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
Tuesday, February 3,
2004
9:00 a.m.
Advisors and Consultants Staff
Conference Room
5630 Fishers Lane
Rockville, Maryland
2
PARTICIPANTS
P. Joan Chesney, M.D., Chair
Thomas H. Perez, MPH, Executive Secretary
SGE CONSULTANTS (VOTING):
Mark Hudak, M.D.
David Danford, M.D.
Richard Gorman, M.D., FAAP
Robert Nelson, M.D., Ph.D.
Susan Fuchs, M.D.
Robert Fink, M.D.
Victor Santana, M.D.
Norman Fost, M.D., MPH
Judith O'Fallon, Ph.D.
Ralph D'Agostino, Ph.D.
Mark Fogel, M.D.
Tal Geva, M.D.
Craig Sable, M.D.
Vasken Dilsizian, M.D.
Marilyn Siegel, M.D.
Phillip Moore, M.D.
MEMBERS (VOTING):
Mary Glode, M.D.
Steven Ebert, Pharm.D. (Consumer
Representative)
FEDERAL EMPLOYEE (VOTING):
Mario Stylianou, Ph.D.
INDUSTRY REPRESENTATIVE:
Samuel Maldonado, M.D.
FDA:
Shirley Murphy, M.D.
Solomon Iyasu, M.D.
Hari Sachs, M.D.
Julie Beitz, M.D.
Sally Loewke, M.D.
Shavhree Buckley, 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.,
Executive Secretary 8
Welcome, Rosemary Roberts, M.D., Office
of
Counterterrorism and Pediatric Drug
Development 11
Adverse Event Reports Per Section 17 of
BPCA,
Solomon Iyasu, M.D., Division of
Pediatric
Drug Development 12
Use of Imaging Drugs in Conjunction
with Cardiac
Imaging Procedures in the Pediatric
Population:
Pediatric Regulatory Update, Susan
Cummins, M.D.,
Division of Pediatric Drug
Development 58
FDA Perspective, Sally Loewke, M.D.,
Division of
Medical Imaging and
Radiopharmaceutical Drug
Products 73
American Academy of Pediatrics
Perspective,
John Ring, M.D., University of
Tennessee
Health Science Center 91
Cardiologist Perspective, Tel Geva, M.D.,
Children's Hospital Boston 106
Q&A for Speakers 126
Contrast Enhanced Magnetic Resonance
Imaging,
Mark Fogel, M.D., The Children's
Hospital
Philadelphia 143
Contrast Enhanced Cardiac Computed
Tomography,
Marilyn Siegel, M.D., Washington
University
School of Medicine 172
Contrast Enhanced Invasive Cardiac
Imaging,
Phillip Moore, M.D., UCSF Children's
Hospital 174
Contrast Enhanced Cardiac Ultrasound,
Craig Sable,
M.D., Children's National Medical
Center 213
Radiopharmaceuticals in Nuclear Cardiac
Imaging,
Vasken Dilsizian, M.D., University of
Maryland
School of Medicine 234
Q&A for Speakers 253
4
C O N T E N T S (Continued)
Open Public Hearing:
Michael J. Gelfand, M.D., Children's
Hospital,
Cincinnati 296
Manuel D. Cerqueira, M.D.,
American Society of Nuclear
Cardiology 311
Peter Gardiner, MB ChB, MRCP,
Bristol-Myers Squibb 316
Jack Rychik, M.D., American Society of
Echocardiography 320
5
1 P R O C E E D I N G S
2 Call to Order and
Introductions
3
DR. CHESNEY: Good morning and
welcome to
4
what should be a very fascinating day and a half.
5 I
would like to start by saying that there is the
6
potential for us to finish our work today if we
7
stay very focused and very attentive to the
8
specific issues that the FDA is asking us to
9
address. But first we need to
have the
10
introductions and I think maybe we could start with
11 Dr.
Maldonado and go around this way, please.
12
DR. MALDONADO: Samuel Maldonado,
from
13
Johnson & Johnson.
14
DR. MOORE: Phillip Moore, from
the
15 University of California San Francisco,
pediatric
16
cardiology.
17
DR. SIEGEL: Marilyn Siegel, from
18
Washington University in St. Louis, pediatric
19
radiologist.
20
DR. DILSIZIAN: Vasken Dilsizian,
21
University of Maryland, Director of Nuclear
22
Cardiology, both adult and cardiology and nuclear
23
medicine.
24
DR. SABLE: Craig Sable,
Children's
25
National Medical Center in Washington, Director of
6
1
Echocardiography.
2
DR. GEVA: Tel Geva, Department of
3
Cardiology at Children's Hospital in Boston.
4
DR. D'AGOSTINO: Ralph D'Agostino,
Boston
5
University, statistician.
6
DR. FOGEL: Mark Fogel, pediatric
7
cardiology, Children's Hospital, Philadelphia.
8
DR. SANTANA: Victor Santana,
pediatric
9
hematologist, oncologist at St. Jude's Children's
10
Research Hospital in Memphis, Tennessee.
11
DR. GORMAN: Rich Gorman,
pediatrician,
12
private practice, Ellicott City, Maryland.
13
DR. EBERT: Steve Ebert,
infectious
14
disease pharmacist, Meriter Hospital, Professor of
15
Pharmacy, University of Wisconsin, Madison.
16
MR. PEREZ: Tom Perez, executive
secretary
17 to
this meeting.
18
DR. CHESNEY: Joan Chesney,
Professor of
19
Pediatrics at the University of Tennessee in
20
Memphis and also at St. Jude's Children's Research
21
Hospital.
22
DR. FOST: Norm Fost, Professor of
23
Pediatrics and Director of the Beioethics Program
24 at
the University of Wisconsin, Madison.
25
DR. NELSON: Robert Nelson,
Critical Care
7
1
Medicine, Children's Hospital, Philadelphia.
2
DR. FINK: Bob Fink, pediatric
3
pulmanology, Professor of Pediatrics, Children's
4
Medical Center, Dayton, Ohio.
5
DR. O'FALLON: Judith O'Fallon,
6
biostatistician, recently retired from the Mayo
7
Clinic.
8
DR. FUCHS: Susan Fuchs, pediatric
9
emergency medicine, Children's Memorial Hospital,
10
Chicago.
11
DR. DANFORD: Dave Danford,
Professor of
12
Pediatrics, Section of Cardiology, University of
13
Nebraska Medical Center and Crayton University in
14
Omaha.
15
DR. GLODE: Mimi Glode, pediatric
16
infectious disease at Children's Hospital,
17
University of Colorado in Denver.
18
DR. HUDAK: Mark Hudak, Professor
of
19
Pediatrics and Neonatology, University of Florida,
20
Jacksonville.
21
DR. SACHS: Hari Sachs, Professor
of
22
Pediatrics and medical officer at FDA.
23
DR. IYASU: Solomon Iyasu. I am team
24
leader at the FDA.
25
DR. S. MURPHY: Shirley Murphy,
the "other
8
1
Murphy." I am the Director
of the Division of
2
Pediatric Drug Development and I am going to be
3
sitting here today because the "other Murphy" may
4
have to deal with counterterrorism.
5
DR. CHESNEY: Thank you and we
6
particularly welcome our cardiology and imaging
7
consultants so that we have some expertise on the
8
committee. We are going to be
very dependent on
9 you
to talk to us about degrading nuclear particles
10 and
so on in the major session for this morning.
11 But
next we would like Tom to give us the meeting
12
statement, please.
13 Meeting Statement
14
MR. PEREZ: Thank you. The following
15
announcement addresses the issue of conflict of
16
interest with respect to Section 17, Best
17
Pharmaceuticals for Children Act Adverse Event
18
Reporting, and is made a part of the record to
19
preclude even the appearance of such at this
20
meeting.
21
This morning you will hear from Dr.
22
Solomon Iyasu, lead medical officer with the
23
Division of Pediatric Development.
As mandated in
24 the
Best Pharmaceuticals for Children Act, Dr.
25
Iyasu will report on adverse events for the
9
1
following drugs that were granted market
2
exclusivity under 505(a) under the Federal Food,
3
Drug and Cosmetic Act, Paxil, paroxetine; Celexa,
4
citalopram; Pravachol, pravastatin and Navebjne,
5
vinorelbine.
6
Because the agency is not seeking advice
7 or
recommendations from the subcommittee with
8
respect to these products there is no potential for
9 an
actual or apparent conflict of interest.
10
The following announcement addresses the
11
issue of conflict of interest with respect to the
12 use
of imaging drugs in conjunction with cardiac
13
imaging procedures in the pediatric population and
14 is
made a part of the record to preclude even the
15
appearance of such at this meeting.
Based on the
16
agenda, it has been determined that the topics of
17
today's meeting are issues of broad applicability.
18
Unlike issues before a committee in which a
19
particular firm's product is discussed, issues of
20
broader applicability involve many sponsors and
21
their products. All subcommittee
participants have
22
been screened for their financial interests as they
23 may
apply to products and companies that could be
24
affected by the subcommittee's discussions of
25
imaging drugs used in conjunction with cardiac
10
1
imaging procedures in pediatric populations.
2
To determine if any conflicts of interest
3
existed, the agency has reviewed the agenda and all
4
relevant financial interests reported by the
5
meeting participants. Based on
this review, it has
6
been determined that there is no potential for an
7
actual or apparent conflict of interest at this
8
meeting.
9
With respect to FDA's invited industry
10
representative, we would like to disclose that Dr.
11
Samuel Maldonado is participating in this meeting
12 as
an industry representative acting on behalf of
13
regulated industry. Dr. Maldonado
is employed by
14
Johnson & Johnson.
15
In the event that the discussions involve
16 any
other products or firms not already on the
17
agenda for which FDA participants have a financial
18
interest, the participant's involvement and
19
exclusion will be noted for the record.
20
With respect to all other participants, we
21 ask
in the interest of fairness that they address
22 any
current or previous financial involvement with
23 any
firm whose product they may wish to comment
24
upon.
25
Ted Treves is Chief of the Division of
11
1
Nuclear Medicine at Children's Hospital, Harvard,
2 who
was an invited speaker for today, will not be
3
able to attend.
4
DR. CHESNEY: Thank you. Our first
5
speaker this morning will be Dr. Rosemary Roberts,
6 who
is going to offer a welcome on behalf of the
7
Office of Counteterrorism and Pediatric Drug
8
Development.
9 Welcome
10
DR. ROBERTS: Good morning. I would like
11 to
take this opportunity to thank you all for
12
coming today. I would also like
to thank the
13
"Murphys" for allowing me to come up and speak. I
14
rarely get to do it; you know, I am sort of the guy
15 in
the middle. I know some of you had to
16
experience much worse weather than we have here
17
today in order to get here so we certainly
18
appreciate all of your dedication in coming.
19
Our office, as you know, has two high
20
priority areas, counterterrorism which we might be
21
dealing with today unfortunately, and also
22
pediatric drug development, and we are certainly
23
happy that we have this program today.
24
We are excited about learning more about
25
cardiac imaging and having this opportunity to
12
1
discuss it and have such a distinguished group of
2
people here to help us see how to move forward in
3
this area. So, thank you very
much for coming. I
4
hope that you have a good day and we appreciate all
5 the
advice that you can give us.
6
One other thing, as you know because Diane
7
Murphy mentioned it yesterday, with the recent
8
legislation, the Pediatric Research Equity Act, we
9 now
have a full pediatric advisory committee.
We
10 are
working on that charter and hope to have
11
something going on with that in the next couple of
12
months and then we will be setting up that advisory
13
committee. Thank you.
14
DR. CHESNEY: Thank you, Dr.
Roberts. Our
15
next speaker is Dr. Solomon Iyasu who is going to
16
bring us up to date on the adverse event reports as
17
required by the BPCA.
18
Adverse Event Reports per Section
17 of BPCA
19
DR. IYASU: Good morning. Yesterday I
20
presented adverse event reports for paroxetine and
21
citalopram pertaining to psychiatric adverse
22
events. Today I will be
presenting on adverse
23
events reported for paroxetine and citalopram and
24
then, subsequently, I will report on adverse events
25 for
vinorelbine and pravastatin.
13
1
[Slide]
2
First I would like to acknowledge the
3
contributions of these individuals.
4
[Slide]
5
First I will speak about paroxetine and
6
citalopram and then vinorelbine and pravastatin.
7
[Slide]
8
The data source for the adverse events is
9
from the FDA's Adverse Event Reporting System which
10 is
a spontaneous and voluntary system. This
system
11 has
several limitations which I wanted to bring to
12
your attention. The
under-reporting is a very
13
significant problem. There are
reporting biases
14
that may be associated with either media publicity
15 or
depending on how long the drug has been on the
16
market. The quality of the
reports is variable,
17
often very scanty. And, this
database only
18
includes the numerator data, therefore, it is very
19
difficult to estimate the true incidence rate of
20
events or exposure risk.
21
[Slide]
22
Since I will be talking about the use of
23
these medications in the pediatric population, I
24
would like to also tell you a little bit about this
25
database that FDA has. The first
is IMS Health,
14
1
National Prescription Audit Plus which measures
2
prescriptions dispensed from retail pharmacies, but
3 the
disadvantage is that it does not provide
4
demographic information or prescription use. So,
5 it only
gives you total prescriptions dispensed.
6
The other database is the National Disease
7 and
Therapeutic Index, which is a survey based on a
8
sample size of about 2,000 to 3,000 office-based
9
physicians. The small sample size
can make these
10
data projections unstable, particularly when use is
11 not
very prevalent as in the case of the pediatric
12
population.
13
[Slide]
14
Another database available to FDA is based
15 on
a large prescription claims database but, again,
16
these data cannot be projected nationally. There
17 is
no methodology developed for that.
18
Premier is another database which contains
19
inpatient drug use from about 400 acute,
20
short-stay, non-federal hospitals.
There is
21
national projection methodology available for this
22
data, but accurate national estimates are
23
selectively available. Drug use
cannot be linked
24 to
diagnosis or procedures, and the treatments
25
administered at hospital outpatient clinics are not
15
1
included in this database.
2
[Slide]
3
There is one more inpatient database,
4
which is the Child Health Corporation of American
5
Pediatric Health Information System which captures
6
information from about 26 free-standing children's
7
hospitals with charge level drug utilization data.
8
Again, although this is very pediatric specific,
9 the
data are from a limited number of hospitals
10
and, therefore, cannot be projected nationally.
11
[Slide]
12
Now coming to the drugs that I will be
13
talking about today, there is some background about
14
Paxil which I mentioned in yesterday's
15
presentation. It is an
antidepressant which is
16
marketed by GlaxoSmithKline, first approved in
17
December, 1992. Its adult
indications are several
18
psychiatric conditions--major depressive disorder,
19
obsessive-compulsive disorder, panic disorder,
20
social anxiety disorder and generalized anxiety
21
disorder, post-traumatic stress disorder. There
22 are
no approved pediatric indications. Exclusivity
23 for
this drug was granted on June 27, 2002.
24
[Slide]
25
The relevant safety information on the
16
1
label as it currently exists refers to pregnancy
2
category C, which means that the drug has not been
3
studied in pregnant women and, therefore, when
4
using it in pregnant women the risks and the
5
benefits have to be weighed.
6
I talked about precautions specifically
7
pertaining to psychiatric events yesterday. Today
8 I
have listed them here but what is specifically
9
important here are the seizures and the adverse
10
reactions with abrupt discontinuation of this
11
medication, and in patients with a history of
12
seizures caution should be exercised with the use
13 of
this medication.
14
[Slide]
15
Additionally, there is information in the
16
adverse event section of the label pertaining to
17
pre-marketing reports and that includes
18
hypertension, diabetes, dysphagia and nausea and
19
vomiting.
20
In post-marketing reports there are
21
reports of serotonin syndrome, hepatic dysfunction
22 and
anaphylaxis, and also in the overdose section
23 of
the label about dangerous hepatic dysfunction.
24
[Slide]
25
Coming to the use data for this
17
1
medication, it is the second most commonly used
2
SSRI in children. For some of you
who were here
3
yesterday at the other meeting this is a repetition
4
but, for the benefit of the others who were not at
5
that meeting I am repeating this information. Both
6
pediatric and adult prescriptions have increased
7
steadily in recent years.
Pediatric diagnoses most
8
often linked with use of this medication include
9
depression, anxiety and obsessive-compulsive
10
disorders. And, pediatric
patients account for
11
approximately 3.5 percent of total U.S.
12
prescriptions of Paxil between July, 2002 and June,
13
2003.
14
[Slide]
15
When we looked at the one-year
16
post-exclusivity determination period, there was a
17
total of 127 pediatric adverse event reports.
18
After my review and excluding all the duplicates,
19
these are the unique reports for pediatrics in one
20
year. We categorized them into
different
21
categories and psychiatric adverse events accounted
22 for
about 68. The rest of them are
discontinuation
23
syndrome, about 7 patients.
Maternal exposure was
24
about 33; neurologic about 8; accidental ingestion
25 in
2 and then others were 9. So, today we
will be
18
1
talking mostly about the non-psychiatric which
2
includes the 5 categories that I have here which
3 are
on this slide.
4
[Slide]
5 First I will talk about the adverse
events
6
pertaining to pediatric deaths.
There were about
7 10
deaths involving direct pediatric exposures; 9
8
completed suicides, which I discussed yesterday;
9 and
1 case of Stevens-Johnson syndrome. That
10
patient was also receiving valproic acid, with a
11
known association with Stevens-Johnson syndrome.
12
[Slide]
13
There were 3 deaths among patients with
14
pediatric exposure. The pediatric
exposures
15
included congenital heart disease and 36 premature
16
infants who died after 75 days postnatally. The
17
second case was a 53-day old infant who was also
18
getting OxyContin and immediate-release oxycodone
19 and
Paxil exposure prenatally--not the kid.
20
Autopsy was done and it was determined to be a SIDS
21
death by the medical examiner.
The third case was
22 a
multiple congenital anomaly, possibly a genetic
23
syndrome. This was an aborted
fetus and it was a
24
fetal death.
25
[Slide]
19
1
Going into detail about the 33 in utero
2
exposures or breast feeding exposures, there was a
3
possible withdrawal syndrome reported in 11
4
patients, one of the fatalities previously
5
described; and congenital anomalies in 5 patients
6 and
seizures in about 4 patients; developmental
7
delay or abnormality in 4 and murmur or congenital
8 heart disease in about 3; and insufficient
weight
9
gain in 2 patients; and there were others that
10
included various events that could not be
11
classified.
12
[Slide]
13
Focusing on the direct exposures, there
14
were 8 patients with neurologic events.
Among
15
these, 3 patients had extrapyramidal or movement
16
disorders. Two of these involved
other medications
17 as
well that are listed here, which are known drugs
18
associated with this kind of syndrome.
Seizures
19
were reported in 3 patients. Two
of these patients
20 had
existing seizure disorders and were also
21
receiving Paxil.
22
There was one patient where there was a
23
loss of consciousness and hallucinations. The
24
patient was also on amphetamine-dextro-amphetamine
25 at
the same time. Then, there was one
patient
20
1
where serotonin syndrome was reported as an adverse
2
event.
3
[Slide]
4
Continuing with the pediatric adverse
5
events, there were also reports of accidental
6
ingestion. One was a 2-year old
who ingested 6
7
tablets of paroxetine and recovered without
8
sequelae. A 2-year old was a
comatose patient with
9
ingestion of multiple medications including
10
paroxetine who recovered after an ICU course.
11
There were a number of medications that were
12
involved as concomitant medications, including
13
other psychotropic agents, theophylline,
14
amytriptyline--there were several of them so this
15 was
a very complicated polypharmacy case.
Other
16
events--there were 9 single occurrences and the
17
majority were labeled.
18
[Slide]
19
In closing, most of the events were
20
labeled or related to labeled events.
Unlabeled
21
events involved maternal exposures.
And, the
22
safety of paroxetine will continue to be monitored
23 in
the future. We could not determine
causality of
24 any
of these medications because of the multiple
25
medications and also the scant histories in some of
21
1 the
case reports. Nevertheless, we will
continue
2 to
monitor adverse events for paroxetine in the
3
Adverse Events Reporting System.
4
[Slide]
5
Now I will talk a little bit about Celexa,
6
citalopram which is also an antidepressant,
7
marketed by Forest Pharmaceuticals.
Its only adult
8
indication is for major depressive disorder and the
9
typical adult dose is about 20-40 mg/day. Again,
10
there are no approved pediatric indications. This
11 was
first marketed in July, 1998 and pediatric
12
exclusivity was granted in July, 2002.
13
[Slide]
14
Again just mentioning some of the relevant
15
safety labeling associated with this drug, it is
16
again a pregnancy category C drug.
It is also
17
excreted in breast milk so caution should be
18
exercised when used in nursing mothers.
19
In the precautions section there are
20
precautions regarding impairment of intellectual or
21
psychomotor functions with the use of citalopram.
22
Also, there is danger of seizures, especially in
23
ones who have history of seizure, and citalopram
24
should be used with care. In the
post-marketing
25
reports and overdose section of the label, there
22
1 are
adverse events pertaining to QTc prolongation.
