists are continuing to
12 hit, the notion that the FDA does not allow a
13
waiver of parental permission for conducting
14
pediatric research, and specifically said they
15
would not adopt the same waiver that is found in
16 the
HHS regulations. The report says that is
not a
17
good idea; we think they should, in fact, be the
18
same regulation.
19
The notion of harmonization is part of
20
that component, that these two regulations ought to
21 be
harmonized. The one point that is not
22
harmonized is, in fact, the waiver under 46.408(c).
352
1 A
controversial issue, but the report said that
2
should be harmonized. Will it
be? Who knows?
3
Payments--this is an opportunity to say
4
something good about industry. It
says that people
5
should be compensated for injury during trials. I
6
will point out that in most of my experience
7
industry trials do offer that kind of compensation;
8
other trials do not generally. It
does talk about
9
investigator payments where it says investigators
10 and
staff should be compensated for the costs
11
associated with conducting research.
However,
12
finders fees and those kinds of kickbacks, if you
13
will, to individuals referring subjects are
14
unethical and should not be permitted.
So, this is
15
another tree.
16
Now, there is a PhARMA principle for
17
conduct of clinical trials which does, I think, say
18
something similar, that payment to clinical
19
investigators is appropriate if it is reasonable
20 for
the work to be done; that you can, in fact,
21
provide additional payment if there is more work to
22
recruitment of subjects. You
could read this as
353
1
precluding finders fees but it doesn't come out and
2 say
that clearly. I would interpret it as
3
precluding that. Just as an
aside, so would James
4
Sheehan, the Associate U.S. Attorney of the Eastern
5
District of Pennsylvania, who often has looked at
6
industry-sponsored research who says that there is
7 no
law prohibiting payment to doctors for
8 recruiting
study subjects but a jury would find the
9
practice wrong. Even in the
absence of an
10
expressed statutory prohibition on finder fees,
11
they are problematic. I think
most industry
12
protocols do not include finders fees, in my
13
experience, and I think most sponsors would
14
interpret PhARMA's principle as excluding those but
15
they are still out there.
16
Regulatory compliance--there is a need for
17
data. We probably know more about
what is going on
18 in
the FDA arena and to some extent in the NIH
19
arena, but we have no idea what is going on in
20
pediatric research in other arenas in terms of the
21
distribution of protocols among minimal risk, minor
22
increase over minimal risk, what is happening with
354
1
investigator-sponsored, single institution, locally
2
funded--we have no idea what is going on in that
3
area. It would be helpful to know
what is
4
happening in terms of development of guidance. So,
5
that was one of the recommendations, to get better
6
data.
7
Finally, responsible conduct of research,
8
which I interpret as systems improvements, the
9
first is something that I think has been brought
10
up. It was brought up by the
original national
11
commission and again by the Institute of Medicine,
12
that federal law should require all clinical
13
research to be governed by the same set of rules.
14 The
way it is worded is conducted under the
15
oversight of a formal program for protecting human
16
participants in research. In
other words, you have
17 the
FDA and industry sponsored research; you have
18 NIH
and its funded research. But if you are
not
19
going to submit any of the data to the FDA or if
20 you
are not funded by the NIH and you are not
21
working in an institution that would require you to
22 be
obedient to those rules, you can carry research
355
1 out
in your basement and whether you are breaking
2 the
law or not is a separate question.
3
I might point out that there is a draft
4 Bill
that I believe Sen. Kennedy--I don't know if
5 it
is officially coming out in a draft but I read
6
about this in a BNA medical policy report, where he
7
wants to provide statutory authority for OHRP to,
8 in
fact, issue rules that would apply a common rule
9 to
all research, and to require all greater than
10
minimal risk research to gain approval from an
11
accredited IRB, effective by those two dates.
12
Whether this will happen or not--open question.
13 But
that is at least one attempt to move that
14
along.