2
[Slide]
3
Summarizing some of the use data for
4
citalopram, it is the fourth most commonly used
5
SSRI in children. Both pediatric
and adult
6
prescriptions have, again, increased steadily in
7
recent years. Pediatric patients
account for
8
approximately 3.3 percent of the total U.S.
9
prescriptions of Celexa.
Pediatric diagnosis is
10
often linked with its use in depressive disorders,
11
obsessive-compulsive disorder and attention deficit
12
disorder.
13
[Slide]
14
For the one-year period of review, which
15 includes
the post-exclusivity period, there were 42
16
unduplicated pediatric reports after this review
17 was
undertaken, and 16 out of the 42 were in utero
18
exposures and mostly resulted in unlabeled events
19 and
one death that I will discuss later; 26
20
children involved direct exposure and 8 resulted in
21
unlabeled events and no deaths.
As I mentioned
22
yesterday, there were 16 serious adverse events, 10
23
hospitalizations and about 4 life-threatening and 2
24 with
disability.
25
[Slide]
23
1
Going to the gender and age distribution
2 of
these adverse events, they were both in females
3 in
both direct and in utero exposure. As
expected,
4 the
in utero exposures were reported in 4 patients
5 who
were less than 2 years. The majority of
them
6
were actually less than 1. In the
direct exposure
7
they were mostly in the older patients, 9 from 6-11
8
years and 15 patients in 12-16.
9
[Slide]
10
Looking at the reasons for exposure to
11
citalopram in these reports, as I mentioned, 16 of
12
them were in utero and included 13 patients who
13
were receiving citalopram for the treatment of
14
depression. Two involved
ingestion of another
15
person's prescription and then other events which
16 are
post-traumatic syndrome and GAD and RDD and
17
also anxiety, aggression and one was ADHD, just one
18
single occurrence of those conditions.
Then, in 6
19
patients it was unknown why they were receiving
20
citalopram.
21
[Slide]
22
Focusing on the known adverse events, of
23 the
16, as I mentioned, there was one death.
There
24 was
an autopsy done and there was no cause of death
25
identified by the medical officer.
It was signed
24
1 out
as a SIDS death in a 4-month old. There
were
2
congenital anomalies in 7 patients.
Three were
3
unrelated kidney malformations; 1 eye malformation;
4 1
cardiac defect; 1 cleft lip and 1 congenital
5
megacolon. Then, there were 5
patients where
6
potentially there was a neonatal withdrawal
7
syndrome, and then there were 3 other patients with
8
myoclonus and otitis in 1 patient and delayed head
9
control at 1-month in 1 patient.
In the last
10
patient there was a report of fetal asphyxia.
11 [Slide]
12
Among the direct exposure group there were
13 21
patients, excluding the 5 psychiatric events
14
that I reported on yesterday.
There were 4
15
patients in which cardiovascular events were
16
reported. One was a
supraventricular tachycardia
17 in
an 8-year old with a prior history of similar
18
episodes. It resolved after
Celexa was
19
discontinued. There were 2
patients with prolonged
20
QTc. One involved syncope and
seizure in a 13-year
21 old
who was also taking other medications
22
concomitantly, albuterol, cetirizine and
23
montelukast. There was also a
patient where an
24
overdose of citalopram was involved in a 14-year
25
old. Whether this was an
intentional overdose or
25
1
accidental was not reported so we cannot give you
2
additional details on that. There
was 1 patient
3
where arrhythmia was reported in an 8-year old with
4
overdose of citalopram.
5
[Slide]
6
In the group where there were reports of
7
neurological or special senses adverse events,
8
there were 8 patients. One
involved demyelinating
9
spinal lesion in a 13-year old who was also on
10
methylphenidate and multivitamins.
There was a
11
patient with a visual field cut in a 15-year old
12 who
was also on Depo Provera and who improved after
13
discontinuation of Depo. There
was one patient
14 with
a cataract, a 10-year old, also on
15
risperidone, and 5 patients with seizures.
16
[Slide]
17
Among other events that were reported
18
there were 2 patients where serotonin syndrome was
19
predominantly given but also, as part of the
20
syndrome, seizures occurred in both of these cases.
21
Then, there was 1 where only syncope was reported
22
with the use of Celexa.
23
There was one curious report of a
24
false-positive drug screen for cocaine on crushed
25
tablet. We tried to get
additional information on
26
1
this and from the chemistry point of view there is
2 no
relationship between these two structurally or
3
chemically. It may have been a
problem of
4
adulteration of the patient's medicine.
We do not
5
have any details but this involved a police test
6
that tested a crushed tablet found on a person
7
found to be positive for cocaine.
There were
8
others. Five patients involved
concomitant
9
medications and/or complicated underlying disease
10
which could not be categorized into a specific
11
category.
12
[Slide]
13
In summary, unlabeled events included in
14 the
non-psychiatric adverse events are the ones
15
that I mentioned involving in utero exposure and
16 the
case where demyelinating spinal cord lesion was
17
reported for one patient; visual field cut in one
18
patient and the supraventricular tachycardia in
19
another patient. These are single
occurrences.
20
Supraventricular tachycardia is not specifically
21
labeled but tachycardia and sinus tachycardia are
22 in
the label.
23
[Slide]
24
In conclusion, we will continue to monitor
25
these adverse events but I wanted to bring to your
27
1
attention that there will be updates that will be
2
provided in the future meetings regarding three
3
issues that are under review, neonatal withdrawal,
4
ophthalmologic malformation and then the QTc
5
prolongations. We will be
reporting on this in
6
future meetings.
7
So, I am done with
paroxetine and
8
citalopram and if there are questions about this
9
section I will entertain any questions.
There are
10
more details that are needed but Dr. Hari Sachs
11
will work very closely with me on these issues and
12 we
will have some details about the cases if there
13 are
any questions. Yes?
14
DR. CHESNEY: Yes, Dr. Nelson?
15
DR. NELSON: Remind me, given our
16
discussion yesterday, can you tell from the data
17 or,
if you can't is it worth finding out what the
18
timing of the suicide events on paroxetine is in
19
respect to when the drug was started?
In other
20
words, within a week, the first two weeks of
21
exposure to the drug?
22 DR. IYASU: It varied.
It varied from
23
patient to patient. There was no
clear pattern.
24
Most of them were on therapy at the time that the
25
suicide events occurred. It
varied from about 14
28
1
days to about a year in terms of how long they had
2
been on therapy. The events that
were reported
3
varied also. But there was not
much detail so that
4 we
can make a clear, distinct pattern as to when.
5
Some of them were early; some of them were later.
6 It
was very difficult, as I mentioned yesterday, to
7 try
to pin it down because of the scanty
8
descriptions that were provided in the case reports
9 but
most of them were on therapy. There were
a few
10
that were post-therapy and during the withdrawal
11
period.
12
DR. CHESNEY: Dr. Ebert?
13
DR. EBERT: Of the 33 maternal
exposures
14 you
noted with paroxetine, do you know what
15
proportion of those were in utero versus breast
16
feeding?
17
DR. IYASU: Out of the 33, about 6
of them
18
involved also breast feeding exposure.
19
DR. EBERT: I noticed there was no
caution
20
regarding breast feeding, or you didn't mention one
21
specifically with that product in the labeling.
22
DR. IYASU: Yes, I think I may not
have
23
mentioned it but there is also in the label
24
information about nursing mothers.
25
DR. CHESNEY: Dr. Glode?
29
1
DR. GLODE: I just want to
clarify, as
2
part of the pediatric exclusivity there is no
3
requirement for the sponsor to do any sort of
4
random sample or active surveillance for safety
5
issues or adverse events? They
just also use this
6
passive reporting system? Is that
right?
7
DR. IYASU: Well, as part of the
BPCA, it
8 is
my understanding that the manufacturers are
9
required, just by FDA regulations, to report all
10
adverse events that come to them to the FDA. But
11
this is for the passive surveillance system.
12
Unless there are specific sorts of adverse events
13 that
are agreed upon in the pediatric studies for
14
follow-up, they do not have to report on follow-up.
15
Diane can add to this.
16
DR. D. MURPHY: The only thing I
wanted to
17 add
is that we have asked for specific
18
post-studies, you know, completion of study
19
surveillance for certain products.
But it has to
20 be
asked for in the written request.
Outside of
21
exclusivity there are Phase IV commitments that
22
could be asked for. But, in
general, what you
23
heard is what usually happens--studies are
24
completed and unless there is a specific
25
requirement they revert to the passive reporting
30
1
system unless a company notices a signal that they
2
then bring to the attention of FDA.
3
DR. S. MURPHY: Joan, I just
wanted to add
4 for
our guests that are here from imaging that this
5 is
mandatory one-year reporting required under the
6 Best
Pharmaceuticals for Children's Act in which a
7
drug gets pediatric exclusivity, which you will
8
learn about in a little while from Susan's talk.
9
Then we are required by law to report to this
10
committee publicly the adverse events that occur
11
forward for one year. So, that is
why you are
12
seeing reporting on these drugs.
They have
13
triggered a time point for the committee to hear
14
about the reports.
15
DR. CHESNEY: Could I ask a
question,
16
please? Could you clarify
this--Dr. O'Fallon
17
mentioned in the van this morning reading about
18
this neonatal withdrawal syndrome and it didn't
19
come up yesterday. I notice with
paroxetine you
20
commented that these are unlabeled events involving
21
maternal exposure. What exactly
is the withdrawal
22
syndrome, and is this something that should be in
23 the
label? Could you elaborate a little?
24
DR. IYASU: These are issues that
are
25
under review right now, but to give you sort of
31
1
additional information on what the concern is I
2
have some notes here. It is
usually associated
3
with reports that involve nervous or neuromuscular
4
effects after birth when the mother is exposed to
5
some of these SSRIs, including citalopram or Paxil.
6
This may include symptoms like irritable or
7
agitated crying, hyperreflexia, hypertonia,
8
seizures or seizure-like movements, and also
9
include some breathing difficulties as well as
10
feeding difficulties. So, this is
sort of a
11
syndrome that is increasingly being recognized with
12
babies who have been exposed prenatally to some of
13
these drugs. It is still under
continued review
14
right now to see whether this is information that
15
needs either to be communicated to the public or be
16 put
in the label. I can't give you more
details
17
except that we are looking at it very closely.
18
DR. CHESNEY: Presumably, these
were
19
serious enough to cause somebody to make a report
20
which is impressive to me. This
is quite an
21
impressive number for just voluntary reporting. Do
22 you
have any more information about whether they
23
needed to be managed? I assume if
they had
24
seizures they had to have some specific management
25
issues.
32
1 DR. IYASU: I don't have additional
2
information right now about what specific measures
3
will be taken regarding this, except to say I think
4
this is something that we are concerned about and
5
specific recommendations as to what would happen as
6
follow-up are still open.
7
DR. CHESNEY: Maybe I can ask some
of the
8 FDA
folk, is there anything that we can do to help
9
move this along? This seems like
it might be a
10
significant issue.
11 DR. S. MURPHY: I think just what you have
12
done is expressing your concern and we will take
13
that back to the Division. I
think that it is
14
under review right now and I think that is why
15
Solomon can't say more.
16 DR. IYASU: Yes.
17
DR. CHESNEY: Dr. Gorman?
18
DR. GORMAN: Are you aware of the
Canadian
19
literature surrounding this withdrawal syndrome
20
from the unit in Toronto that looks at
21
maternal-fetal exposure rate and has noted an
22
increased transfer to NICUs for babies born with
23
these agents?
24
DR. IYASU: Yes, I am and it is
good that
25 you
are pointing that out, and the Division is also
33
1
aware of the data.
2
DR. CHESNEY: I have one other
question
3
relative, I guess, to yesterday's discussion, the
4
paroxetine 68 psychiatric adverse events in
5
children, were those along the lines of what we
6
were talking about yesterday, which is activation
7 of
stimulant syndrome, or do you have any further
8
breakdown of those?
9
DR. IYASU: Actually, we were
talking
10
about this with Hari. Hari, do
you want to comment
11 on
that?
12
DR. SACHS: You know, as Solomon
pointed
13 out
yesterday, there are the 9 completed suicides
14 and
17 suicide attempts. I went back and
just
15
checked the case reports to see how many of them
16
were associated with agitation. I
picked up 8, 2
17 of
which have resulted in completed suicide, 2 with
18
suicidal ideation, 2 with suicide attempts and 2
19
with self-mutilation.
Interestingly enough, for 4
20 of
them the kids' reasons for treatment were not
21
major depression; they were OCD and anxiety; 4 of
22
them were for depression and it was pretty split,
23
half female, half male, and half of them were on
24
concomitant medications, including other
25
psychotropics or having a history of substance
34
1
abuse. So, it is definitely a
very mixed bag.
2
DR. CHESNEY: If we subtract out
the
3
suicidal issues, that still leaves a significant
4
number of other children. What
were their adverse
5
events?
6
DR. S. MURPHY: The other
psychiatric
7
adverse events, as I said, the totals were the 9
8
completed suicides, 17 suicide attempts, several
9
cases of suicidal ideation and 10 of self-injury.
10
Then, the rest of them were kind of emergence of
11
other psychiatric symptoms such as mania. So, it
12
depends I guess on what you look at but what I was
13 thinking
was that the agitation was picked up, or
14 at
least the other suicidality issue was picked up
15 as
well as the agitation. It wasn't that
agitation
16
looked, you know, linked to anything else at least
17 in
these 68 reports.
18
DR. IYASU: Yes, I think just
looking at
19
these case reports there was tremendous variability
20
also. But you can find some
agitation in some of
21 the
case reports and no mention of it in others.
22 So,
it was hard to sort of see which one is
23
predominant there; there is a mixture.
24
DR. CHESNEY: Dr. Nelson?
25
DR. NELSON: I realize this
suggestion may
35
1 be
naive from a resource point of view but, given
2 the
discussion, does it make sense to do a more
3
in-depth case ascertainment both for the cases you
4
have got and to see if there are other cases, and
5 to
see if someone could do a case study design
6
approach to see if they could ascertain that
7
this--you know, similar to what happened with the
8
rotaviral vaccine--might be a hint relative to the
9
timing and to this issue of agitation?
I mean,
10
that might be one way to try to sort this out?
11
DR. IYASU: I think that is a good
12
suggestion. These kind of studies
always require
13
additional resources that the Office of Drug Safety
14 may
not have available, but theoretically I think
15 you
can go back and try to ascertain some of these
16
cases. But one thing that we have
to be careful
17
about is that the cases that come to our attention
18 are
a selected few and we don't know what they
19
actually represent because, you know, it is really
20 a
small percentage of an unknown group of adverse
21
events. So, it requires I think
careful assessment
22 of
what the cases actually represent. Do
they
23
represent other cases that are occurring in the
24 population?
But it is a good suggestion.
25
DR. CHESNEY: Dr. Glode?
36
1
DR. GLODE: I would just like to
2
emphasize, and I think this came up for many people
3
yesterday, that with a database of between 3,000
4 and
4,000 children with regard to safety issues, it
5 is
a very inadequate number for safety. So,
there
6
needs to be some mechanism I think, other than this
7
passive surveillance reporting, for doing
8
additional safety studies whether that is by Phase
9 IV
studies from the sponsor, or whatever, but there
10
needs to be more safety data beyond 3,000 to 4,000
11 I
think for children for these drugs.
12
DR. IYASU: I think your point is
well
13
taken.
14
DR. CHESNEY: Thank you.
15
DR. IYASU: All right, thank you.
16
[Slide]
17
Now I will report on two other medications
18 that
have received exclusivity. The first
drug is
19
vinorelbine which is an anti-tumor drug marketed by
20
GlaxoSmithKline. The indications
which are
21
approved are in adults as a single agent or in
22
combination with cisplatin for the first-line
23
treatment of ambulatory patients with unresectable,
24
advanced non-small cell lung cancer.
Again, there
25 are
no approved pediatric indications for this
37
1 medication.
Exclusivity was granted on August 15,
2
2002.
3
[Slide]
4
Summarizing the use data, there wasn't
5
much in terms of our databases that revealed a lot
6 of
use for this medication in the pediatric
7
population.
8
In CHCA, which is a children's hospital
9
corporation database which is 26 children's
10
hospitals that I mentioned before, which is a
11
discharge-based database, there were 5 discharges
12 in
2001 and about 21 discharges in 2002 that
13
indicated that this medication may have been used.
14 The
diagnoses that were closely linked with its use
15
were put under the category of chemotherapy and
16
most of them were Hodgkin's disease.
17 [Slide]
18
Looking at the adverse event reports for
19
vinorelbine, the total raw number of adult and
20
pediatric reports that were received were about
21
495, and 181 of them were domestic and 314 were
22
international reports. These are
not adjusted for
23
duplicates so this includes duplicates also.
24
Looking at the pediatric reports for the
25 one
year, there were 3 unduplicated pediatric
38
1
reports and 1 was U.S. and 2 were foreign. All
2
were reported as having serious outcomes but there
3
were no deaths with the use of this medication in
4 the
one-year period that was evaluated. Five
of
5 the
16 adverse events that were reported were
6
considered unlabeled. The
diagnosis or the reason
7 its
use was for the treatment of rhabdomyosarcoma
8 in
2 of the patients and 1 of the patients had
9
neuroblastoma and the drug was being given for that
10
treatment.
11
[Slide]
12
I am just summarizing the 3 patients who
13
were reported to us with adverse events.
The first
14 one
is a 14-year old with rhabdomyosarcoma who
15
developed neutropenia, a labeled event, and was
16
successfully treated with Nupogen.
17
The second patient was a 2-year old with
18
rhabdomyosarcoma who developed life-threatening
19
adverse events including unlabeled events that
20
included epidermolysis, muscle inflammation,
21
somnolence and tachypnea. This
patient was also on
22
cytoxan. The patient was
hospitalized for about 16
23
days and eventually recovered and was discharged.
24
A 6-year old was diagnosed neuroblastoma
25 and
developed adverse events including one of the
39
1
unlabeled events, the muscle spasm, but the adverse
2
events that reported for this patient resolved
3 after
lowering the dose of vinorelbine.
4
[Slide]
5
So, it was a small number of reports that
6 we
got for the labeled and unlabeled adverse events
7
were reported, as I mentioned before.
The
8
unlabeled events have also been reported in adults
9 and
are not unique to pediatrics. The FDA
will
10
continue its routine monitoring of additional data
11 on
adverse events in all populations, including
12
pediatrics, to follow-up on the significance of any
13 of
these events.
14
[Slide]
15
The last drug I will be presenting on is
16
pravastatin, which is one of the statins. It is
17
marketed by Bristol-Myers Squibb.
In adults it is
18
indicated for the prevention of coronary and
19
cardiovascular events and hyperlipidemia. In
20
children it is approved for 8 years and older for
21 the
treatment of heterozygous familial
22
hypercholesterolemia. Pediatric
exclusivity was
23
granted on July 10, 2002.
24
[Slide]
25
Drug use databases indicate that the total
40
1
dispensed prescriptions have increased by about
2
17.5 percent between September, 1999 and August,
3
2003. That is, from 13.4 to 15.8
million per year
4 for
pravastatin and that is adults and pediatrics.
5
This is total dispensed prescriptions.
6
Pediatricians wrote about 47,000 or about 0.4
7
percent of the total of the 15.8 million
8
pravastatin prescriptions during that period.
9
[Slide]
10
Looking at the proportion of pediatric
11
prescriptions, an estimated 7,900 prescriptions
12
were dispensed nationwide to pediatric patients
13
aged 1-16 years. This is based on
a calculation of
14 the
proportions that were obtained from advanced
15
PCS, which is a database that I mentioned before
16
which has demographic information, and applying it
17 to
the total dispensed prescriptions. It is
a
18
small number but this has to be interpreted with
19
caution because really this is an estimate.
20
[Slide]
21
There was a total number of adult reports,
22
about 993 reports during the exclusivity period and
23 691
were U.S. and 302 were international reports.
24
There were no pediatric adverse event reports that
25
were mentioned in the one-year exclusivity period.
41
1
[Slide]
2
Therefore, I don't have any additional
3
comments on pravastatin in the pediatric
4
population, except to say that we will continue to
5
monitor the database and see if there are any
6
adverse events that emerge. Thank
you very much.
7
DR. CHESNEY: Thank you. Are there any
8
questions? Yes, Dr. D'Agostino?
9
DR. D'AGOSTINO: Could you tell me
or us
10
what the physicians do with the statins in terms of
11
muscle, liver and so forth in the pediatric
12
population? Do they do anything
routinely in terms
13 of
the side effects? I mean, what do you do
with a
14
child with muscle problems? The
children are
15 growing
and so forth so how do you recognize that
16
that is happening?
17
DR. IYASU: Well, from the adverse
event
18
reports there is no way to tell, or there is no
19
information as to what actually is being done to
20
treat that, except in the cases that were presented
21
today where they were admitted but what actual
22
treatment was given was not clearly specified.
23
DR. D'AGOSTINO: Do we know if
there is
24
withdrawal of the drug in the children where things
25
like that might be happening?
That is not an
42
1
adverse event necessarily but if the children are
2
complaining about muscle pains and so forth.