15
The need for IRB expertise--I think this
16 is
one thing where IRBs, particularly small
17
community IRBs, may struggle with.
I would propose
18
that this could be something that the FDA audit
19
procedure could, in fact, look at if they wanted
20
to. IRBs reviewing pediatric
protocols should have
21
adequate expertise in child healthcare and
22
research. At least three
individuals with such
356
1
expertise present and voting, and among the
2
pediatric clinical care research, psychosocial
3
aspects of child and adolescent healthcare and
4
research, and then ethics of research involving
5
children.
6
Where did we get this list of three?
7
Well, if you look at ICH-E11, at the end it says
8
there should be adequate expertise on an IRB in
9
these three areas, but it doesn't translate that to
10
three people. We had a lot of
discussion about
11
this. Is one enough? Is five too many? We just
12
decided to say you should have three because often
13 if
you are on a general IRB with only one
14
pediatrician it is very easy for that voice to be
15
drowned out. I think there are
many IRBs that
16
might struggle with this particular requirement.
17 We
also say they should consult with other child
18
healthcare experts, parents, children, etc.,
19
relevant family and community perspectives.
20
It goes on and actually you could say
21
potentially it would impact the pediatric advisory
22
committee. The Institute of
Medicine report
357
1
advises that standing pediatric advisory committees
2 and
pediatric IRBs, but standing pediatric advisory
3
committees include at least one non-scientific,
4
unaffiliated member who can represent explicitly
5 the
perspectives of parents and children.
6
I would propose this is not quite the same
7 as
a consumer perspective. The argument
here was
8
that this is very much a sort of participant
9
standpoint perspective as far as those who would
10
potentially be the subjects of this kind of
11
research, which would be very different than a
12
consumer perspective. So, that is
one of those
13
recommendations.
14
Multicenter studies in terms of
15
coordination was another recommendation.
Here is
16
another guidance. This is another
tree I would
17
like to call your attention to.
Ideally, there
18
would be coordinated guidance among NIH, FDA, OHRP
19 and
the like, and that doesn't currently exist.
20 Let
me give you one example that I came upon when I
21 was
reviewing the NIH guidance on the inclusion of
22
children in research for another talk.
358
1
You heard earlier today that the FDA does
2 not
require sponsors to do efficacy studies if, in
3
fact, you can extrapolate data from the adult to
4
children. The ethical argument is
that it is
5
really unnecessary to expose children to the risk
6 of
that research if, in fact, the efficacy data can
7 be
extrapolated and you would only need PK data or
8
safety data. I think that has
been a fairly
9
consistent approach over the last decade. But if
10 you
look at the NIH guidance on the inclusion of
11
children, it actually says if the disease is the
12
same you can include children and you don't even
13
need to include enough children to do a meaningful
14
subgroup analysis. So, it is the
opposite and I
15
think the FDA approach is right; I think the NIH
16
approach is wrong. In fact, you
shouldn't include
17
children in the research unless you can have
18
meaningful data about them as a population. It
19
raises the same issue as the inclusion of women so
20 it
is a broader issue than that, but if you look at
21
that guidance, it is the exact opposite.
22
So, I think that is something that ought
359
1 to
be discussed between the two, and I might point
2 out
there was recently an RFA for inclusion of
3
adolescents in sleep studies but then a proposal
4
failed to get past what would be equivalent of a
5
50.54 review because the review panel and then the
6
OHRP and the Secretary agreed that it was unethical
7 to
do the study in adolescents even though the RFA
8
specifically asked for adolescents to be included.
9
Finally, 50.54 which was alluded to.
Here
10 I
think again the FDA has a leg up on the process.
11 It
has a federally mandated public, accessible
12
advisory committee where, if there is a 50.54
13
application that the sponsor then pursues and
14
doesn't take off the table because they don't want
15 the
publicity, and that is a whole separate
16
question, there is now a venue for that to happen
17
which is open and publicly accessible.