3
DR. IYASU: I can't tell you because the
4
narratives that were provided to us were very
5
scanty. So, what treatment was
given to these
6
individual patients is not clearly stated in those
7
narrative reports, except that there was an ICU
8
course for one of them where it was considered to
9 be
serious enough that the patient was admitted.
10 In
terms of the complaints, they were elicited and
11
reported by a health professional.
Whether these
12
were based on clinical records or medical records
13 or
whether they were just clinical encounters, I
14
couldn't tell from the narrative.
15
DR. CHESNEY: Dr. Santana?
16
DR. SANTANA: Can you clarify for
me a
17
process issue? My understanding
is that when an
18
agent is granted exclusivity there is a commitment
19 to
do a number of studies and those studies may
20
occur in different time lines.
When does that data
21
from those studies surface in adverse event
22
reporting to this committee?
Because it seems to
23 me
that what we are seeing are reports that are
24
coming from different sources, more public kind of
25
usage sources, but the data from the actual studies
43
1
that are being done or have been done under the
2
exclusivity--when does that surface for us to see
3 in
these reports?
4
What made me think about that question is
5
that for a lot of the oncology drugs that may be
6
granted exclusivity, and I think this one is a good
7
example, those studies will occur in a semi-closed
8
system either through the cooperative group
9
mechanism or through large oncology institutions,
10 and
those data may not necessarily show up in these
11
other databases. For the oncology
drugs, why don't
12 you
go to the NCI and request their adverse event
13
reporting for the pediatric patients that are
14
participating in those studies under drugs that
15
have been granted exclusivity?
That would be a
16
more enriched data set than using this other
17
system. Can you comment, please?
18
DR. IYASU: My comment is that the
adverse
19 events
are reported to FDA, again, through this
20
passive system. The exclusivity
is granted on a
21
specific data and then, if there is a change in
22
labeling for example, it may not happen for several
23
months after exclusivity is granted.
So, in
24
theory, what you would expect is that there would
25
have been a change in the label and then there
44
1
would be increased usage of the medication and then
2 we
have to monitor or would pick up if there are
3 any
adverse events that emerge as use expands.
But
4
with many of these drugs maybe the indication is
5 not
approved and, secondly, there is a time lag
6
between the use and the period that we are looking
7 at
because this is immediately the one-year after.
8
Now, we depend on adverse event reporting
9
with the system that we have. We
don't have any
10
other system. But an active
surveillance mechanism
11 is
where we actually go to do case finding and
12
querying other databases is something that is a
13
good idea. But, again, as I said
before, that
14
system is not in place to go after that.
15
DR. SANTANA: So, the data that is
being
16
collected by the sponsors for the studies that may
17 be
related to exclusivity, when does that data
18
surface for us to see?
19
DR. IYASU: Oh, that is a question
that--
20
DR. S. MURPHY: Yes, the medical
officers'
21
reviews have to be posted on the web 180 days after
22
exclusivity is granted. I think
you bring up an
23
excellent point. I think what we
are trying to do
24 is
interpret the law and figure out the best way to
25 report to you, and that is one of the things I
was
45
1
going to ask you, if this is the best information.
2
What we are doing now is going to the AERS passive
3
system and picking up all the reports for a year
4
after exclusivity. We are not
going into the
5
trials and pulling those out.
6
DR. SANTANA: Yes, what
highlighted my
7
comment was the oncology example.
8
DR. S. MURPHY: That is a very
good
9
example.
10
DR. SANTANA: You would not pick
up a lot
11 of
the oncology adverse event reports through these
12
databases. You would have to go
to a very enriched
13
data set that already exists.
14
DR. IYASU: I agree.
15
DR. SANTANA: There is a lot of
16
under-reporting here.
17
DR. S. MURPHY: Yes, there is a
lot of
18
under-reporting.
19
DR. SANTANA: This drug is an
example but
20 I
suspect if we continue that practice with
21
oncology drugs we will see a lot of under-reporting
22
that will not come out until years later when the
23
drugs are being used in a different way.
24
DR. S. MURPHY: Well, I agree with
you. I
25
think that the reporting of a lot of this, you
46
1
know, can be enhanced and we have sort of taken a
2
year now to report this way. I
think we also
3 realize
that the label is going to get out there
4 for
six months at least. So, is there
really,
5
after exclusivity, a big peak in pediatric use, or
6
does the use come later, or was it used off-label
7
before?
8
DR. D. MURPHY: I think the
question is
9
really good but it gets to a different process and
10 I
think it is an important process for this
11
committee to think about because it has huge
12
ramifications. What the law
mandates we do is, as
13 has
been noted, to report on the adverse event
14
reporting after exclusivity. At
some period in
15
that exclusivity the product will be approved and
16
labeled.
17
The issue is that the BPCA has said that
18
this information will be posted.
The studies will
19 be
posted on the web and theoretically in the
20
medical review information on the oncology
21
product--I mean, the information that came out
22
during the studies should be up on the web at that
23
point.
24
Now, I think the other issue though that
25
people are pointing out, and that I think this
47
1
committee is now very familiar with is that if you
2 have a new label and that label is supposed
to
3
reflect the adverse events that were defined in
4
those studies, then that is the way of
5
communicating to the public what those adverse
6
events were that were found in that better process,
7
which is controlled studies, versus this passive
8
adverse event reporting. That
label sometimes is
9 not
available except up on the web site somewhere
10 for
different periods of time depending on how many
11
labels are out there already, etc.
So, it will
12
vary.
13
So, I think you are bringing forth a very
14
important question which is access to this
15
information, which we talked about yesterday quite
16 a
bit. Second is the issue--and I really
think the
17
committee needs to think about this for a long
18
time--are you asking us to review every study that
19 is
approved under exclusivity? There have
been
20
over a hundred determinations and over 60, 70
21
labels. That would be 60 meetings
literally to go
22
over each of the studies. So, I
think that is a
23
different question. I just want
to make sure that
24 we
define when the information will be available.
25
DR. CHESNEY: Dr. O'Fallon had her
hand up
48
1
next.
2
DR. O'FALLON: I have another
process
3
issue. I was curious because in
looking at
4
pravastatin, or whatever it is, there are two
5
different estimates of the size of the prescription
6 to
the pediatric population. On one slide
it says
7
pediatricians wrote 47,000 of the total
8
prescriptions during that year and the other one
9
says an estimated 7,900 prescriptions were
10
dispensed. Now, I realize you are
working off two
11
different sets but the difference between 8,000 and
12
47,000 is big in my mind and I am wondering is that
13
sort of a very high upper bound and a very low
14
lower bound, or what. You are
trying to get at
15
what is the piece of the pie that goes for
16
prescriptions to this age group.
17
DR. IYASU: Yes, I think that is
an
18
important point. There is
obviously a big
19
discrepancy between the two estimates.
One is
20
referring to dispensed prescriptions written by
21
different specialties. The other
one is getting
22
proportions out of a database that is not
23
nationally representative and applying the
24
demographic percentage to the national database.
25 So,
we are trying to get sort of two estimates but
49
1
they are giving us different estimates and we don't
2
know how to sort of marry the two.
But we thought
3
that we would give these databases and explain what
4 the
limitations of both of these databases are,
5
which I mentioned before. So,
that is a good
6
point. It is something that we
have to work on to
7 try
to get better databases that could give us
8
better estimates and not miss significant portions
9 of
dispensed prescriptions. That is a good
point.
10
Thanks.
11
DR. CHESNEY: Dr. Gorman and then
Dr.
12
D'Agostino.
13
DR. GORMAN: I can explain four of
those
14
pravastatin prescriptions, I wrote them for my
15
mother.
16
[Laughter]
17
So, a pediatrician wrote them but it
18
didn't go to a pediatric patient. So, that is four
19 and
you only have 47,000 more to go. So.
20
The other issue that I think is a little
21 bit
more global is that I think I hear a different
22
theme emerging from our discussion which is that we
23
have listened to the AERS data reporting system and
24 its
weaknesses and we have listened to the concerns
25
that there are safety signals we will not meet
50
1
during the controlled clinical trials for efficacy.
2 I
think the AERS system grew up in a totally
3
different generation of information collection and
4
distribution and perhaps there needs to be a more
5
active system looking for safety signals than we
6 presently have. I think I heard Dr. Glode say that
7 and
I have heard other people say that with active
8
case finding there is a more active searching, and
9 I
am not sure that is inside the charge of the FDA
10 but
I am sure that that is something that would
11
enhance the safety of these agents.
Rather than
12
demanding of sponsors that the clinical trials get
13
larger and larger and larger, look for clinical
14
safety signals and perhaps there can be another
15
mechanism that allows us to look for safety signals
16 for
the rare events after post-marketing.
17
DR. CHESNEY: Dr. D'Agostino?
18
DR. D'AGOSTINO: My comment is
similar to
19
that. I mean, in some fields like
cardiology with
20 the
statins we have an idea, we have a very good
21
idea of what some of the problems are and there are
22
lots of different companies and lots of different
23
trials, but it is quite quick in some cases to put
24
together how many problems are developing. Instead
25 of
each study being reported separately, I know
51
1
with the OTCs and things that we do in some of the
2
cardiology we can quickly find out how many muscle
3
problems are developing, how many liver problems
4 are
developing without having a list of each study
5
being laid out but these companies are constantly
6
surveying. They know what some of
the problems are
7 and
they have active ways of getting at them.
Are
8 we
doing the same here? I mean, I presume
we are
9 and
the question is how do we get that information
10 to
the committee here and how you are actually
11
pulling that data together because, as we said, the
12
AERS is not really going to do it.
13
DR. D. MURPHY: The companies are
required
14 to
report this to us so it is coming into AERS.
If
15 the
company knows about it, it is coming in to us.
16
DR. D'AGOSTINO: What I was saying
is some
17 of
these are doing active registries, surveillances
18 and
so forth so they are actively looking.
They
19 are
not just waiting for a passive.
20
DR. D. MURPHY: I think what Dr.
Gorman
21 and
you all are trying to say is that you have
22
heard the limitations, and we have sort of pounded
23 you
with it multiple times, and that there needs to
24 be
a better way but that we can't power safety
25
studies for rare events. That
just won't go
52
1
forward; it is not feasible.
2
I was just trying to see if somebody from
3 our
ODS Office was here because it would be good
4 for
them to hear your concerns and we will relay
5
those back to them, how can we improve the process?
6 Can
we target--I think one of the questions is can
7 we
target areas, which it sounds like others have,
8
where we think there needs to be an active
9
surveillance system? Certainly,
as I mentioned
10
earlier, we have done that in a few cases where we
11
know what the safety signal is.
If you know what
12 the
safety signal is, then it is a lot easier to
13
design that kind of surveillance system.
So, you
14
know, it gets back to that kind of focused system
15
versus finding in kids unexpected results which I
16
don't know that we are able to do yet.
17
DR. CHESNEY: Dr. Danford?
18
DR. DANFORD: To briefly address
Dr.
19
D'Agostino's earlier question about what would the
20
response of a pediatric cardiologist be to muscle
21
pains, myalgias or muscle problems we might
22
encounter in starting these medicines in children,
23 I
think that we would be pretty quick to withdraw
24 the
medicines under those circumstances. I
don't
25
think, watching the people who handle our childhood
53
1
lipid problems in our town--I don't think that the
2
discovery of that or any of the other relatively
3
well-known complications discovered by our adult
4
colleagues would necessarily trigger a report that
5
would show up in AERS. You know,
we know about
6
these things; we stop the medicines and we don't
7
think about it. It highlights
once again the
8
inadequacies of this approach and our need to look
9 for
other ways.
10
DR. IYASU: I think these are all
very
11
good comments and, in terms of the limitations of
12 the
AERS database, I think everybody recognizes
13
that it has very limited utility in terms of
14
picking up adverse events. It is
useful to sort of
15
maybe generate some potential signals, especially
16
rare events that have not been picked up in
17
clinical trials, but to confirm the existence of an
18
event in association with a particular drug it is
19
terribly inadequate and I understand and I hear
20
what you are saying in terms of are there any
21
better ways of looking at adverse events and
22
monitoring them that would be a step forward. But
23
there are also limitations in terms of whether you
24 do
it for specific adverse events for a specific
25
drug or whether you do it for all the medications
54
1
that are regulated by FDA. As
Diane said, it has
2
been done for certain specific events of concern
3 but
when you try to do it to capture all potential
4
adverse events, that is a big undertaking and we
5
look forward to having some specific
6
recommendations from the committee.
Thank you very
7
much.
8
DR. CHESNEY: Thank you. Just thinking
9 out
loud, Dr. Danford raises a very interesting
10
point which is that if there were a difference in
11 the
incidence of a labeled adverse event in
12
children we would never pick that up because we
13
would just say, well, yes, we know that happens but
14 if
it were more common in children than adults we
15
wouldn't pick that up. Does that
make sense?
16
DR. IYASU: Well, we look at sort
of the
17
pediatrics and compare whether it is more common in
18
pediatrics for a specific event than in adults.
19 But
it is always very difficult also to sort of
20
have a relative rate of the event in the two
21
populations because of the different use patterns
22 and
different frequencies of use in the different
23
populations. So, a sort of
head-to-head comparison
24
sometimes doesn't work but it gives us some idea in
25
terms of whether there is a potential signal that
55
1 we
need to look further into.
2
DR. CHESNEY: Right, but a lot of
these
3
wouldn't be reported to AERS because, "well, this
4 is
something that we know happens" and unless it
5 may
be happening much more often in pediatrics it
6
wouldn't be reported because it is a labeled
7
adverse event.
8
DR. IYASU: Absolutely. Under-reporting
9 is
one of the big issues in AERS. Thank
you.
10 DR. CHESNEY: Thank you very much. I
11
think we have one new person at the table, Dr.
12
Stylianou, would you mind introducing yourself,
13
please?
14
DR. STYLIANOU: Mario Stylianou,
15
statistician from NIH. I do some
work with
16
pediatric clinical trials at the National Heart,
17
Lung and Blood Institute.
18
DR. CHESNEY: Thank you. There is nobody
19
scheduled to speak at the open public hearing but
20 let
me ask if there is anybody not scheduled who
21
would like to come to the microphone.
Apparently
22
not. We are scheduled for a
15-minute break.
23
Given the small room and small number of people and
24
potential to move ahead today, maybe we could take
25 10 minutes and, according to this clock, be
back
56
1
between 10:20 and 10:25 to begin our discussion of
2 the
cardiac imaging drugs. Thank you.
3
[Brief recess]
4
DR. CHESNEY: Let's get started if
5
everybody could find their seats, please. We do
6
have some new people at the table so I thought we
7
might take this opportunity to let them introduce
8
themselves and start over here.
9
DR. BEITZ: I am Julie Beitz. I am the
10
Deputy Director of the Office of Drug Evaluation
11
III.
12
DR. LOEWKE: I am Sally
Loewke. I am the
13
Acting Division Director of the Division of Medical
14 Imaging
and Radiopharmaceutical Drug Products.
15
DR. BUCKLEY: Hi, I am Shavhree
Buckley.
16 I
am a medical officer in the Division of Pediatric
17
Drug Development, and a pediatrician.
18
DR. CHESNEY: Thank you. Just one
19
technical or business detail, it was brought to my
20
attention that some people would be willing to
21
either forego lunch or make it a brief 15-minute
22
lunch in order to keep on going.
So, please keep
23
that in mind and we will raise it again at the end
24 of
this morning's session as to whether you want to
25 do
that.
57
1
The rest of our session very briefly, as I
2
understand it--and this will be repeated to us a
3
number of times but for the committee's benefit and
4 for
me thinking out loud, our challenge is to help
5 the
FDA determine what cardiac imaging drugs, not
6
devices or procedures but what cardiac imaging
7
drugs do we need pediatric labeling for.
Very few
8 of
these imaging agents or drugs currently have
9
pediatric labeling, and how many need it and for
10 how
many could the use simply be extrapolated from
11
adult labeling? Specifically,
they are interested
12 in
what imaging drug classes need further study.
13
Secondly, what patient populations would be
14
available to receive these drugs.
Along that line,
15
utilization information is particularly important.
16 In
other words, how many children would undergo a
17
procedure involving the agent such that there would
18 be
enough to do a study with the agent?
19
So with that, I am pleased to introduce
20 Dr.
Susan Cummins who is the lead medical officer
21 in
the Division of Pediatric Development. I
22
understand that in addition to introducing this
23
session, she may have some comments for us about
24 the
previous issue of adverse drug reporting.
25
Use of Imaging Drugs in
Conjunction with
58
1
Cardiac Imaging Procedures in the Pediatric
2
Population Pediatric Regulatory Update
3
DR. CUMMINS: Good morning. First, just
4 to
comment on the adverse drug reporting feedback
5
that you gave us, I wanted to let you know that we
6
kibitzed over the break and what we will do for our
7
next meeting and into the future is provide you
8
with the medical officers' summaries for the drugs
9
that are granted exclusivity. We
will also provide
10 you
with the labeling changes, as well as the AERS
11
summary that you get now in the summary that is
12
provided to you in your packets.
13
Diane Murphy has already shared your
14
concerns with the Office of Drug Safety who,
15
themselves, are always interested in strengthening
16
drug safety reporting to the FDA and we will be
17
talking with them about your concerns and see how
18 to
go forward with them.
19
[Slide]
20
I want to welcome you all here.
There are
21 a
lot of new faces at the table. I am
Susan
22
Cummins. I am a medical team
leader in the
23
Division of Pediatric Drug Development and Shirley
24
Murphy asked me to tell you a little bit about
25
myself so here is a 30-second story.
59
1
I came to the Division from the National
2
Academy of Sciences a little over a year ago where
3 I
was the Director of the Board on Children, Youth
4 and
Families. This board was a joint board
with
5
both the Institute of Medicine and the National
6
Research Council.
7
I also brought along a long experience
8
with environmental health, especially in childhood
9
lead poisoning. For many years I
managed the
10
childhood lead poisoning prevention program for the
11
State of California. In that role
we used meetings
12
such as this one, advisory committees, extensively.
13 We
were actually mandated by state law to use
14
advisory committees to help us with complex issues
15 of
science, medicine, public health and policy.
16 So,
I have a lot of experience with meeting
17
processes both at the National Academy of Sciences
18 and
in California, and I love meetings like this.
19 I
think your input is just so valuable and really
20 helps
us be able to move forward.
21
I want to thank you in advance for all
22
your time and wisdom, and at the end of the day for
23 the
advice that you are going to give us.
Many of
24
you, in addition to coming today, participated in a
25
series of scoping interviews that we conducted to
60
1
plan this meeting and to help us define the issues
2
that we needed to address. That
was just
3
unbelievably helpful. I don't
know that we could
4
have moved forward in planning this meeting without
5 the
input that you have given us already. We
also
6
look forward to a very stimulating and productive
7 day
so I want to thank you already for all that you
8
have done.
9
[Slide]
10
What I am going to do today is give you a
11
brief overview of the last decade of pediatric drug
12
development efforts at the FDA. I
am also pleased
13 to
report that the agency is fully engaged in
14
efforts to strengthen labeling of products for use
15 in
the pediatric populations.
16
Today I am going to talk about the issues
17
listed here. First I am going to
review pediatric
18 issues,
especially pediatric safety issues which
19
have long influenced the evolution of FDA law,
20
regulation and policy. That said,
today I am going
21 to
focus on recent milestones, those of the last
22
decade.
23
I will also briefly review the written
24
request process, discuss current pediatric labeling
25 and
exclusivity statistics, the big goals of these
61
1
efforts and pediatric resources that are available
2 at
the FDA Internet web site. For the
standing
3
committee members this will be yet another review
4 and
I apologize for that, though I appreciate Joan
5
Chesney's gracious comments yesterday that no
6
review could be too many.
However, many of you are
7
new, as I just mentioned and have just come for
8
this meeting and this topic is intended to provide
9 you
with a quick primer on how these issues have
10
unfolded at the FDA.
11
[Slide]
12
As in every field, we at the FDA conduct
13 our
work with many acronym shortcuts. You
have
14
your MRI, your PET, your SPECT, your XR, and we
15
have our FDAMA, BPCA, PREA and WR.
The acronyms I
16
will use for my talk are listed here.
The first
17
three refer to recent laws. FDAMA
is the Food,
18
Drug and Cosmetic Modernization Act.
BPCA is the
19
Best Pharmaceuticals for Children Act.
PREA is the
20
Pediatric Research Equity Act. WR
refers to a
21
written request and PPSR refers to a proposed
22
pediatric study request. I will
describe all of
23
these throughout the course of my talk.
24
[Slide]
25
In 1994 FDA issued pediatric regulations
62
1
that required data review for pediatric labeling.
2
This rule required sponsors to review both their
3
existing data as well as available published
4
literature to see if enough data was available to
5
support pediatric labeling. No
clinical studies
6
were required by this rule.
Importantly, this rule
7
introduced the concept of extrapolation of efficacy
8
data from adults to children when that
9 extrapolation
seemed scientifically appropriate.
10
[Slide]
11
In 1997 FDAMA was passed by Congress.