To date,
18
that has not been available. To
date, all of the
19 reviews that have happened within OHRP have
been
20
done non-publicly by individuals offering opinions
21
individually to the OHRP, which then correlated it
22 all
and went through that process.
360
1
There is a proposal that, in fact, that
2
will be done publicly but OHRP can't establish a
3
FACA committee so it will be a public, non-FACA
4
process. If you can imagine, all
of us, after we
5 have
our discussion, we leave the room, write our
6
individual reports and we send them in to the
7
office and then they try to collate that. That
8
would be the equivalent. Here now
at least there
9 is
a process that the FDA has where they can do
10
that if such request comes up. I
know there are
11
discussions about how to coordinate that if there
12 is
a coordinated product that would be both NIH
13
reviewed and FDA regulated.
14
So, I have kind of given you a whirlwind
15
tour. Hopefully, you can see the
forest but I
16
wanted to point out where I see some trees that are
17 of
interest and the impact of this report
18
potentially on FDA activities and the like, and in
19 may
ways, I hope in a positive sense, where there
20 are
some things that are being done well and some
21
things that could be done better.
22
Again to remind you of the three-part
361
1
theme, clinical research is essential to improve
2 the
health of future children; a robust system is
3
necessary for protecting child research
4
participants; and effective protection requires
5
expertise in child health at all stages of the
6
design, review and conduct of such research. I
7
hope that hasn't been too fast.
8
DR. CHESNEY: Outstanding. Any technical
9
questions for Dr. Nelson? Dr.
Luban?
10
DR. LUBAN: I don't know if you
would
11
define this as technical, but could you flesh out a
12
little bit more some of the discussion on the
13
multicenter studies? That appears
to be, at least
14 at
our institution, just one of the most difficult
15
problems to deal with, particularly when they are
16
NIH-sponsored and large and excessively
17
multi-institutional. DR.
NELSON: I guess the
18
committee felt since the institute process is meant
19 to
be evidence-based, and there are a lot of models
20 out
there in terms of independent IRBs that are
21
often used by industry; central IRBs that are used
22 by
cooperative groups--I know there is a National
362
1 Cancer Institute initiative that is primarily
in
2
adults; there are a lot of initiatives out
3
there--the feeling was that right now there is not
4
good data to say which of those models is best.
5
Part of that was also recognition that many
6
institutions are very reluctant to hand over
7
authority or responsibility for some of that
8
decision-making. So, one of the
obstacles to
9
centralization is often local concerns about
10
liability. Basically, the report
runs through some
11 of
those issues and suggests that we just need more
12
work done in sorting out what is the best way to
13
conduct those multicenter studies.
I have my own
14
bias about that, but we didn't feel there was any
15 clear winner in all of that to be able to say,
from
16 an
evidence perspective, what is the best approach.
17
DR. CHESNEY: Local concerns about
18
liability.
19
DR. GORMAN: Just to reengage the
20
discussion on minimal risk, which average, healthy,
21
normal children were you planning on using? Would
22
that be country dependent or the traditional
363
1
suburban child in east Philadelphia or west
2
Philadelphia?
3
DR. NELSON: You know, if you look
at the
4
report, the way I would answer that is one of the
5
principles that I think has been under-discussed in
6
research ethics is justice. What
we are really
7
talking about is under what conditions is it
8
appropriate to expose a child to increased risk.
9 Is
one of those conditions that you happen to live
10 in
inner city Baltimore in lead-affected housing?
11
Well, maybe in one protocol the answer is yes and
12 in
another protocol the answer is no. So,
figuring
13 out
the notion of condition requires both evidence
14 and
then I think an understanding of the risk
15
within that. With average,
healthy, normal
16 children the intent was to not use what would
be
17
considered research irrelevant characteristics to
18
justify increased risk exposure.