12
FDAMA actually brought the FDA law up to date. It
13 was
a big law that modernized the Food, Drug and
14
Cosmetic Act. Included in this
law were several
15
pediatric provisions, most importantly the
16
exclusivity incentive, which is a big carrot based
17 on
compliance with terms of a written request
18
issued by the FDA to drug sponsors.
Before the
19
passage of FDAMA the pediatric market, with the
20
exception of perhaps antibiotics and a few other
21
product classes, was too small to support a drug
22
development program so pediatric studies were not
23
done. Pediatric exclusivity
changed all of that,
24 as
you will see in a minute. The pediatric
25
exclusivity provisions of FDAMA sunsetted on
63
1
January 1, 2002.
2
[Slide]
3 Now, what is pediatric exclusivity?
4
Pediatric exclusivity is an additional 6-month
5
period during which a sponsor retains exclusive
6
marketing control of all forms of a drug product
7
line. It requires either an
existing patent or
8
exclusivity and is not a patent extension. FDA
9
doesn't have the authority to grant a patent
10
extension; only the Patent Office can do that.
11
Pediatric exclusivity attaches to an existing
12
patent or to other exclusivities which have been
13
granted by the FDA.
14
This is a very powerful economic incentive
15 for
pediatric drug development because it confers
16 to
the entire drug moiety and every product that
17
contains that active drug product.
It delays for 6
18
months the introduction of generic products. As
19
soon as the generic product is introduced the sale
20 of
the branded product declines dramatically.
21
For example, consider the steroid
22
fluticasone. When exclusivity was
granted to
23
fluticaszone it attached to Flovent, the inhaled
24
product; to Flonase, the nasal spray; to Cutivate,
25 the
topical product; and to Advair, the combined
64
1
fluticasone and salmeterol product.
Imagine, for
2
example, a product with 2 billion dollars annually
3 in
sales. Exclusivity translates to an
additional
4 1
billion dollars in sales. So, this is a
very,
5
very powerful economic incentive for pediatric
6
studies, and this was the carrot that made
7
pediatric studies economically feasible.
8
[Slide]
9
I want to touch on one part of FDAMA about
10
which there has been some confusion on the part of
11
industry, the FDAMA priority list.
The priority
12
list consisted of several hundred drugs that were
13
prioritized for pediatric studies by the FDA. If a
14
drug was on the priority list it did not require
15 FDA
to issue a written request. Issuance of
a
16
written request if a drug was on the priority list
17 was
optional. But important for now, this
list has
18
sunsetted. Its sunset was on
January 1, 2002. So,
19 it
sunsetted when the pediatric provisions of FDAMA
20
sunsetted so now this list is a piece of history;
21 it
really no longer exists.
22
[Slide]
23
The next advance I want to mention is the
24
Best Pharmaceuticals for Children Act, the BPCA,
25
which became law on January 4, 2002.
The BPCA
65
1
re-authorized the exclusivity provisions of FDAMA
2 for
on-patent drugs. In addition, it also
includes
3 an
additional mechanism for obtaining information
4 on
the safe and efficacious use of off-patent drugs
5 in
the pediatric populations.
6
There is a slide missing so I am going to
7
tell you what it says. The Best Pharmaceuticals
8 for
Children Act--as I just mentioned, BPCA
9
establishes mechanisms for study of both on-patent
10 and
off-patent products. It requires in
addition
11 the
FDA to collaborate with NIH on these studies.
12 For
off-patent products that is the major focus of
13 the
work of our Office and for on-patent products
14
that industry does not want to study.
So, if
15
industry does not want to study an on-patent
16
product we have a mechanism through BPCA to get
17
studies done aon that product for pediatric
18
labeling, as well as mechanisms for doing studies
19 of
off-patent products. For both on-patent
and
20
off-patent products industry has the right of first
21
refusal to conduct studies that are requested
22
through the written request process.
23
[Slide]
24
There are two paths to a written request.
25
First, FDA can itself issue a written request and
66
1
this happens when the agency determines that there
2 is
a public health need for the studies that are
3
being requested. The definition
of a public health
4
need can vary on many factors, such as whether
5 there
is substantial off-label use; if the proposed
6 use
is a significant pediatric issue; and whether
7
there are other treatment options available.
8
Having a disease be prevalent is not the only
9
factor that we fold into a decision about the
10
public health need. Pediatric
studies for drugs to
11
treat rare diseases may also have a high priority,
12
especially when no other treatment options are
13
available.
14
The other path is when industry submits a
15
PPSR to the FDA. In that
circumstance the FDA may
16
accept the proposal as it is and issue a written
17
request. It may modify the
proposal and issue a
18
modified written request, or it may not accept the
19
proposal at all and the factors that we just
20
described fold into the decision-making process.
21 In
that case, if the FDA decides not to issue a
22
written request then it will issue an inadequate
23
letter.
24
[Slide]
25
Now, what is a written request? A
written
67
1
request is a legal document that provides a
2
detailed outline of the studies needed by the FDA
3 to
adequately label the product for us in the
4
pediatric population. It is an
outline, a detailed
5
outline that does not have the kind of detail you
6
usually see in a protocol. Once a
study is moving
7
forward based on a written request, then a protocol
8 is
developed. The written request specifies
all
9 the
study needs to label the product, including
10
indication, population, types of studies, PK,
11
safety and efficacy studies for example, safety
12
parameters that need to be monitored, whether there
13 is
a need for long-term follow-up and what that
14
might be and the time frame for response. In the
15
next few slides I am going to review the written
16
request process.
17
[Slide]
18
These slides focus on the on-patent
19
process. The off-patent process
is fairly similar.
20 In
this example the industry sponsor submits the
21
proposed pediatric study request to the agency and
22 the
FDA reviews the PPSR to determine whether there
23 is
a public health benefit to the proposed studies.
24
Again, the public health benefit issue here is
25
important. The agency only issues
a written
68
1
request if it determines that there is a public
2
health benefit to the studies. If
so, it issues a
3
written request and, again, if not, it issues an
4
inadequate letter.
5
[Slide]
6
Once the FDA has issued its written
7 request, the industry has 180 days to respond
to
8
that request. If it declines the
request, then the
9 WR
may be referred to the National Institutes of
10
Health Foundation for funding of the requested
11
studies. I would add though that
currently there
12 are
very limited funds available within the NIH
13
Foundation to conduct studies of on-patent
14
products.
15
[Slide]
16
I am not going to talk about this slide.
17 I
want to move on and talk a little bit more about
18 the
on-patent drug exclusivity process because that
19 has
been somewhat of a mystery, what happens at the
20 FDA
in this on-patent written request review
21
issuance, and then review studies once they come in
22 to
the FDA.
23
[Slide]
24
This slide addresses all of that and I
25
want you to focus on the right side of the diagram,
69
1
this column right here. Prior to
issuing a written
2
request the agency does background research on the
3
drug product and the issues at hand and conducts a
4
literature review. That
literature review is used
5 to
inform the drafting of a written request.
The
6
draft request is then reviewed by PdIT, the
7
pediatric implementation team which is a
8
cross-functional team that meets regularly within
9 the
agency to discuss draft written requests.
10
Once the draft is reviewed, has been
11
discussed, has been revised and finally approved,
12 it
is issued to industry by the review division.
13 The
studies are completed by the sponsor, if the
14
sponsor agrees to perform them, and the results are
15
submitted to the agency. So, we
are right here.
16
Once the FDA receives the submitted study
17
reports a time clock starts. It
has 60-90 days to
18
review the reports and make an exclusivity
19
determination. The submission is
reviewed
20
eventually by the exclusivity board which is a
21
cross-CDER team. It is a very
formal meeting and
22 the
team is chaired by Dr. John Jenkins. The
23
review focuses not on whether efficacy has been
24
demonstrated but, rather, on whether the sponsor
25 has
fairly met the terms of the written request.
70
1
That is the legal standard that we must meet. This
2 is
determined by making a very careful comparison
3 of
the submission that we have received from the
4
sponsor compared to the written request that was
5
issued.
6
If, for example, the written request asks
7
that 10 children between the ages of 6 and 10 be
8
included in the study population, then the review
9
carefully checks to see if, in fact, 6 [sic]
10
children were included in the study population in
11 the
submission. If exclusivity is granted,
then
12
that notice is posted on the pediatric page and on
13 the
web. Other actions to the label follow
within
14 a
few months.
15
[Slide]
16
This incentive has really been a
17
tremendous success. Please note
here, this slide
18
reports on industry response to the written request
19
process as of January, 2004. Your
handout may say
20
2003. It is one of those last
minute errors you
21 see
after looking at a slide a dozen times.
To
22
date we have received over 300 proposals from
23
industry. We have issued nearly
300 written
24
requests. We have made
exclusivity determinations
25 for
101 cases and granted exclusivity in 91 of
71
1
those cases. This effort has led
to 63 new labels.
2
The significance of these new labels
3
really cannot be underestimated.
It isn't just
4
data; the labeling changes determine how we use
5
these drugs and provide new information on how to
6 use
these drugs safely in the pediatric population
7 on
issues such as dose, unanticipated adverse
8
events and the like.
9
[Slide]
10
I want to move forward to the present.
On
11
December 3, 2003 the President signed the Pediatric
12
Research Equity Act, PREA, into law.
PREA mimics
13 the
Pediatric Rule which was overturned by the
14
courts in 2002, and this form provides the stick
15
that balances the carrot that I talked about
16 earlier. PREA is retroactive for applications back
17 to
April 1, 1999.
18
[Slide]
19
PREA requires pediatric studies of certain
20
drugs and biologics for the issues listed here: if
21
there is a new indication; if there is a new dosage
22
form; a new route; a new dosing regimen; or a new
23
active ingredient. Biologics are
included because
24
biologics have not been eligible for exclusivity in
25 the
past because they don't have patents.
72
1
The Act also establishes, as was mentioned
2
earlier, a formal pediatric advisory committee and
3
this committee will be seated at the Commissioner's
4
level so it will advise the agency on pediatric
5
issues for most of the FDA centers--for drugs,
6
biologics, foods and devices, probably not
7
veterinary medicine. Its range of
issues will be
8
even broader than that of the current subcommittee
9 which has tackled a number of issues. The range of
10
issues we have tackled since I have been here is
11
just extraordinary.
Implementation of the Act is
12
still under discussion within the agency. The FDA
13 is
currently in the process of developing a
14
guidance to advise on how we plan on implementing
15 the
Act.
16
[Slide]
17
This is our goal for all of these efforts,
18 to
add new pediatric information to the labels of
19
drug products that are commonly used in children.
20
Before pediatrics came to the FDA drugs were
21
commonly used off-label, as I know you all know,
22 and
in that circumstance each child was an N of 1.
23
Little was learned from any of these individual
24
treatment experiments and we already have gathered
25 a
lot of very valuable information since this
73
1
effort has started.
2
[Slide]
3
I want to close by mentioning just a
4
couple of resources that are available on the FDA
5
Internet. If you go to the FDA
home page, which is
6
shown here, at www.fda.gov and you look at the
7
lower right corner--this little arrow right here,
8
there is a little link to the pediatrics web home
9
page.
10
[Slide]
11
Then if you go to the pediatric home page
12
there is a lot of valuable information--statistics,
13
guidances, information about pediatric advisory
14
subcommittee meetings and much, much more.
15
That concludes my comments. I
want to
16
thank you for your attention and I will turn the
17
podium over to Sally Loewke.
18
DR. CHESNEY: Just in advance of
Dr.
19
Loewke, I wonder if all of the speakers who follow
20
her, and including her, could tell us just very
21
briefly, 30 seconds, about your background, please.
22 FDA Perspective
23
DR. LOEWKE: Good morning and
welcome all.
24
[Slide]
25
My name is Sally Loewke. I am the
Acting
74
1
Division Director for the Division of Medical
2
Imaging and Radiopharmaceutical Drug Products. I
3 am
a nuclear medicine physician and I am going to
4
note some bias here. I am a
mother of twins with a
5 son
who has had some cardiac problems, who has
6
actually had to have cardiac catheterization and
7
some cardiac procedures. So, I am
going to throw
8
that out just so you know.
9
[Slide]
10
Dr. Chesney and panel members, I really
11
want to thank you very much for coming here today
12 and
taking time out of your busy schedules to talk
13
about this very important topic, the use of imaging
14
drugs in conjunction with cardiac imaging
15
procedures in the pediatric population.
As you
16
know, cardiac imaging plays an important role in
17 the
management of patients with cardiac disease and
18 to
date we have very few drugs that are approved
19 for
cardiac indications in the pediatric
20
population.
21
We are here today to get needed input from
22 you about the use of these products in the
23
pediatric population. The
information that you
24
will bring forward will be invaluable to the agency
25 as
we proceed in our efforts to provide safe and
75
1
effective drugs for the pediatric population.
2
[Slide]
3
These are several areas that I will be
4
addressing over the course of this presentation
5
this morning.
6
[Slide]
7
The FDA is a regulatory agency.
It is
8
made up of 6 centers. The center
that is
9
responsible for review of drugs for human use is
10 the
Center for Drug Evaluation Research. We
are
11
also known as CDER. An important
piece of
12
information to also take away from this slide is
13
that the devices are regulated by a different
14
center within the FDA, CDRH, Center for Devices and
15
Radiologic Health.
16
[Slide]
17
CDER's mission is to assure that safe and
18
effective drugs are made available to the American
19
people.
20
[Slide]
21
The Division of Medical Imaging and
22
Radiopharmaceutical Drug Products is one of 18
23
divisions that makes up the Office of New Drugs
24
within CDER. The Division is
responsible for the
25
review of drugs that are utilized for diagnostic
76
1
imaging including some radiotherapeutic products as
2
well. The medical imaging drugs
have been broken
3
down into two categories, the contrast agents and
4 the
radiopharmaceuticals. The definitions
you are
5
about to see come from the FDA draft guidance which
6 is
in your packet.
7
[Slide]
8
A contrast agent is a medical imaging
9
agent used to improve the visualization of tissues,
10
organs and physiologic processes by increasing the
11
relative difference of imaging signal intensities
12 in
adjacent regions of the body. Some
common
13
examples of these types of agents include iodinated
14
contrast, gadolinium and microspheres.
15
[Slide]
16
A diagnostic radiopharmaceutical is an
17
article that is intended for use in the diagnosis
18 or
monitoring of a disease or a manifestation of a
19
disease in humans that exhibits spontaneous
20
disintegration of unstable nuclei with the emission
21 of
nuclear particles or photons, or any radioactive
22
reagent kit or nuclide generator that is intended
23 to
be used in the preparation of such an article.
24 One
of the common radioactive tags that is used in
25
nuclear medicine imaging, including nuclear cardiac
77
1
imaging, would be technetium 99-M.
2
[Slide]
3
As an aid to your understanding of the
4
Division and its thinking about the development of
5
medical imaging drugs, you were provided with the
6
draft guidance for developing clinical imaging drug
7 and
biologic products in your preparatory package.
8
This document provides information on important
9
areas that need to be discussed during the course
10 of
drug development. I refer you to the
guidance
11 for
specifics, however, I will briefly touch upon
12 the
types of indications that could be sought for
13
both the pediatric and adult indications.
14
Structure delineation--an imaging agent is
15
able to locate and outline normal anatomic
16
structures and, in doing so, can clarify the
17
spatial relationship of that structure with respect
18 to
other body parts or regions.
19
Disease or pathology detection--an agent
20 is
able to detect and locate specific disease or
21
pathological states.
22
Functional, physiological or biochemical
23
assessment--an agent is able to evaluate function,
24
physiology of biochemistry of a tissue, organ
25
system or body region. This type
of indication
78
1
could apply to an agent that is used to detect
2
either a decrease or an increase of a normal
3
function or physiological or biochemical process.
4
Diagnostic or therapeutic patient
5
management--a medical imaging agent would improve
6
patient management decisions or improved patient
7
outcomes, including predicting survival or patient
8
response to specific therapies.
9
[Slide]
10
To provide you with a framework of the
11
types of information we routinely see when new drug
12
applications come into the agency, I have this one
13
slide. It is not all-inclusive
for the clinical
14
assessment and it is not all-inclusive for the
15
information that we seek in a new drug application
16 but
it highlights a couple of points I wanted to
17
discuss further. For efficacy,
obviously, we
18
review the data and review the studies to make sure
19 an
appropriate dose has been selected that is going
20 to
give you a useful image. We look at the
21
pharmacokinetics and make sure they are well
22
defined.
23
The pivotal Phase III trials are the
24
trials where we get most of our efficacy
25
information and what we like to see is a trial
79
1
design that includes clinically relevant endpoints,
2
relevant patient populations and an appropriate
3
standard of truth.
4
The question is what does all that mean?
5 I
am going to give you an example to help
6
illustrate my point here. It is
not a cardiac
7
example but I still think it makes the point
8
effectively. If you are
developing a medical
9
imaging agent that you felt could distinguish
10
between benign versus malignant lesions, having an
11
agent that could identify a malignant lesion
12
obviously has clinical utility.
Physicians will
13
know what to do with that information and it is
14
very useful. So, you would then
pursue study of
15
that agent in a patient population who would
16
present with a tumor or a lesion that needed
17
further evaluation. Ultimately,
how do you
18
validate the performance of the new drug? You
19
would do so in this case by getting biopsy and
20
confirming the pathology of those lesions.
21
From a safety perspective, we identify any
22
major toxicities that might have come about during
23 the
course of drug development and we put together
24 an
adverse event profile that, if the drug is
25
approved, generally is put into drug labeling.
80
1
So, overall our review and action on a
2
drug, whether it be approval or non-approval, is
3 based
on a risk/benefit assessment. In this
case
4
risk can mean a safety hazard or risk. It could
5
also mean hazard could be occurring from a
6
misdiagnosis as a result of the imaging drug.
7
[Slide]
8
The Division has several drugs in which
9
cardiac indications are approved.
This slide lists
10
drug classes and some of the general indications
11
that are approved in both the adult and pediatric
12
populations. The iodinated
contrast drug class is
13 the
only drug class that has a cardiac indication
14
approval in both the adult and pediatric
15
populations, that being for conventional
16
angiography. The pediatric
approval goes down to
17 the
age of 1.
18
The gadolinium drug products are not
19
approved in either the adult or pediatric
20
populations for a cardiac indication, however they
21 do
have other indications that are approved in both
22
populations.
23
The radiopharmaceuticals--we have approval
24 for
myocardial perfusion identifying cardiac
25
ischemia and other myocardial functional
81
1
assessments such as ejection fraction, wall motion
2 and viability. Again, those are studied and
3
approved in the adult population.
4
Microspheres are one of our most recent
5
drugs that have been on the market.
They have been
6
approved for left ventricular opacification and
7
endocardial border delineation but have only been
8
approved in the adult population.
9
[Slide]
10
Historically, children were felt to be
11
considered like little adults and we could dose on
12 a
milligram/kilogram basis and, therefore, research
13 in
children really wasn't necessary.
However, in
14 the
1970s there was a change in that thinking where
15
people actually felt it was unethical not to study
16
drugs in the pediatric population as many new drugs
17
were flooding the market and were being used in
18
this population.
19
Today, as Susan has mentioned, we have the
20
Best Pharmaceuticals for Children Act and the
21
Pediatric Research Equity Act which are
22
congressionally mandated, and Congress has clearly
23
stated that children deserve the same level of
24
evidence as that provided for the adult approvals.
25
[Slide]
82
1
The agency has tried to foster pediatric
2
drug development and, in doing so, has made
3
comments about the potential use of extrapolation
4
from efficacy data from adults to the pediatric
5
population. Therefore, if the
course of disease
6 and
the effects of the drug are similar in adults
7 and
pediatric patients, then the FDA may conclude
8
that pediatric efficacy can be extrapolated from
9
adequate and well-controlled studies in adults,
10
usually supplemented with other information
11
obtained in the pediatric population such as
12
pharmacokinetic and safety studies.
13
[Slide]
14
When may it not be appropriate to
15
extrapolate? When the disease is
different in
16
etiology, pathophysiology or in its manifestations;
17
when the response to therapy is different; when the
18
pathophysiology may be comparable but the response
19
unpredictable; or when pharmacokinetic parameters
20 are
not well-defined in the adult population.
21
[Slide]
22
We know that there are differences in
23
pathophysiology of cardiac disease between the
24
pediatric and adult populations.
Pediatric
25
population presents with congenital heart disease
83
1 and
the adults with atherosclerotic heart disease,
2 and
most of our drug approvals for cardiac
3
indications in adults have revolved around patient
4
populations that have signs and symptoms of
5
atherosclerotic disease. So, the
question to
6
ponder later today is do differences in the
7
etiology and pathophysiology affect imaging drug
8
performance?
9
[Slide]
10
We have had great difficulty in getting
11
accurate use data of these products.
In an effort
12 to
try to give you some perspective, we looked at
13 the
Child Health Corporation of America's Pediatric
14
Health Information System database.
Currently,
15
this is inpatient data from 31 free-standing
16
children's hospitals with charge level drug
17
utilization information. It is
our first access to
18
pediatric inpatient drug use and, since many
19
children's hospitals are the sites of research
20
trials, we feel that we probably get great
21
information on potential off-label use of these
22
products.
23
This database, however, has a lot of
24
limitations to it. You cannot
nationally project.
25 The
FDA only has access to data dating back to
84
1
1999. There is no direct link
between drug and
2
diagnosis procedure. It does not
capture
3
outpatient use and free-standing image center use.
4
And, the contrast media radiopharmaceuticals are
5
usually bundled together with the imaging procedure
6 and
cannot be specifically separated out.
7
[Slide]
8
So, this is the result of our database
9
search and this is specifically from 26
10
free-standing children's hospitals at the time this
11 was
done. These are drug mentions in the
pediatric
12
population for the years 2001 and 2002 out of the
13
total discharges that you see at the bottom of the
14
slide. The iodinated contrast
agents have the most
15
drug mentions for both 2001 and 2002, followed by
16 the
gadolinium contrasts, radiopharmaceuticals and
17 the
microspheres.
18
[Slide]
19
Since most of our products are not
20
approved in pediatrics we have little knowledge
21
about their safety. I just want
to step back for
22 one
second to make one more comment about that
23
database information on use. We
are fully aware
24
that it is not an accurate representation of the
25 use
of these products because we know many imaging
85
1 procedures are performed on an outpatient
basis and
2 are
performed at free-standing imaging centers.
3 So,
we hope that the discussions later today and
4 the
presentations from our experts will help
5
enhance our knowledge of the frequency of use of
6
these products.
7
[Slide]
8
Unfortunately, we have a limited knowledge
9
base for pediatric safety data as well since we
10
have few approvals. So, in an
attempt again to
11
give you some kind of flavor of what we do know, we
12 did
a data search of the Adverse Event Reporting
13
System, also known as the AERS database.
It is a
14
spontaneous and voluntary reporting system and it
15 too
has many limitations which you heard about
16
earlier today. There is
under-reporting; reporting
17
bias; the quality of the reports is very limited;
18 and
you cannot estimate the true incidence rate of
19
events or exposure risk.
20
[Slide]
21
I just want to go over the
methodology
22
briefly of our search. We did not
want this whole
23
meeting to revolve around any one specific drug
24
but, rather, the drug classes so in an attempt to
25
keep that theme with the search of this database we
86
1
selected two drugs per drug class which we thought
2
were relative market leaders and did a search of
3 the
database in both the adult and pediatric
4 population.
5
Once we got those results, we then
6
combined them and, as you will see, the slides that
7
will be forthcoming are combined data for the drug
8
class per se. We report out the
most common
9
adverse events reported in 10 percent of the total
10 or
greater. We report out the deaths and
the
11
search time frames were variable depending on the
12
specific drug product that we used and their
13
original approval dates. Again,
be warned that
14
this database has its limitations and cannot be
15
construed as an accurate representation of the
16
adverse event profiles for these drug classes.
17
[Slide]
18
This is the data we generated for the
19
iodinated contrast agents. As you
can see here,
20
there were 2,997 reports in the adult population
21
versus 68 in the pediatric population.
The common
22
event types were pruritus, dermatitis and urticaria
23 in
the adults and urticaria, dyspnea and facial
24
edema in pediatrics. There was a
total of 274
25
deaths in the adults and 2 reported in the
87
1
pediatric population.
2
Those 2 deaths in the pediatric population
3
included a 9-year old male having an abdominal CT
4 who
had an anaphylactic reaction and died.
This
5
patient was noted to have a history of asthma. The
6
other patient was a 7-month old with multiple
7
cardiac anomalies who died approximately 6 hours
8
after a cardiac cath procedure.
As you can note,
9
these common events are really a hypersensitivity
10
type reaction and these are very common for
11
iodinated contrast agents.
12
[Slide]
13
This slide represents the gadolinium drug
14
class. There is a total of 5,163
reports in the
15
adult population versus 233 in the pediatric
16
population. Common events in
adults include
17
urticaria, vomiting, nausea, dyspnea and pruritus,
18 and
in children vomiting, nausea and urticaria.
19
There was a total of 108 deaths in the adult
20
population and 3 in the pediatric population.
21
Those 3 deaths were as follows, a 7-month
22 old
with gastroenteritis had an MRI to exclude
23
meningitis. The patient had spina
bifida and the
24
patient died 2 hours after the procedure from
25
septic shock.
88
1 A 12-year old female died from
2
complications of brain stem glioma and a 5-year old
3
male with meningeal toxemia died approximately 8
4
hours after an MRI from complications of
5
hemorrhagic stroke. Again, as I
stated earlier,
6 the
gadolinium drug class does not have a cardiac
7
indication approval in either population.
8
[Slide]
9
The radiopharmaceutical drug class--a
10
total of 334 reports in the adult population versus
11 no reports in the pediatric population. Common
12
events in adults include dermatitis, pruritus,
13
urticaria, nausea, cough, headache and dyspnea and
14 a
total of 16 deaths were reported.
15
[Slide]
16
The microsphere drug class--a total of 107
17
reports in the adult population, no reports in the
18
pediatric population. Common
events in adults are
19
back pain and headache and no deaths reported.
20
[Slide]
21
Overall, to date we have few approvals of
22
cardiac imaging drugs in the pediatric population.
23 We
have limited use data and limited safety data,
24 and
we have the question to ponder whether the
25
differences between cardiac disease processes in
89
1
adults and kids can actually allow us to
2
extrapolate the efficacy data.
3
[Slide]
4
These are basically the questions for the
5
panel that will be coming up either later today or
6
tomorrow. I just flash them on
the screen for the
7
benefit of the audience so you can understand as
8 you
listen to the speakers talk later.
9
The first question basically revolves
10
around extrapolation. Is it
possible? If so,
11
when? The second question is a
series of questions
12
that we would like addressed per drug class
13
category, asking whether there is needed study for
14 the
drug class and, if so, what patient
15
populations, what disease states, etc.
16
[Slide]
17
The third and last question is the
18
relevance of new drug developments in the field of
19
adult cardiac imaging and whether they are
20
applicable to the pediatric population.
21
[Slide]
22
So, we would really like today's focus to
23 be
on the imaging drugs. I know it is hard
to
24
separate the imaging procedure and the device but I
25 ask
that people try. We also know that there
are
90
1
many ethical issues in pediatric research. Again,
2 we
would like today's discussion to focus on the
3
science and trial design issues.
Do we need
4
additional drug labeling, and for what classes, and
5
what do we need to know? How are
these products
6
being used and for what purpose and what
7
population? And, how do they
alter your management
8
decisions, the information that you gather? The
9
bottom line, do you feel that extrapolation is
10
potentially possible?
11
[Slide]
12
I want to thank you very much for
13
attending today. As Susan had
alluded to, we
14 counted
on many people on this panel and others who
15 are
not present to help organize this meeting and
16
your help has been very invaluable and I thank you
17
very much.
18
DR. CHESNEY: Thank you, Dr.
Loewke. We
19
will have time for questions and answers of the
20
speakers after the next two presentations. The
21
next presentation is by Dr. John Ring, representing
22 the
American Academy of Pediatrics, to give their
23
perspective on the issues Dr. Loewke just outlined.
24
American Academy of Pediatrics Perspective
25
DR. RING: One of the advantages
of
91
1
becoming middle aged is that you get a bit
2 farsighted
over time so I am thinking that this
3
will probably work.
4
[Slide]
5
Apropos Joan's request to identify
6
oneself, I have found, now that I am clearly
7
unequivocally middle aged, that it is important for
8 me
to start each day by orienting myself to a
9
person, place and time--
10
[Laughter]
11
--so, this is who I am. This is
where we
12 are
and this is who you are, in case any of you
13
require this type of orientation as well.
14
The five physicians sitting to my right
15
along this part of the table will offer detailed
16
information this afternoon regarding the
17
application of intravascular contrast agents and
18
radiopharmaceuticals to various pediatric cardiac
19
diagnostic modalities. My
assignment is more
20
general. It is to present the
position of the
21
American Academy of Pediatrics as to whether these
22
agents should be studied at all.
I believe I have
23
been selected for this role because I have
24
practiced pediatric cardiology for over 20 years
25
with extensive experience in the cardiac
92
1
catheterization lab and because I am also a member
2 of
the national AAP Committee on Drugs. My
two
3
sons, Jack and Patrick who are sitting in the
4
audience feel that I was selected for this
5
presentation today so that they could miss three
6 days of school.
7
[Laughter]
8
[Slide]
9
The four points which I am about to
10
summarize represent what we know for sure about the
11 use
of intravenous contrast agents and
12
radiopharmaceuticals in pediatric cardiology.
13
These points are that congenital and acquired heart
14
disease is common in children and of considerable
15
clinical importance; that accurate diagnosis is
16
central in order to effect a good clinical outcome;
17
that the diagnostic use of intravascular contrast
18
agents and probably radiopharmaceuticals is likely
19 to
increase in the target patient population; and,
20
finally, that our current use of these agents is
21
guided really by good intentions rather than by
22
data.
23
Taken together, these points identify a
24
clinical problem that is of major clinical
25
significance in children. They
indicate that there
93
1 is
a trend toward increased utilization of these
2
diagnostic units and they highlight what the
3
Academy feels is a glaring deficiency in our
4
knowledge base regarding their use.
5
[Slide]
6 As a good academician I did a
literature
7
search. I did a literature search
in large part
8
because the American Academy of Pediatrics has not
9
given these agents focused consideration and, thus,
10
there are no official AAP policies, technical
11
reports or practice guidelines that speak to their
12
use. Regardless, the AAP
recognizes that in
13
general children's health care needs are unique,
14
that these needs commonly vary with the patient's
15
age, and that optimal pediatric therapy, regardless
16 of
type, is predicated on the performance of
17
appropriate scientific studies performed in
18
children.
19
[Slide]
20
Put very simply, knowledge is good and
21 children
are not little adults. I spoke a minute
22 ago
in regards to a literature search in order to
23 see
what guidance we had there. With the
help of
24
three research librarians at two institutions, the
25
University of Tennessee and St. Jude Children's
94
1
Research Hospital, we searched key words such as
2
intravascular contrast agents and
3
radiopharmaceuticals. We focused
the search on
4
children rather than adults. We
specified that we
5
were most interested in cardiac disease and we had
6 a
particular interest in identifying complications.
7
[Slide]
8
The databases searched are those that are
9
listed and the time frame for the search is a
10
particularly long one.
Unfortunately, but not to
11
much to my surprise, what we found is that there is
12
virtually no information extant in the literature
13
which speaks to the contemporaneous usage of
14
contrast agents in pediatric cardiology or, by
15
extension, radiopharmaceuticals.
16
Something has happened to my script.
17
Well, let's go back to the four things that we
18
actually know for sure.
19
[Slide]
20
What in particular is the scope of the
21
problem? The reported frequency
of congenital
22
heart disease in the population is 2.03 to 8.56 per
23
1,000 live births, with a median figure of 5.93.
24 The
figure that is generally quoted for the quiz is
25 the
higher of these. Even when one requires
more
95
1
firm diagnostic criteria, for example cardiac
2
catheterization, intraoperative inspection or
3
postmortem examination, the figure is still
4
substantial, up to 4.3 per 1,000 live births.
5
We have a population of children with
6
congenital heart disease which is aging.
An
7
article from The American Journal of Cardiology, in
8
1982, so a relatively dated reference, indicated
9
that there were at that time approximately 8,500
10
children with operated congenital heart disease
11
reaching adulthood each year.
Thanks to advances
12 in
diagnosis and therapy that number is actually
13
increasing. In addition, those
patients constitute
14 an
aging population, the natural history for which
15 is
entirely unclear. So, we are obviously
on a
16
voyage of discovery.
17
As far as inflammatory cardiac disease is
18
concerned, the first two points indicate that the
19
incidence and prevalence of Kawasaki syndrome and
20
acute rheumatic fever are substantial in the
21
pediatric population. As far as
myocarditis is
22
concerned, more frequent myocardial biopsy in
23
children coupled with better diagnostic modalities,
24 for
example PCR analysis, are beginning to extend
25 the
scope and define the specificity of this
96
1
diagnosis which to date has been largely
2
descriptive.
3
[Slide]
4
One of the ways in which pediatrics
5
differs from adult medicine is with its focus on
6 the
future. The mission statement of the
American
7
Academy of Pediatrics is very clear on this point:
8 The
AAP is committed to the attainment of optimal
9
physical, mental and social health and well being
10 for
all infants, children, adolescents and young
11
adults. Balance this against the
fact that
12
congenital anomalies are the fifth ranked cause of
13
premature mortality in the United States. That is
14
taken from a reference in Morbidity and Mortality
15
weekly reports in 1998. Of
interest for this
16
group's deliberations, structural congenital heart
17
diseases account for 6 of the 15 most lethal
18
congenital malformations in this group.
19
[Slide]
20
Optimal interventions in pediatric
21
cardiology really do depend, in large part, on good
22
imaging. A good picture is worth
a thousand words.
23
Pediatric cardiologists and cardiovascular surgeons
24 are
visually oriented practitioners. We
cannot
25
treat effectively what we cannot see well. This
97
1
applies both to surgical and catheterization
2
laboratory interventions.
3
Our patient population today is undergoing
4
higher risk interventions both in the cath lab and
5 in
the operating room. These interventions
reduce
6
what we consider to be the acceptable margin of
7
diagnostic error. Our patients
are usually
8 younger,
sometimes much older--for example, adults
9
with grown up congenital heart disease--and usually
10
sicker. They have a limited
tolerance for long,
11
stressful procedures. Accurate
imaging then
12
provides the road map to reach our therapeutic
13
destination in a timely fashion.
Just as the
14
children's oncologist can now choose the safest,
15
most effective treatment for his or her patients
16
with leukemia through use of genetic subtyping, so
17 the
pediatric cardiologist can choose, at least to
18 a
degree, the safest, most effective dilation
19
balloon or closure device provided that he or she
20 has
a detailed and accurate image with which to
21
work.
22
Finally, different imaging modalities are
23
complementary rather than competitive.
The
24
echocardiogram, for example, will certainly
25
satisfactorily define the basic anatomy of
98
1
tetralogy of flow. Angiography,
however, is
2
necessary to dilate and stent the focal pulmonary
3
artery stenoses that often complicate this lesion
4 and
affect its clinical outcome.
5
[Slide]
6
The use of these agents is likely to
7
increase. The volume, for
example, of
8
interventional cardiac procedures performed in
9
children is increasing rapidly and in most centers
10
interventional procedures take place in a third to
11
two-thirds of cardiac catheterizations.
These
12
interventional procedures oftentimes require more
13
angiograms, though of a different type or programs,
14 and
smaller but more frequent injections.
15
The number of adult patients with
16
congenital heart disease is increasing as well.
17
Thus, the assessment of myocardial function and
18
blood flow becomes clinically of greater
19
significance. This may be
particularly true in
20
those structural cardiac lesions which involve
21
abnormalities of coronary arteries, for example
22
transposition of the great arteries or anomalous
23
origin of the left coronary artery from the
24
pulmonary artery. This may apply
particularly to
25
children who survive acute Kawasaki disease but may
99
1 go
on to be at cardiac risk for myocardial
2
ischemia.
3
Our colleagues in interventional radiology
4
apply procedures to non-cardiac areas in pediatric
5
practice as well. For example
embolization of
6
venous malformations in the central nervous system
7 and
catheter-directed thrombolysis have
8
implications for the use of these agents as well.
9
[Slide]
10
Young people search extensive databases on
11 the
web. Older people, like myself, pick up
the
12
telephone and call respected colleagues at big
13
programs. So, what I did to
prepare for this
14
meeting was to query the cardiac cath lab directors
15 at
five programs throughout the United States.
16
Four of these five programs are university
17
affiliated. One is a respected
adult in a
18
pediatric multi-specialty clinic that does a large
19
volume of pediatric cardiac disease.
These five
20
centers do a total of approximately 3,000 pediatric
21
cardiac catheterizations in a year's time. The
22
number of children they catheterize who are under
23 one
year of age is 30-50 percent and in some
24
programs somewhat greater. The
number of
25
interventional procedures performed during these
100
1
cardiac catheterizations at present are upwards of
2 50
percent of these cases. Each of the
programs
3 did
a handful, in one case approaching 5 percent of
4
their cath lab volume, of immediate postoperative
5
catheterizations. All of the
centers had an
6
increasing population of adults with congenital
7
heart disease, 10-15 percent and in some cases
8
larger.
9
What do these inquiring pediatric
10
cardiologists want to know? the
first thing they
11
want to know is are nonionic contrast agents really
12 that
safe or have they just been lucky or good in
13
their practice? The type of
complications that we
14 are
talking about do not really reference nausea
15 and
vomiting; they reflect the sort of
16
complications which are meaningful to this
17
gun-slinging subgroup of pediatricians.
That would
18 be
death, shock, anaphylaxis, life-threatening
19
respiratory distress, gross hematuria, acute renal
20
failure and so on.
21
Their experience is that with the
22 development
of nonionic contrast agents those
23
complications, all of which were seen previously in
24
frighteningly high numbers, have now disappeared
25
almost completely. But there
still is a question
101
1 in
the mind of the practitioners as to what is
2
safe. That is important
particularly when we
3
consider whether there is a maximum volume of
4
contrast that I can inject safely.
Most pediatric
5 centers will limit contrast injection to a
total of
6
somewhere between 5-7 cc/kg of body weight during
7 the
course of a single cardiac catheterization.
8
Some centers have hinted that as they approach that
9
contrast wall they will forego indicated diagnostic
10
procedures till another day for safety-related
11
reasons. Is that a good
practice? Nobody really
12
knows.
13
So, cardiologists would like to know how
14
safe these contrast agents are and does that safety
15
factor vary with age, vary with lesion, vary with
16
co-morbidities, or vary with the program of
17
injection? Are a couple of great,
big angiograms
18
like we used to do better or worse for the patient
19
than a whole bunch of small angiograms that might
20
guide an intervention during a dilation and
21
stenting? The data is simply not
there.
22
Finally, is there an agent that will give
23
adequate opacification at lower volumes of contrast
24
administered in large patients?
This is
25
particularly apropos to that increasing patient
102
1
population, the adult with congenital heart
2
disease.
3
The final question is one that many
4
pediatric cardiologists ask themselves at the end
5 of
the day, especially if their day is ending in
6 the
middle of the night, how can I earn as much as
7 my
colleagues in internal medicine do? I know
that
8 is
beyond the scope of this committee to answer.
9
[Slide]
10
Why wouldn't you study these agents?
That
11 is
the question that I came to ask myself as I
12
tried to prepare these comments.
There may be
13
philosophical considerations at work here. Some
14
feel that data-driven decision-making is of no
15
particular value. Others may feel
that children
16 are
unable for some reason to receive the benefits
17
that accrue to the adult patient through scientific
18
study. Evidence-based medicine
has refuted, I
19
think quite effectively, both of these contentions
20 and
Congress has mandated that the benefits of
21
study should be available to children as well as to
22
adults. There may be some who
believe that
23
clinical resources do not exist to study this
24
problem effectively in children.
25
Each of the institutions I have surveyed
103
1
indicated that they would be pleased to participate
2 in
studies to answer some of the questions that
3
were raised. That doesn't
represent written in
4
stone commitment but it certainly does indicate
5
interest and, coupled both with the incidence and
6
prevalence factors that I spoke of initially,
7
indicates that I think there is a patient
8
population there readily available for study.
9
Finally, there may be some hard-core
10 skeptics
who are either unfamiliar with or frankly
11
doubtful that important practice improvements have
12
been made as the result of the fruits of FDAMA.
13
[Slide]
14
Dr. Cummins pointed you toward the FDA web
15 site
which, much to my surprise, I was actually
16
able to access in a user-friendly fashion. That is
17 a
comment on me; that is not a comment on you.
18
What I found is that the FDA has so far issued
19
approximately 300 written requests and that, as a
20
result of the studies requested, there have been
21
over 90 changes in labeling. I
can say as a
22
pediatrician that fully 15 of those 90 changes are
23
changes that impact my practice, five of which very
24
directly and I am a niche practitioner--studies on
25
midazolam, studies on fentanyl, studies on all of
104
1 the
statins, studies on all of the prils have been
2
important to me as a practicing pediatric
3
cardiologist. As the Carpenters
would say, we have
4
only just begun to gather this information.
5
[Slide]
6
If you look at the exclusivity statistics
7 you
will see that some divisions have been very
8
active in requesting studies in pediatric patients,
9 and
one particular division has not, the Division
10
of--what do you call yourselves?--Medical Imaging
11 and
Radiopharmaceutical Drug Products. We
single
12
this out because it is the subject of today's
13
discussion. We feel clearly, as
pediatricians,
14
that this area deserves study as well.
15
[Slide]
16
So, what are the recommendations of the
17
American Academy of Pediatrics?
We feel that the
18 FDA
should exercise its authority to require that
19
appropriate studies be performed regarding the use
20 of
intravascular contrast agents and
21
radiopharmaceuticals in children cardiac disease.
22
We feel that those contrast
studies should
23
focus on dosing considerations, balancing safety
24
concerns with imaging effectiveness.