You know, if you
19
were going to do non-lead related research you
20
wouldn't use the risk of living in Baltimore as a
21
justification for going into that population to do
22
something riskier than you wouldn't do in the
364
1
suburb that you live in, for example, maybe. So,
2
that would have to be the same.
3
On the other hand, if you define the
4
condition, then there might be a justification for
5 a
protocol that would have an increased risk
6
exposure in that population. So,
that is part of
7 the
balancing that we went through. So,
average,
8
normal, healthy--I don't think we defined it but,
9 on
the other hand, since it is not defined at all I
10
think the first step is to at least agree that that
11 is
what we are talking about. Right now you
could
12
define as minimal risk, although most IRBs don't, a
13
child who has leukemia, if you wanted to, in terms
14 of
their daily experience. I don't think
IRBs do
15
that but they could.
16
DR. CHESNEY: Thank you very much. I
17
think Dr. Buckman had a comment that she wanted to
18
make in response to Dr. Maldonado's comments.
19
DR. BUCKMAN: Sorry, you probably
thought
20 you
had finished hearing from me today. Just
very
21
briefly, I just wanted to bring up one point of
22
clarification. As you know, the
FDA is all about
365
1 the
details. It is just responding to a
comment
2
that was made a little bit earlier and maybe a
3
concern that was raised to me regarding adverse
4
events in the pediatric population versus the adult
5
population for the Duragesic patch, and whether
6
there were similarities or differences, and I
7
wanted to give the exact information because I
8
didn't want it to go into the record without it
9
being very clear.
10
DR. CHESNEY: Do I need to recuse
myself
11
from listening to this?
12
DR. BUCKMAN: I don't think
so. No
13
questions; this is just a point of clarification.
14 The
top 20 reported adverse events in the adult
15
population were compared to the pediatric
16
population for that 1-year post-exclusivity period
17 for
Duragesic. Of the events that were
captured in
18 the
adult population, there were 4 unique events
19
that were captured in the pediatric population that
20
were not seen in the top 20 for the adults. I just
21
want to read what those were: cardiac arrest;
22
respiratory arrest; self-medication; and anxious
366
1
parent. Those were not captured
in the top 20
2
adverse events for the adult population.
3
Now, I cannot say whether below the top 20
4
those adverse events were also captured in the
5
adult population, but I just wanted to give that as
6 a
point of clarification.
7
DR. CHESNEY: Thank you.
8 DR. SANTANA: So, besides the message that
9
there were some differences, the underlying message
10 is
that when you look at pediatric adverse events
11 for
these drugs you also look at the adverse event
12
reporting for adults and do a backside comparison.
13
DR. BUCKMAN: Right, right, we try
to.
14 You
know, we look for similarities and differences.
15 I
don't want to go on the record making a global
16
statement that they are the same or that they are
17
different, but there are some similarities and
18
there may be some differences as well.
19
DR. MALDONADO: I wasn't focusing
on that
20
particular drug because I heard the same thing for
21
Effexor. So it is in general. I mean, I said
22
don't frame it in pediatrics if it is not only a
367
1
pediatric problem; just frame it generally because
2
otherwise people will focus and say this is a
3
pediatric liability issue. It is
general. That is
4
all.
5
DR. BUCKMAN: The point is well
taken.
6
DR. SANTANA: Yes, it is, but we
did see a
7
couple of examples of the drugs that were reviewed
8
this morning in which they were very similar and no
9
issues were made of that. I
remember at least two
10 of
the oncology drugs in which the profiles were
11
very different and no further issues were created
12
because of that analysis.
13 DR. BUCKMAN: Thank you.
14
DR. CHESNEY: Well, thank you,
all, very,
15
very much. Tom tells me that the
vans to take
16
anybody wherever they want to go are outside. I
17
don't know what to say except thank you, all,
18
again.
19
DR. D. MURPHY: Thank you, all.
20
[Whereupon, at 4:42 p.m., the proceedings
21
were adjourned.]
22 - - -