As an aside,
25
there is a question in the mind at least of all the
105
1
practitioners as to whether the new nonionic
2
contrasts achieve a comparable level of
3
opacification and, therefore, diagnostic
4
information. Inadequate data or
erroneous data can
5 be
as damaging as no data at all. So,
clearly,
6
that has to be balanced against the safety
7
consideration.
8
Finally, we wonder, and this is just a
9
question, whether a different regulatory posture
10 may
be needed on the part of the FDA in order to
11
study these agents as effectively as others have
12
been studied. It is our
understanding that
13
currently intravascular contrast agents and
14
radiopharmaceuticals are regulated or studied under
15 the
auspices of a device rather than a drug, and we
16 are
not certain, if that is the case, whether this
17 is
the most effective way to pursue that.
18
Regardless of whether it is a drug or whether it is
19 a
device, whether it is done through this division
20 or
that division, we feel there is a substantial
21
problem to address a large pediatric population
22
which can potentially benefit from an informed
23
consideration of these agents.
Thank you.
24
DR. CHESNEY: Thank you, Dr. Ring.
25
Because of how these meetings are run, since Dr.
106
1
Ring is not at the table this is our only
2
opportunity to ask him questions that the committee
3 may
have. Once our next speaker begins we
can no
4
longer ask him questions. Are
there any questions
5 for
Dr. Ring?
6
[No response]
7
Thank you very much.
8
DR. LOEWKE: Excuse me, I just
wanted to
9
clarify that the contrast agents and
10
radiopharmaceuticals are approved at the Center for
11
Drugs.
12
DR. CHESNEY: Our next speaker is
Dr.
13
Geva, from the Children's Hospital Boston. Please,
14 do
give us a few seconds of your background.
15 Cardiologist Perspective
16
[Slide]
17
DR. GEVA: My name is Tel Geva and
I am
18
from the Children's Hospital in Boston.
Just give
19 me
a second here to set this up. I spend
the
20
majority of my time--I divide my time between
21
taking care of children with congenital heart
22
disease and imaging. With regard
to imaging, I
23
divide my time between the cardiovascular MRI
24
program in Children's Hospital in Boston, which I
25
direct, and the echocardiography laboratory.
107
1
[Slide]
2
My task this morning is to give you an
3
overview of progress in the field of pediatric
4
cardiology. This is, of course, a
mammoth task but
5
what I will focus on are the following areas, first
6 the
scope of congenital heart disease; trends in
7
congenital heart disease outcomes; trends in
8
management; trends in imaging of pediatric and
9
adult congenital heart disease; and, finally, I
10
will try to identify some of the gaps in knowledge
11 as
they pertain to imaging.
12
[Slide]
13
As the previous speaker has alluded to,
14 the
incidence of congenital heart disease as widely
15
quoted is approximately 8 per 1,000 live births.
16
This comes from the American Heart Association.
17
With approximately 40,000 patients born every year
18
with some form of congenital heart disease there
19 are
presently approximately a million Americans
20
currently living with congenital heart disease.
21
An extensive review by Hoffman and Kaplan,
22
published in The Journal of the American College of
23
Cardiology in 2002, analyzed 62 studies on the
24
incidence of congenital heart disease published
25
since 1955. They found an
incidence ranging from
108
1
4-50 per 1,000 live births. It
turned out that the
2
variations between those studies had mostly to do
3
with the inclusion of small ventricular septal
4
defects and it has to do with what kind of imaging
5 or
diagnostic modality was used to identify those
6
ventricular septal defects.
7
However, moderate and severe congenital
8
heart disease--the incidence of those is
9
approximately 6 per 1,000. Those
are patients that
10
require some active management of their heart
11
disease, and the incidence of 6 per 1,000 relates
12 to
the population of patients without excluding
13
bicuspid aortic valve. If you
include bicuspid
14
aortic valve, then the incidence increases to
15
approximately 19 per 1,000 live births.
16
[Slide]
17
Here is a rundown of the types of
18
congenital heart disease, and that is taken from
19
that paper published in JACC and the numbers here
20 are
the median incidence per one million live
21
births excluding non-stenotic bicuspid aortic
22
valves and silent PDAs. Also
excluded are tiny
23
ventricular septal defects.
Still, VSD or
24
ventricular septal defect is the most common form
25 of congenital heart disease, followed by
several
109
1
acyanotic congenital heart diseases.
Tetralogy of
2
flow is the most common form of cyanotic congenital
3
heart disease, followed by transposition of the
4
great arteries. If you look down
here, at the
5
bottom, all cyanotic congenital heart diseases
6
account for approximately 1,270 per million of live
7
births; all congenital heart disease, approximately
8
7,600, which is close to the 8 per 1,000; and then
9
bicuspid aortic valve being the commonest form of
10
congenital heart disease. However
it manifests
11
clinically oftentimes later in life.
12
[Slide]
13
Moving on to outcomes of
congenital heart
14
disease first looking at mortality, mortality has
15
consistently decreased over the years.
This is a
16
paper that originated here from the CDC, published
17 in
Circulation in 2001, showing the deaths per
18
100,000, age adjusted, and showing a trend of
19
declining overall mortality from congenital heart
20
disease from 1979 through 1993.
21
[Slide]
22
When you look at age at death, then it
23 turns
out that 51 percent of the deaths occur in
24
infants; additional 7 percent between 1-4 years of
25
age. So, the majority of deaths
occur early in
110
1
life and then it plateaus for several decades until
2 it
starts to pick up again in the elderly.
There
3 are
some racial differences with approximately 19
4
percent higher mortality in Blacks compared with
5
Whites, as found in that paper, and slight gender
6
variations, as you can see from this graph.
7
[Slide]
8
This is data from Children's Hospital in
9
Boston looking at the cardiac intensive care unit
10
admissions--the blue bars here, from 1992 through
11
2003. Here, in red, is the
overall mortality from
12 all
causes in cardiac patients. This does
not
13
capture all deaths from congenital heart disease,
14
nevertheless, the majority do occur in the cardiac
15
intensive care unit and that is a relatively
16
accurate representation of mortality in a large
17
tertiary care acute care referral facility. If you
18
look at the numbers, about 14 years ago overall
19
mortality was approximately 6 percent and that has
20 decreased
quite consistently in the last several
21
years to somewhere between 2.5 and 2.8 percent for
22
overall mortality.
23
[Slide]
24
Still, despite the overall decrease in
25
mortality there are some pockets of resistance and
111
1
there are certain types of lesions that are still
2 at
a high level of mortality. I am just
bringing
3 as
an example pulmonary vein stenosis which is
4
nearly universally a fatal condition.
There are
5
fortunately not too many similar conditions,
6
nevertheless, there are some challenges in the
7
field of pediatric cardiology even when it comes to
8
mortality.
9
[Slide]
10
However, the majority of patients with
11
congenital heart disease survive and the majority
12 of
the therapeutic interventions--surgeries,
13
interventional catheterization, medical therapy--do
14 not
lead to cure. Residual anatomical and
15
functional abnormalities are very common in our
16
patients. Neurodevelopmental
issues are of
17
substantial interest, as well as social and
18
insurability issues.
19
[Slide]
20
As survival of patients with congenital
21
heart disease improved attention shifted from
22
getting these patients alive out of the hospital to
23
improving their functional, psychological and
24
social outcomes. These are just a
few slides
25 showing some of the work that has been done
in that
112
1
field. This is from the
circulatory arrest versus
2 low
flow cardiopulmonary bypass trial where
3
patients with transposition of the great arteries
4
were randomized into circulatory arrest versus low
5
flow cardiopulmonary bypass, and this is the 8-year
6
full-scale IQ results showing that in patients
7
transposition in ventricular septum--their
8
full-scale IQ is nearly normal as a group, whereas
9
patients with transposition in ventricular septal
10
defect who were randomized to the circulatory
11
arrest arm actually as a group,had lower overall
12 IQ.
13
[Slide]
14
There is similar data on patients after
15 the
Fontan operation, again showing full-scale IQ
16
verbal and performance tests, and showing that
17
overall these patients are doing nearly as well as
18 the
normal population.
19
[Slide]
20
Here is a group that doesn't do as well,
21
albeit a small group of patients with interrupted
22
aortic arch. Their performance is
sub-normal in
23 all
levels of tests.
24
[Slide]
25
It is interesting to compare patients with
113
1
congenital heart disease to other pediatric
2
patients with different problems.
This is what
3
this work did, published in Circulation in 2001,
4
comparing physical health summary and psychosocial
5
summary in patients with transposition, asthma,
6
juvenile rheumatoid arthritis and attention deficit
7
disorder and you can see the comparison in this
8
slide. Patients with congenital
heart disease
9
don't do particularly worse than some other common
10
forms of pediatric illnesses.
11
[Slide]
12
I mentioned earlier that patients with
13
congenital heart disease, despite the excellent
14
survival, overall have residual anatomical and
15
functional abnormalities. This is
an example of a
16
22-year old woman who had coarctation repair in
17
infancy so even when we think that our treatment
18
leads to cure, these are some of the complications
19 or
residuals that could develop--a huge aneurism.
20 You
can see part of the dissection right here in a
21
patient about 20 years after repair of congenital
22
heart disease.
23 [Slide]
24
This is an example of a common problem in
25 a
fairly large and rapidly growing population of
114
1
patients, survivors of TOF repair.
Most of them
2
survive and they reach adulthood.
However, most of
3
them have significant pulmonary regurgitation. It
4 is
essentially part of the operation to repair the
5
tetralogy and they have free pulmonary
6
regurgitation which you can see here on this image.
7
Here is a 4-chamber view showing the markedly
8
dilated right ventricle and right ventricular
9
dysfunction. So, these types of
functional
10
abnormalities are quite common in our patient
11
populations.
12
[Slide]
13
Let me switch gears to trends in
14
management of congenital heart disease.
Many
15
variables account for the dramatic progress in
16
treatments of congenital heart disease:
Better
17
understanding of the anatomy, embryology, molecular
18
genetics, pathophysiology and natural history and
19
improved diagnosis and I will come back to that as
20
this is the focus of this meeting.
Support
21
technology has improved dramatically, including
22
cardiorespiratory support and monitoring technology
23 in
the intensive care unit, operating room and the
24
like, development of extracorporeal membrane
25
oxygenators, mechanical assist devices.
Those are
115
1
some examples of improved support technology;
2
pharmacotherapy such as pressors, ACE inhibitors,
3
beta-blockers and the like.
Surgical techniques
4
have improved and transcatheter therapy is playing
5 a
major role in management of congenital heart
6
disease.
7
[Slide]
8
Let me briefly touch on the overall
9
progress in our surgery for congenital heart
10
disease. There has been a revolution
in surgical
11
management of congenital heart disease with early
12
emphasis on a staged palliative approach, with
13
emphasis on treatment of symptoms.
Examples
14
include aortic pulmonary shunts to treat cyanosis
15 in
patients with reduced pulmonary blood flow, or
16
placement of a pulmonary artery band to control
17
pulmonary over-circulation. That
was then.
18
Nowadays there is a growing emphasis on
19
early anatomical repair, with emphasis on
20 restoration of normal physiology with
complete
21
repair of complex anomalies done soon after birth
22 in
patients that are as small as 1.8 kg, with or
23
without the use of cardiopulmonary bypass.
24
Other areas of improvement include
25
protection of vital organs. Areas
of research
116
1
include circulatory versus low-flow bypass that I
2
have mentioned earlier; improved myocardial
3
protection; improved oxygen delivery; and then
4
development of minimally invasive surgeries such as
5
video-assisted thoracoscopic surgery and robotic
6
surgery as an example.
7
[Slide]
8
This is the Da Vinci robotic surgery.
For
9 the
purpose of this presentation, this is in fact a
10 pig
with a coarctation model and the surgeon, in
11
fact, sits right here and this is the robot. The
12
surgeon controls the robotic arms, which you can
13 see
here, from a distance. In this case he
sits
14
next to the operating table. In
fact, it is
15
possible to do that from thousands of miles away.
16
Here is an example of coarctation surgery. This is
17
practice coarctation surgery using robotic surgery.
18
This particular experiment was done by Dr. Pedro De
19
Lido from our hospital. You can
see that the
20
robotic arms are essentially doing pretty much
21
everything that the human arm can do.
What Pedro
22 is
telling me is that the degree of accuracy and
23
control is far superior with this type of approach.
24 In
the interest of time, I will stop here but
25
essentially all of these surgeries can be
117
1
accomplished robotically.
2
[Slide]
3
Moving on to another area where there has
4
been tremendous progress, this is transcatheter
5
therapy of congenital heart disease.
The
6
interventionalists are able to treat a growing
7
number of conditions without the need for a
8
thoracotomy or full cardiopulmonary bypass, valve
9 and
vessel stenosis using balloon stents, radio
10
frequency energy, occlusion procedures for atrial
11 and
ventricular septal defects, collateral vessels,
12
fistulae and the like. There is a
variety of
13
occluding devices and coils available.
Arrhythmia
14
therapy and fetal interventions are only some of
15 the
excellent work that is done in the
16
catheterization laboratory.
17
[Slide]
18
There has been a trend in the
19
catheterization laboratory. This
is the annual
20
case volume in the cath laboratories in Boston from
21
1990 through 2003. I would just
like to turn your
22
attention to two things. Number
one, the overall
23
case load has gone up and down a little bit but
24
hasn't changed dramatically. What
has changed is
25 the
proportion of cases, in pink, of purely
118
1
diagnostic procedures. Not only
did they go down
2 in
absolute terms, but even more so in relative
3
terms. So, the percentage of
non-interventional
4
procedures, in fact, has gone down to less than 25
5
percent. That is, more than 75
percent of cases
6
are, in fact, interventional.
7
[Slide]
8
Moving on to a different area, that is,
9
improved diagnosis which is the focus of this
10
discussion, there has been obviously an evolution
11 in
introduction, development and use of various
12
imaging modalities in the field of pediatric
13
cardiology. Cardiac
catheterization with the use
14 of
X-ray angiography has been the first, dating
15
back to the late 1930s. I am not
exactly sure when
16
nuclear radioactive tracers were first introduced
17 but
I am told that goes many, many years back.
18
However, the modern use of radionuclear cardiology,
19 if
you will, is not as old.
20
Echocardiography came into the clinical
21
arena sometime in the late 1970s.
Use of
22
ultrasound in medicine goes back several years
23
earlier than that but echo has truly revolutionized
24 the
way that pediatric cardiologists practice.
I
25
will not spend time on that.
Needless to say, that
119
1
technology has evolved dramatically and is the
2
primary imaging tool used in the field of pediatric
3
cardiology.
4
CT came to the clinical arena sometime in
5 the
mid-1970s and is continuously improving in
6
terms of resolutions and its role in imaging
7
patients with congenital heart disease certainly
8 has
a place.
9
MRI is the newest kid on the block and is
10 of
particular interest to me. The success
of MRI
11 in
congenital heart disease has to do with the
12
transition from being primarily an anatomical
13
imaging modality to being a much more diverse tool
14
that allows for a comprehensive evaluation of the
15
cardiovascular system including anatomy, function,
16
flow analysis, effusion viability and so on and so
17
forth. Dr. Fogel, I am sure, will
get into that
18
into more detail.
19
[Slide]
20
Just to give you a perspective with regard
21 to
the use of these imaging tools in congenital
22
heart disease, here is the breakdown of use of
23
imaging techniques. I didn't
include CT simply
24
because we don't really have an identifying code
25 for
cardiac CT as opposed to chest CT for various
120
1
lung diseases. So, we don't
really know how many
2 CTs
we perform. Nevertheless, you can see
here
3
that echo by far has exceeded every other imaging
4
modality.
5
[Slide]
6
So, the excellent overall survival of
7 patients with congenital heart disease and
the
8
associated high rate of residual anatomic and
9
functional cardiovascular impairments in these
10
patients result in a rapidly growing population of
11
individuals with a life-long need for surveillance
12
that includes cardiac imaging. In
other words, the
13
patient population that we will be asked to image
14 is
rapidly growing.
15
[Slide]
16
Here is some of the evidence for that.
17
Here is the annual case load in echocardiography at
18 our
hospital. I can tell you that this is
not
19
because of improved marketing or because we have
20
changed dramatically our capture of the local
21
market. This is based on analysis
of the data and
22
mostly has to do with simply the growing
23
population. This is a reflection
of improved
24
survival and the fact that these patients come back
25
again and again and again because they are not
121
1
cured and they need to have continued imaging.
2
[Slide]
3
Similarly, in the cardiovascular MRI
4
program, albeit there are much smaller numbers,
5
this not only reflects evolution of the technology
6 but
also the fact that the same patients come back
7
again and again, and it gives you a flavor as to
8 how
these imaging modalities are used in clinical
9
practice.
10
[Slide]
11
The last issue I would like to
touch on
12 are
safety issues in pediatric cardiac imaging.
13
There are many safety issues that are worthy of
14
in-depth discussion. Not all of
them directly
15
relate to this committee or the other committee or
16
this body of the Food and Drug Administration. I
17 am
listing as many as I could think about.
18
The issue that is unique to pediatrics or
19
nearly unique has to do with sedation.
Young
20
children cannot cooperate with many imaging tests
21 and
the more involved the imaging procedure is, the
22
greater the need for sedation for the patient to
23
stay still, calm, to alleviate anxiety, etc.
24
There are inherent risks of invasive
25 diagnostic procedures that I will not go into
but
122
1
they have to be taken into account.
So, when you
2
have a choice of making a diagnosis or getting
3
information by a non-invasive technique or an
4
invasive technique, the inherent risks of invasive
5
techniques must be taken into consideration.
6
Ionizing radiation exposure--I will come
7
back to that briefly. Contrast
agents is the focus
8 of
this discussion so I will not discuss those.
9
Radiopharmaceuticals, the same.
Auditory trauma is
10
something that is relevant to magnetic resonance
11
imaging. Pharmacological
testing--I am not sure if
12
Mark will touch on that but we are doing a growing
13
number of pharmacological testing in the MRI suite
14
with children. Just to give you
an example,
15
children with Kawasaki disease who have large
16
coronary aneurysms are being sent to us for
17
assessment of myocardial ischemia and viability.
18 So,
we are doing adenosine stress, gadolinium
19
perfusion and viability exams in those children.
20
Lastly, improper use of imaging
21
technology, including an unfavorable risk/benefit
22
ratio--this is not an obvious safety issue but I
23
think it is. I think if a patient
is set for a
24
test such as cardiac catheterization or CT with its
25
risk of ionizing radiation and there is an
123
1
alternative at least as good non-invasive test
2
without those risks, then that patient is exposed
3 to
an unnecessary risk.
4
[Slide]
5
Let me finish off by touching on ionizing
6
radiation exposure. Briefly, this
is a paper that
7 was
published in 2001 in AJR. I am sure many
of
8 you
are familiar with it and, if not, the reference
9 is
available. It looked at the estimated
risk of
10
radiation-induced fatal cancer from pediatric CT.
11
[Slide]
12
This is a graph of pharmacokinetics from a
13
subsequent article. This is the
estimated lifetime
14
attributable risk of fatal cancer in pediatric CT.
15 On
the X axis is age and on the Y axis is the
16
percent risk. So, 0.1 means
1/1,000 will die from
17
cancer related to radiation from CT examination.
18
Notice the relation between age and risk. Here is
19 a
unique issue relevant to the pediatric
20
population. As you get to the
first decade of
21
life, especially during the first 4 years of life,
22
these patients are particularly susceptible to risk
23 of
ionizing radiation.
24
[Slide]
25
Dr. Brenner estimated that above the dose
124
1 of
50-100 mSv protracted exposure or 10-50 mSv
2
acute exposure there is direct epidemiologic
3
evidence from human populations that demonstrate
4
that exposure to ionizing radiation increases the
5
risk of some cancer.
6
[Slide]
7
It takes years to realize the risk from
8
ionizing radiation, as it did for realizing the
9
relationship between cigarette consumption and lung
10
cancer. So, with regard to
cardiac catheterization
11 in
the pediatric age group, this is the first
12
direct evidence or the first paper that I was able
13 to
find that actually demonstrated that link.
This
14 is
a paper published in the International Journal
15 of
Epidemiology in 2002. The reference is
up on
16
top. This group looked at 674
children who
17
underwent cardiac catheterization between 1950 and
18
1970 in Israel, and 28.6 had more than one
19
catheterization. The mean age at
cath was just
20
about 9 years. Mean age at
follow-up was 37.5
21
years. They compared the data to
a national
22
database and the expected number of malignancies
23 was
4.75 whereas the observed number of
24
malignancies was 11, yielding a standardized
25
incidence ratio of 2.3 and you can see the 95
125
1
percent confidence intervals. Of
the 11
2
malignancies, 4 were lymphomas and 3 were
3
melanomas.
4
[Slide]
5
In summary, advances in diagnosis and
6
management of congenital heart disease have led to
7 a
dramatic decline in overall mortality to less
8
than 3 percent. With the rapidly
expanding
9
population of patients with congenital heart
10
disease, currently estimated between 1-2 million in
11 the
United States and growing, patients are rarely
12
cured. Frequent anatomic and
hemodynamic
13
abnormalities require surveillance, that is,
14
imaging. And, there is an
increasing use of
15
transcatheter and minimally invasive surgical
16
interventions that also are based on imaging.
17
[Slide]
18 Consequently, the number of
cardiovascular
19
imaging procedures in patients with congenital
20
heart disease will continue to increase, and there
21 is
an urgent need for research in pediatric cardiac
22
imaging with regard to safety and efficacy of
23
radiopharmaceuticals; the cost and risk/benefit
24
ratio of various imaging strategies; and minimizing
25
exposure to ionizing radiation.
Thank you.
126
1
DR. CHESNEY: Thank you very
much. Your
2
graphics were wonderful. We now
can take questions
3 for
Dr. Cummins, Dr. Loewke and Dr. Geva.
Dr.
4
Fost?
5 Q&A for Speakers
6
DR. FOST: I doubt that you have
numbers
7 on
this but I am interested in how commonly you get
8
adventitious findings with the expanded use of
9
these various imaging procedures.
You mentioned
10 one
study showing 50/1,000 congenital heart disease
11
picking up some clinically insignificant lesions
12 but
I am wondering if there were wider use of
13
various imaging procedures how common do you think
14 it
would be that clinically insignificant findings
15
would be picked up which could lead to both medical
16
risks, that is, impulsion to do further studies and
17
possibly even unneeded therapeutic studies but more
18
invasive diagnostic studies, and psychosocial
19
issues, stigmatization, confusion, parents thinking
20
their child had some severe cardiac disease? How
21
common is that and how do cardiologists handle that
22
now?
23
DR. GEVA: No, I don't have
numbers but,
24 in
the spirit of an overview, I think that overall
25 the
problem is not widespread. I don't think
it is
127
1 a
major problem. Perhaps I have a skewed
view
2
residing in a tertiary referral center.
There are
3
some issues with identification and proper
4
diagnosis of congenital heart disease that have to
5 do
with some of these imaging tests performed by
6
non-experts or by people who don't do that for a
7
living. There has been, for
example, an excellent
8
paper published from UCSF where they looked at
9
accuracy of diagnoses, accuracy of identifying
10
congenital heart disease by echocardiography
11
comparing pediatric echocardiography laboratory to
12
adults and showing significant differences with
13
either misdiagnoses or wrong diagnoses when echo
14 was
done in non-expert hands. Certainly from
15
anecdotal experience, that is true for other
16
diagnostic testing in congenital heart disease.
17
DR. FOST: I was more interested
in the
18
issue of over-diagnosis rather than
19
under-diagnosis, but I am also interested in
20
adventitious findings of extracardiac lesions.
21
That is, you do scans of various types and you pick
22 up
lesions that you weren't even concerned about
23
which are in the body, in the kidney, brain and so
24 on,
some of which may be clinically significant and
25
variable but many and probably most which will be
128
1 of
very uncertain clinical significance. Is
that a
2
common phenomenon? Do you have
any thoughts about
3 the
expanded discovery of such adventitious things
4
with the standard use of imaging, particularly in
5
following up children over the years, and so on?
6
DR. GEVA: It happens. I don't know how
7
common it is. I simply don't have
data that I can
8
provide you with. In the course
of either an
9
echocardiographic examination or cardiac MRI
10
examination we have discovered all sorts of
11
non-cardiac abnormalities, anywhere from thyroid
12
cancer in young patients who get an MRI for
13
congenital heart disease to bronchial cyst picked
14 up
on echocardiogram, and so on. This is
15
anecdotal. I am not aware of a
systematic data set
16
that, in fact, looks at it, that I am aware of.
17
DR. CHESNEY: Yes, Dr. Santana?
18
DR. SANTANA: As a
non-cardiologist, can
19 you
help me understand how these modalities are
20
used in different historical time points for the
21
patient? Do you always get an
echo, a diagnostic
22
cath or MRI diagnosis and then after that you say I
23 am
going to use this modality from now on or I am
24
going to complement it with something else? That
25 is
one question, if you could clarify it for me.
129
1
The second is you obviously come from a
2
large center where you have done a lot of cardiac
3
caths historically. Have you
looked at your data
4 set
in terms of second malignancies in relation to
5
radiation exposure, and how do you quantify the
6
radiation experience for patients receiving all
7
this imaging?
8
DR. GEVA: Let me answer the
second one
9
while it is still fresh in my mind.
We have not
10
looked at the relationship between cardiac
11
catheterization, ionizing radiation exposure and
12
cancer in our center, and that would be an
13
important study to do. We
certainly have the
14
patient population, both in terms of how long the
15
cath laboratory in Boston has been active as well
16 as
sheer numbers. But that study, to my
knowledge,
17 is
not under way.
18
We do have the standard--whatever is
19
mandated by the regulatory bodies--elements in
20
place to monitor radiation but then I have to say
21
that as I started looking into radiation exposure I
22
discovered that this is not as simple as meets the
23
eye. There are various standards
and measures and
24
what is often measured and recorded is not
25
necessarily what is biologically important. My
130
1
suspicion is that you would have to go in and
2
prospectively set up a system to, in fact, evaluate
3 the
amount of radiation that patients are exposed
4 to
that is biologically relevant. Again, I
don't
5
think that we or other places do that.
6
With regard to your first question, I
7
would say that echocardiography is being used
8
widely almost as an extension of the stethoscope.
9 When a question about congenital heart
disease
10
comes up based on clinical suspicion, it almost
11
automatically triggers an echocardiogram. Other
12
tests or other diagnostic imaging testing that
13
comes after that varies quite substantially across
14 the
field, even within a center from cardiologist
15 to
cardiologist whether to catheterize, when to
16
catheterize. Use of cardiac MRI
as a widely
17
available clinical tool is in its infancy. I
18
suspect that is the case for the high quality
19
cardiac CT technology and similarly radionuclear.
20
DR. CHESNEY: Dr. Fink?
21
DR. FINK: Just a quick question,
you
22
presented the spectrum for CT for head and abdomen.
23
Where would cardiac CT fit in that in terms of
24
radiation exposure?
25
DR. GEVA: Closer to abdomen,
number one,
131
1 but
what I did not mention is the fact that these
2
analyses were performed from standard CT
3
examinations. The modern CT
angiography studies
4
using multidetector CTs, in fact, expose patients
5 to
much higher doses of radiation.
6
DR. CHESNEY: Dr. Siegel?
7
DR. SIEGEL: Two comments, one is
8
addressing the incidental findings in imaging. I
9 can
address that from a CT standpoint.
Cardiac CT
10 in
children is still a relatively young tool but in
11 our
experience we have really not found incidental
12
lesions I think in anyone in that population. In
13
adults it is different because there are more risk
14
factors. So, in adults we are
going to see those
15
pulmonary nodules and it is a problem--is it
16
inflammatory or is it tumor? In
children that has
17 not
been the case so far in, again, relatively
18
early experience.
19
The other thing, which I will address in
20
some of my presentation, is the radiation risk with
21
CT. In adults, if you do coronary
CT you are using
22 a
limited area and you can get some high radiation.
23 In
children, when we do cardiac CT we are really
24
examining the entire chest. I
will show you that
25
some of the doses are lower now with the techniques
132
1
that we are using.
2
DR. CHESNEY: Yes, Dr. Maldonado?
3
DR. MALDONADO: This question is
for Dr.
4
Cummins. Before I ask the question
I just want to
5
make the comment that I fully agree with her that
6
this carrot that the BPCA has created is really
7
significant, except that not all the drugs are
8
block-buster drugs like fluticasone or Viagra, and
9 I
am sure you know that Viagra has a written
10
request for pediatrics in the FDA.
It had better
11 be
for a different indication.
12
[Laughter]
13
By saying that, I am not trying to
14
minimize the importance even for all the other
15
drugs that are not block-buster drugs.
For me,
16
working in the pharmaceutical industry, it is a
17
very good tool and it is a good tool that helps us
18 to
balance the fears and the disincentives that
19
have been in place for years, like the liability
20
issues that are very big in the minds of the
21
leaders in the pharmaceutical industry.
22
But there is another element that I should
23
mention, and that is that the fact that the
24 government
has created two laws for pediatric drug
25
development by itself makes a strong statement
133
1
that, indeed, you mean business and it is better to
2
respond to that. Indeed, even
when the economic
3
incentive may not be significant, it is
4
significant--those two statements that the
5
government has made.
6
That leads me to the following question,
7 as
chair of the pediatric working group in PhRMA,
8
with all the other members of that group we do an
9
extensive advocacy because we are not just trying
10 to
use these tools but also advocacy. I
went to
11 the
FDA web site in pediatrics--and by the way, as
12 Dr.
Ring said it is a very good, user-friendly web
13
site--trying to look for the list of the sponsors
14 who
have not responded either because we have
15
refused or basically have not responded to a
16
written request, and I know that the list of
17
non-responders was supposed to be made public and
18
maybe I am looking in the wrong place or may have
19
missed altogether that list of drug companies that
20
have not responded. Why I wanted
that list is
21
because if I can identify those, I can do
22
advocacy--not me personally but through all the
23
members of the pharmaceutical industry--to find out
24 why
they are not responding and maybe correct that
25
problem. But maybe I am looking
in the wrong place
134
1 and
I don't know where that list is.
2
DR. CUMMINS: I am going to defer
to my
3
senior management on that one.
4
DR. D. MURPHY: Dr. Maldonado, I
think
5
what you are referring to is the process where if
6 we
issue a written request and it is turned down by
7
industry and we send it forward to NIH or to the
8
Foundation, then it becomes public.
But if we
9
issue a written request to a sponsor for an
10
on-patent product and they decline it and we do not
11
forward it for some reason, such as additional
12
information has occurred and maybe somebody else's
13
study is done in some other way and we are not
14
going to forward it, then we would not make that
15
information public. So, what you
are asking for is
16
really the list of off-patent plus those that are
17
referred to the Foundation. Is
that correct?
18
DR. MALDONADO: Not the
off-patent, the
19
on-patent drugs that have minimal response from
20
industry to forward to the Foundation.
Some people
21
actually questioned that in the law, saying are you
22
trying just to embarrass those companies by making
23 it
public. That is fine, they can be
embarrassed
24 if
you need to embarrass them but, at the other
25
end, I would like to have that information to see
135
1 if,
through the PhRMA pediatric working group we
2 can
do some advocacy for them to respond.
3
DR. D. MURPHY: I guess one thing
I am
4
just not completely sure is once we send it to NIH
5 or
to the Foundation whether at that point it
6
becomes completely public knowledge.
I mean, after
7 we
get the response from the industry that it is no
8 and
we refer it to the Foundation, it is when that
9
process becomes public that we need to follow-up on
10
with you. Okay? Because we do have a couple that
11 we
are referring to the Foundation. We will
be
12
glad to get those to you as soon as we can.
13
DR. CHESNEY: Dr. Fink?
14
DR. FINK: This is a question for
FDA.
15
From a regulatory standpoint, are there any
16
obstacles or hurdles you would face in doing
17
pediatric studies for some of these indications
18
when the adult studies for similar--well, different
19
indications but the same adult studies of cardiac
20 use
of these compounds have not been performed?
21
DR. D. MURPHY: You say this would
be a
22 new
indication for the drug altogether?
23
DR. FINK: No, most of the FDA
regulations
24
seem to be based on the assumption that adult
25
studies have already been performed and pediatric
136
1
studies then follow on. In some
of these places we
2
would actually potentially be jumping pediatrics
3
ahead of adults because there is not an approved
4
adult indication. Is that a
regulatory problem at
5
all?
6
DR. D. MURPHY: Susan?
7
DR. LOEWKE: I don't believe
so. No, if
8
there is a patient population for which there would
9 be
benefit to study this product we would pursue
10
it. Obviously, we like to rely on
a database of
11
information from adults. That
makes us much more
12
comfortable when we move into pediatrics.
13 DR. CHESNEY: Dr. Glode?
14
DR. GLODE: I also have just a
quick
15
question for Dr. Cummins. If, by
virtue of a
16
written request or a proposed pediatric study
17
request, exclusivity is granted and the company
18 does
three studies in children and all three show
19 no
efficacy, is then automatically the label of the
20
drug changed to say studies have been done in the
21
pediatric population which demonstrated no efficacy
22 or
what happens?
23 DR. D. MURPHY: If they do three studies
24 and
they are all negative, and they came in after
25
BPCA was enacted and after they had gotten the
137
1
letter from us saying they were now under BPCA, all
2 of
those will go up on the web. Those
studies will
3 go
up on the web. The controversy really
now is
4 the
label. The divisions have had different
5
approaches to this depending on the risk of putting
6 the
information in and having that information
7
actually lead to improper use versus putting that
8
information in and thinking that they are able to
9
qualify it or modify it in a way so people
10
understand the context. So, the
bottom line is
11
that sometimes they do put that in the label, that
12 a
negative study has been conducted, because they
13
think that, unlike neuropharm where you may get 10
14 or
12 studies, you know, usually you get positive
15 studies fairly rapidly if they are well
designed
16 and
they think it is important to say, and we have
17 had
that happen where they put that information in
18 the
label.
19
One of the problems we have found is that
20 if
you put information in the label, and
21
particularly if you describe the studies and the
22
dosing that occurred in the study, it is taken as a
23 de
facto indication even when you say that that
24
study didn't show efficacy. So,
there is a balance
25 in
trying to provide information in the label that
138
1
describes the context of that information. In
2
other words, this is three studies out of three
3
really good studies, and they try to tell you how
4
many patients and whatever, and they were negative,
5 or
these are three small studies and we don't think
6
that they were able to tell us that much. That is
7 the
quandary because the label, as you know, is
8
what allows marketing. So, that
is why we have to
9 be
careful what we put in it, even if it negative.
10 So,
it is a balance of trying to put very few
11
sentences in that would describe those negative
12
studies and put them in context and that is why you
13 get
some of them not put in the label.
14
DR. CHESNEY: Yes, Dr. O'Fallon?
15
DR. O'FALLON: A follow-up on that
then,
16 say
pediatricians are needing something, this is an
17
indication that is real in the pediatric
18
population, and they got three negative studies,
19
that is, negative for efficacy but they collected a
20
whole ton of adverse events data, what happens?
21
Does the adverse event data information get into
22 the
label?
23
DR. D. MURPHY: The answer is
sometimes.
24 It
would depend on is it already labeled.
In other
25
words, does the adult indication have the same
139
1
adverse event? And, there might
be a statement in
2
there and they may not say anything additional.
3
However, if there are unique adverse events that
4 are
considered important and significant to be put
5 in
there, yes, they would put that in there.
From
6
yesterday's discussion you can see where that cut
7
might vary but the answer is if they are unique
8
adverse events that are safety issues that the
9
division agrees are solid data, then it would go in
10
there. But I think propyphol is
one of those
11
examples where there was a great concern about what
12 it
meant. You had one positive, one
negative.
13
There was a lot of discussion as to one center
14
driving that data; lots of controversy.
Yet, it
15 was
felt that we could find a way to state in the
16
label in a limited way what the problem was so that
17
safety data did go in.
18
DR. O'FALLON: Because yesterday
we did
19 see
examples in which the statement was made that
20 the
adverse events pattern was similar to that of
21 the
adults and, yet, it really wasn't. When
you
22
looked at it the same things were showing up but in
23
rather significantly different frequencies of
24
occurrence. So, you know, they
say "ah, yeah, they
25 are
seeing seizures." Well, they are
seeing
140
1
seizures in half of one percent in adults and five
2
percent of children. Now, is that
similar? That
3
type of thing.
4
DR. D. MURPHY: That gets to be a
5
discussion within the division.
6
DR. CHESNEY: Dr. Fink, you have
another
7
question?
8
DR. FINK: This is I guess also for Diane.
9 It
sounded like your implication was that, let's
10
say, you took a dermatologic topical that had not
11
shown efficacy in young children but the safety
12
data was okay, if you put that in the label the
13
company could potentially then advertise that the
14
product was safe to use for children down to age
15 two
even though efficacy hadn't been shown between,
16
let's say, in age two and five.
17
DR. D. MURPHY: No, they couldn't
market
18 it
as being proven to be efficacious. I
guess what
19 I
would say is that if you got something in the
20
package insert which says it has been studied and
21
there were no adverse events, that might be
22 utilized in a way that wouldn't be optimal.
23
[Laughter]
24
DR. CHESNEY: Yes, another
question?
25
DR. FOGEL: Yes, this is a
question about
141
1 the
exclusivity rule. It just wasn't clear
from
2 the
presentation how many times can industry
3
actually use it? In other words,
if they come out
4
with one indication and they get the exclusivity
5
rule and then they come up with a second indication
6
does the exclusivity rule go into effect so they
7
have a year's worth of exclusivity?
Or, can it
8
only be used once?
9
DR. ROBERTS: They can actually
have two
10
exclusivities. The first exclusivity
is the one
11
that Susan described in her talk where that six
12
additional months of marketing attaches to the
13
entire moiety or the entire product where they have
14
existing exclusivity or patent to attach to. The
15
second period of exclusivity is much more limited
16 and
has not seemed to be of big interest to
17
industry. We have only had maybe
three to five
18
times where they have actually attempted to get the
19
second period of exclusivity. For
the second
20
period it will attach only to the indication that
21
they receive. Therefore, unlike
the first period
22 of
exclusivity, they actually have to submit a
23
supplement that gets approved and then they can get
24 the
six months of additional exclusivity on the
25
three years of Hatch-Waxman exclusivity that they
142
1
would get with the approved indication.
That has
2
always been available to industry; that is not new,
3
except for the six months of additional pediatric
4
exclusivity. They have always had
the ability to
5 get
the three years of Hatch-Waxman. So, we
don't
6 see
that there has been much interest in that.
7 DR. CHESNEY: I think maybe we have
8
exhausted all the questions. We
are scheduled to
9
begin again at 1:15. Unless I
hear a significant
10
outcry for making it 15 minutes instead of half an
11
hour, I think maybe we will stick with the 1:15.
12
Does the committee have any strong feelings about
13
cutting off 15 minutes?
14
[No response]
15
So, we will reconvene at 1:15.
Thank you.
16
[Whereupon, at 12:45 p.m., the proceedings
17
were recessed for lunch, to resume at 1:15 p.m.]
143
1
A F T E R N O O N P R O C E E D I
N G S
2
DR. CHESNEY: We are still looking
at the
3
possibility of finishing up today.
One suggestion
4
that has been brought to my attention is that we
5
could stay as late as 6:00 or 7:00 this evening if
6
that would significantly affect people's travel
7
plans. If everybody is planning
to stay over
8
tonight regardless of when we finish, then maybe it
9 is
not quite so urgent to finish. Does the
10
committee have any feelings about whether we push
11 on
till later or shall we wait until after the
12
break to make that decision? The
question is are
13 we
having cocktails at 5:00?
14
[Laughter]
15
Well, we will wait until we see how the
16
afternoon progresses and at the break we will make
17 a
final decision, and the FDA has offered to help
18
with getting people tickets out this evening if
19
that is our decision.
20
Our first speaker for this afternoon is
21 Dr.
Mark Fogel who will discuss contrast enhanced
22
cardiac magnetic resonance imaging.
23 Contrast Enhanced Cardiac
Magnetic
24 Resonance Imaging
25
DR. FOGEL: While we are waiting,
my name
144
1 is
Mark Fogel. I am Associate Professor of
2
Pediatrics and Radiology at Children's Hospital of
3
Philadelphia. I am a director of
cardiac MRI. I
4
also spend a good portion of my time in the echo
5 lab
as well. I have been doing cardiac MRI
since
6 1990 so I have seen a decade's worth, at least
a
7
decade's worth of development of the field. I did
8
take a three-year hiatus to run large-scale
9
clinical drug trials for a pharmaceutical company
10 so
I have the unique experience of being able to
11 see
drug development from both sides.
12
[Slide]
13
Today I am going to be talking with you
14
about contrast enhanced pediatric cardiac magnetic
15
resonance imaging. Although MRI
is a multi-faceted
16
technique, what I am going to concentrate on is
17
just the contrast enhanced version of it. What I
18 am
going to talk to you today about--and this is
19 the
order in which the talk is arranged--is the
20
description and properties of the most commonly
21
used contrast agents, in particular gadolinium; how
22 it
is used, for what purpose; the dosing and
23
administration; and then just a brief slide about
24 the
future.
25
[Slide]
145
1
I first want to take 30 seconds and step
2
back a little bit for how MRI generates an image.
3
That is important because you need to know where
4
some of the contrast agents act.
MRI can
5
differentiate tissue by its magnetic properties.
6 You
will see on the screen the four major ways of
7 how
cardiac MRI does that: The hydrogen and
proton
8
density of the tissue; the T1 recovery rates, and
9 T1
is also called the longitudinal vertical or
10
spin-lattice relaxation; the T2 recovery rate,
11
which is also called the
12
horizontal/transverse/spin-spin recovery rates;
13
and, finally, the motion/flow properties of the
14
various tissues.
15
Gadolinium, the major contrast agent in
16
MRI, works mostly in T1, right over here at this
17
portion. Gadolinium itself is, as
I said, the most
18
common contract agent that is used by cardiologists
19 for
contrast enhanced MRI. It has 7 unpaired
20
electrons in its outer shell. It
is paramagentic,
21
meaning that it generates a large magnetic moment
22
when placed in a magnetic field.
It is toxic. It
23 is
a heavy metal. So, the way we have gotten
24
around that is that it is bound to a chelator. The
25
most common one, and I will probably pronounce this
146
1
wrong, is diethylenetriamine pentaacetic acid,
2 abbreviated
DTPA. There are other ways in which
3
gadolinium can be bound to large molecules, like
4
albumin which doesn't diffuse through the capillary
5
membranes, making it a blood pool agent.
However,
6
that has yet to be FDA approved.
7
[Slide]
8
It is an extracellular agent. It
has
9
rapid vascular equilibration and extravasation into
10 the
extravascular tissue. The mechanism of
action,
11 the
way it works is that it increases the
12 relaxation rate of the surrounding protons
when it
13 is
injected in a dose-dependent fashion. As
I
14
mentioned before, it does affect T1 mostly and that
15 is
the major effect of gadolinium. It
decreases
16 the
T1 constant and, therefore, increases the
17
signal intensity of the image.
For your reference,
18 T1
of blood is 1,200 measure and it decreases it
19
down to 100 measure at 1.5 tesla.
The formula you
20 see
on the bottom basically is the way people
21
calculate the relaxivities of the various
22
gadolinium agents, R being the relaxivity constant
23 and
the Gd with the brackets around it is the
24
concentration of gadolinium.
25
It does also affect T2 but that is a very
147
1
minor component of it. It
increases the rate of
2
decay of that and what tissues benefit the most
3
from gadolinium targeting. That
is, if the target
4
tissue, the T1 value is similar to the background
5
but, yet, the target tissue takes up the
6
gadolinium, such as blood, and the rest of the
7
background does not, that is the tissues that
8
benefit the most from gadolinium enhancement. As
9
such, because it affects T1 the most, sequences
10
that have short repetition times, shown here as TR,
11
moderately short echo times, or TE, as well as high
12
flip angle studies are the ones that we use
13
gadolinium with the most.
14
[Slide]
15
Pharmacokinetics is what makes this thing
16
work. You will see why in a
second. Free
17
gadolinium, as you know, is a heavy metal and is
18
toxic, as I mentioned. Its
half-life is actually
19
several weeks. The way we get
around it is
20
chelation, but chelation is a tradeoff.
Chelation
21
decreases the efficiency of increasing the T1
22
relaxation rate and, therefore, increasing the
23
signal intensity. At the same
time, chelation
24 allows
the toxicity to be much, much less. It
25
decreases the toxicity because it allows for the
148
1
excretion of the gadolinium very quickly. When it
2 is
chelated there is a 500 time increase in the
3
rate of renal excretion relative to pre-chelation.
4
When it is chelated its half-life is about an hour
5 and
a half.
6
There are two ways in theory that
7
gadolinium can become more toxic.
One is that
8
increased association from the chelated agent will
9
increase the toxicity. You may
see it in the
10
literature called transmetallation.
What happens
11 is
there are competing moieties, for example copper
12 and
zinc, that displace gadolinium from its
13
chelator and, therefore, allows you to have free
14
gadolinium in the body and, therefore, makes it a
15
little bit more toxic. Of course,
increasing the
16
time of gadolinium in the body also increases its
17
toxicity.
18
[Slide]
19
The median lethal dose for gadolinium DTPA
20 is
10 mmo/kg. To put that in a reference
frame for
21
you, it is 60-300 times the diagnostic dose. The
22
LD50 for two of the more common types of gadolinium
23
preparations is highest Omniscan and lowest
24
Magnevist.
25
Its safety profile is better than
149
1
conventional iodinated contrast agents.
There are
2 a
number of studies. I just picked these
three
3
examples that you see here. There
are few reported
4
fatalities that were temporally related to
5
gadolinium administration, and all those reports
6
seem to question the association of the gadolinium
7
administration with the fatality.
As far as I
8
could tell, there are no known contraindications.
9
[Slide]
10
If you look through the literature,
11
adverse events are very low.
Idiosyncratic
12
reactions are rare. There is a
good review article
13 by
Runge in The Journal of Magnetic Resonance
14
Imaging, in 2000, which I believe is in your
15
packet, that reviews that. In
most of the studies
16 the
AEs that are related to gadolinium are
17
approximately less than 5 percent, with the vast
18
majority being minor, and there is a whole host of
19
transient headache, nausea, vomiting, local
20
burning, cool sensation, hives, temporal increase
21 in
bilirubin and a temporary increase in iron.
22
Anaphylactoid reaction is estimated
23
between 1/200,000 and 1/400,000 doses.
And, it is
24
safe in renal patients even at doses of 0.3 mmo/kg,
25 the
normal dose being 0.1 mmo/kg. It has
been
150
1
studied in numerous papers with patients with renal
2
failure, dialysis, renal A stenosis and renal
3
tumors. There are numerous
reports, although I
4
have to say that the reports that I could pick up
5
were very small numbers, and here are examples of
6
some of the reports.
7
[Slide]
8
There are multiple safety studies for use
9 in
children without danger. This is not for
10
cardiac but it is for other indications so not in
11
patients with congenital heart disease.
There are
12
five papers which I have listed here.
The top one
13 for
example by Marti Bonmati, in investigative
14
radiology, looked for example at lab values or
15
vital sign abnormalities. There
were 51 percent in
16 the
contrast group with an N of 39 and 80 percent
17 in
the non-contrast group with an N of 20.
18
If you take all these five studies
19
together and you lump them together, they encompass
20
doses of 0.1-0.2 mmo/kg, 1,368 children ranging in
21 age
from 15 days to 21 years of age. The AEs
vary
22
between 2 to 5 percent, none of which were serious.
23 [Slide]
24
This is the latest I could find in terms
25 of
the approved MRI contrast agents. The
top seven
151
1 are
gadolinium based. The one right below
the
2
purple box is a manganese ion.
The last two are
3
superparamagnetic iron agents.
These two we don't
4
use, we haven't used at all in cardiac.
If you
5
look at some of the gadolinium agents you can see
6
that there are some differences between them, and I
7
will go into that in a second but since I have the
8
table up here, the highest ones in terms of
9
osmolality are Magnevist and MultiHance and the
10
lowest one is Gadovist. The
osmolality is
11 important
because in case of extravasation of the
12
gadolinium agent you can get pain at the site as
13
well as sloughing so that is an important
14
consideration.
15
[Slide]
16
There are similarities between the
17
gadolinium agents, in particular reporting of
18
adverse events in terms of their frequency being
19
less than 5 percent and the types are all similar
20
between the marketed products.
The dose in general
21 for
all the marketed products is around 1.1 mmo/kg.
22 The
packaging is all the same. A 0.1 mmo/kg
dose
23 in
a 0.5 mmo solution gives you a dose of 0.2
24
cc/kg.
25
The relaxivity, which is the amount of T1
152
1 and
T2 relaxation with a given field strength and
2
concentration, meaning how much it increases the
3
signal intensity in the image, is the same
4
throughout. Therefore, you really
can't tell the
5
difference between the gadolinium agents when you
6 are
examining the images. The nephrotoxicity
for
7 all
the marketed products is none.
8
[Slide]
9
There are differences, as I mentioned.
10
Magnevist has been on the market four years longer,
11 at
least four years longer than some of the others.
12
Magnevist was approved in 1988, ProHance and
13
Omniscan in 1992 and 1993 respectively.
Some of
14 the
products are ionic. Magnevist has a
charge of
15 minus 2, and some of them are nonionic like
16
ProHance, Omniscan and OptiMark.
Their osmolality,
17 as
I mentioned, is different between the different
18
marketed products. The upper
dosage of Omniscan
19 and
ProHance has been approved for up to 0.3
20
mmo/kg. Magnevist, for example,
is only 0.1
21
mmo/kg.
22
[Slide]
23
Now that we have talked about the
24
different types and how gadolinium works, when we
25
administer the gadolinium how do we monitor
153
1
patients during the study? The
personnel that are
2
available are cardiologists and radiologists, a
3
sedation nurse and MRI technician.
The monitoring
4 equipment that we use is direct visualization
via
5
video link, direct audio feed from the scanner,
6
ECG, pulse oximetry and when a patient is sedated
7 we
use end tidal CO2 as well as blood pressure
8
monitoring.
9
[Slide]
10
In terms of the frequency of use, it
11
really depends on the institution.
At Children's
12
Hospital Philadelphia we use gadolinium in a vast
13
majority of cardiovascular cases and I would say
14
that would be approximately 70-90 percent of the
15
clinical cases that we do. Out of
approximately
16 400
cases in the 2003-2004 academic year we will do
17
approximately 330 cases with gadolinium.
The
18
notable exceptions are, of course, patients who we
19 do
an MRI on and they are normal; patients in whom
20 we
are just looking at RV dysplasia, although there
21 is
one paper I believe in the literature that has
22
actually looked at gadolinium and RV dysplasia.
23
And, when we are strictly looking at ventricular
24
dysfunction without perfusion we won't use
25
gadolinium.
154
1
The uses of gadolinium break down into
2
three basic categories, anatomy, blood flow and
3
tissue characterization, and we will go into those
4 in
detail in a second.
5
[Slide]
6
There have been multiple studies in
7
congenital heart disease for anatomy, for efficacy.
8 I
just picked two examples here, one published in
9
2001 which took 73 patients looking at pulmonary
10
artery size anatomy with and without breath hold.
11
Then, one that was published in 2000 that took 38
12
patients with various types of congenital heart
13
disease.
14
Studies investigating blood flow and
15
perfusion and tissue characterization are still
16
underway in the pediatric age group.
The imaging
17
itself you can divide up into two categories, first
18
pass, meaning that the gadolinium is injected and
19 we
take the images during the first pass of the
20
gadolinium through the circulatory system, or
21
delayed enhancement, which means we will let the
22
gadolinium circulate for 5-10 minutes and then do
23 the
study itself. The first pass technique,
in and
24 of
itself, can be divided up into two different
25
kinds. One is the time resolved
where we are
155
1
actually watching the gadolinium enter the body and
2
watching it circulate throughout the circulatory
3
system. One is freeze frame where
we will actually
4 try
and get all the pictures in one image and we
5 are
not following it through the body but we are
6
going to get a static image that has all the
7
gadolinium in it in the area of interest.
8
[Slide]
9
This is meant as an overview.
These next
10
three slides are going to be overview slides of the
11
various uses for gadolinium in congenital heart
12
disease. We will go over them in
detail in a
13
second.
14
This is specifically for anatomy.
This is
15 a
gadolinium enhanced MRI looking at a patient with
16 a
coarctation which you can see right here.
We are
17
basically marching through the body from right to
18
left in very thin cuts. There are
maximum
19
intensity projections which give you a much more
20
three-dimensional picture of the cardiovascular
21
system. This is actually a
patient with a right
22
aortic arch with a coarctation.
There is a shaded
23
surface display where we take the gadolinium volume
24
data set and make a shaded surface display. This
25 is
a patient with an isolated subclavian artery
156
1
which you can see right here.
Those two were
2
freeze framed.
3
[Slide]
4
This is a dynamic injection, a time
5
resolved injection, if you will, where you can also
6 see
the anatomy. This is during an
angiography in
7 the
cath lab. This is a patient who had a
stenting
8
procedure and you can see the upper and lower limbs
9 of
the pathway right here.
10
[Slide]
11
In terms of blood flow, which is the
12
second of the three uses, again you can see blood
13
flow to the lungs and you can actually
14
qualitatively see the perfusion in this time
15
resolved injection.
16
[Slide]
17
Then, of course, there is myocardial
18
perfusion where you can actually look at how well
19 the
myocardium is perfused. The cavities
first
20
light up and then the myocardial tissue itself
21
lights up afterwards.
22
[Slide]
23
Finally, there is tissue characterization
24
which is the third use. One can
identify scarred
25
myocardium, also called delayed enhancement. You
157
1 can
see the arrows here. This is actually a
2
patient after tetralogy flow repair and you can see
3 the
bright tissue here of the ventriculotomy.
You
4 can
actually identify scarred or infarcted
5
myocardium, as well as that different tumors of the
6
heart take up gadolinium in different ways and you
7 can
actually characterize a tumor with whether or
8 not
it takes up gadolinium.
9 [Slide]
10
Now that you know the uses, let's see how
11
they help us when we want see a patient with
12
congenital heart disease. This is
that patient
13
whom we saw earlier who has a right aortic arch
14
with a coarctation. It is
actually a circumflex
15
aortic arch where the aortic arch passes over the
16
right, comes across and goes down the left side of
17 the
spine. So, these are the two-dimensional
18
images that we would normally get.
These are axial
19
images so this is anterior, posterior, and that is
20
right and left. You can see the
aortic arch right
21
over here. If we move a little
bit lower down you
22 can
see the ascending aorta, part of the aortic
23
arch here and then another circle here which is
24
actually the descending aorta. If
we go down a
25
little bit further you can see the aorta crossing
158
1
over to that descending aorta on the left and then,
2
finally, if you move down further you can see the
3
descending aorta right here.
4
Although you are cutting at the picture,
5 you
would like to maybe see it a little bit better
6
than to have to go through cuts.
Basically what
7
gadolinium does for anatomy is that it gives a
8
three-dimensional nature to the picture.
9
[Slide]
10
So, you can look at those straight cuts or
11 you
can look at a maximum intensity projection and
12 see
the squiggly cardiovascular structure that is
13 the
aorta, right here, much better than you can
14
visualize it as you are just going through a
15
two-dimensional cut.
16
[Slide]
17
So, not only can we make it a
18
three-dimensional image and twirl it around any
19
which way we want, we can actually make very, very
20
thin cuts and we can make them parallel to each
21
other or we can make them rotate.
For example,
22 this is a rotation as if you were sitting on
the
23 top
of the descending aorta and turning yourself
24
over from posterior to anterior.
If you follow it
25
here you will see it again as it starts in the
159
1
middle. One branch comes out to
the descending
2
aorta and the other branch comes to the ascending
3
aorta. So, it gives you a lot of
flexibility in
4
terms of visualization and getting a
5
three-dimensional picture in your mind.
6
[Slide]
7
Not only can we do straight cuts, we can
8
also do curved cuts. This is a
patient actually
9
after an arterial switch procedure for
10
transposition of the great arteries and with left
11
pulmonary artery stenosis which you can visualize
12
right here in this axial view.
What we asked the
13
computer to do is to take this axial view and to
14 cut
it in this curved cut and show us what it would
15 look like if we cut it in this particular
plane.
16
This is the resulting image. The
computer
17
basically displays it and you can see the stenosis
18 of
the pulmonary artery here very nicely.
This is
19 the
left pulmonary artery, the right pulmonary
20
artery and the main pulmonary artery right here.
21
[Slide]
22
Of course, if you don't like looking at
23 any
one of those, you can also go to a shaded
24
surface display, again, made from the
25 three-dimensional gadolinium images. This is
160
1
another patient with transposition after arterial
2
switch and you can see how the pulmonary arteries
3
drape over the aorta as the surgeon typically does
4 a
LeConte maneuver for that kind of repair.
5
[Slide]
6
Finally, time resolved gadolinium
7
injection can also help. This
injection was done
8 to
rule out a clot in the superior vena cava.
You
9 can
see here is the gadolinium going in, first
10
lighting up the right side and then lighting up the
11
left side. You can see here is
the superior vena
12
cava and you can see that there is no clot or
13
filling defect in this blood vessel.
14
[Slide]
15
What are the kind of patients we use
16
gadolinium for anatomy? Well, we
use it for
17
patients with coarctation to get a
18
three-dimensional picture of the coarct; patients
19
with supravalvular aortic stenosis to get a
20
three-dimensional picture of that for example in
21
William's syndrome; a dilated aorta for patients
22 for
example with Marfan's. This
three-dimensional
23
image down here, maximal intensity projection, has
24
both the dilated aorta right here, as well as two
25
areas of coarctation right up here, in the
161
1
transverse arch and over here as we start going
2
into the abdominal aortic arch; aortic aneurysms
3 and
dissection as well as vascular rings.
This is
4 a
shaded surface display of a double aortic arch.
5 You
can see why it is called a double aortic arch.
6
Right here are the two limbs of the aortic arch.
7 We
can turn it over the lateral dimension,
8
basically fly over it, and you can see the circle
9
there which creates the vascular ring.
That is why
10 it
is called the double aortic arch.
11
[Slide]
12
So, you can see that there is a whole host
13 of
aortic anomalies, anomalies of the aortic
14
branches like the isolated left subclavian which I
15
repeated again down here that you can see so well;
16 the
relationship of the aorta to the pulmonary
17
arteries which we saw earlier, like in
18
transposition after arterial switch; collaterals
19
from the aorta, for example in patients with
20
tetralogy flow with pulmonary atresia; aortic
21
conduits for complex congenital heart disease; or
22
reconstructed aortas such as aortic-pulmonary
23
anastomosis.
24
This is a three-D shaded surface of the
25
aortic-pulmonary anastomosis. You
can see here is
162
1 the
native aorta and here is the native pulmonary
2
artery connecting to each other, right up here.
3
[Slide]
4
Not only do we use it for the aorta, we
5
also use it for the pulmonary arteries as well.
6
Patients with pulmonary stenosis, like in tetralogy
7 of
flow or pulmonary artery dilation like with
8
tetralogy absent pulmonary valves which you can see
9
right here how dilated the pulmonary arteries are;
10
pulmonary origins, for example in patients with
11
truncus or hemitruncus, or pulmonary artery
12
conduits for patients with heterotaxia.
This is
13
actually a maximum intensity projection of a
14
patient with a left ventricle to pulmonary artery
15
conduit. The conduit starts here
at the apex and
16
goes out to the pulmonary arteries.
Or, patients
17
with reconstructed pulmonary arteries like in
18
Fontan patients.
19
[Slide]
20
We also use it for pulmonary
venous
21
anomalies, anomalous pulmonary venous connections.
22 In
the lower left-hand corner here you can see that
23 we
are going to be marching through the body from
24
anterior, posterior and back again.
This is a
25
patient with an anomalous right pulmonary vein that
163
1 is
entering the right atrium. You can see
it right
2
over here as it comes down, entering into the right
3
atrium right near the IVC close to a scimitar vein.
4
Pulmonary vein stenosis or repaired pulmonary
5
veins, or systemic venous anomalies like anomalous
6
systemic venous connections.
7
This is that normal that you saw earlier
8 as
a comparison. You can see the right side
9
lighting up first and then the left side. Now if
10 you
look at this one, this is actually a patient
11
where all the systemic veins go straight into the
12
left atrium, the right and left superior vena cava
13 and
hepatic veins, and you see as soon as the
14
gadolinium hits everything lights up.
You are not
15
seeing the right side nicely and then the left
16
side; everything lights up so you can basically
17
confirm that, indeed, that is what the patient has,
18 as
well you can identify the left and right
19
superior vena cava.
20
[Slide]
21
How does it help us? As I
mentioned, it
22
gives you a three-dimensional nature to the study.
23 It
helps surgeons and cardiologists visualize what
24 the
anatomy is. It also labels the blood so
you
25 can
visualize the third to fifth generation
164
1
branching of blood vessels. You
can identify small
2
collaterals that can be used for coiling or for
3
unifocalization procedure where we take the
4
collaterals off the aorta and connect them back to
5 the
pulmonary arteries.
6
[Slide]
7
Moving on after anatomy to blood flow,
8
remember, there are two kinds.
One is myocardial
9
perfusion and the other would be lung perfusion.
10 For
the myocardial perfusion what happens is that
11 the
gadolinium is injected and it is followed by
12
time resolved imaging, watching the gadolinium
13
enter the circulation. We image
the myocardium in
14 the
region of interest that we want. So,
what
15
happens is that first the chamber lights up and
16
then the myocardium lights up afterwards. Normally
17 you
should see uniform signal intensity around the
18
entire myocardium. Of course,
abnormal is that you
19
have localized areas of decreased signal intensity
20 when it should be uniform across the entire
21
myocardium.
22
We can analyze this in a number of
23
different ways: qualitatively, just basically
24
eyeballing it; semi-quantitatively, looking at it
25
with time intensity curves, looking at the
165
1
intensity as a function of time in a region of
2
interest; finally, quantitatively, which would be
3
mathematical modeling of the perfusion of the
4 myocardium itself. The way imaging works is that
5 the
images at each slice position are taken at
6
different parts of the cardiac cycle.
7
[Slide]
8
This is actually a patient after
9
transposition of the great artery surgery after
10
arterial switch procedure. You
can see here that
11
first the right ventricle cavity lights up and then
12 the
left ventricle and then the myocardium, and you
13 can
see uniform opacification of the myocardium.
14
However, if you now look here towards the apex and
15
look right down here, you can see how decreased
16
signal intensity just remains even throughout the
17
entire injection, meaning that there is some kind
18 of
decreased flow to that particular part of the
19
myocardium near the apex. That
doesn't necessarily
20
translate into functional problems.
Here you can
21 see
that even though there is some decreased signal
22
intensity in this region, you can see that that
23
region of the myocardium is actually contracting
24
pretty well.
25
[Slide]
166
1
Moving from myocardial perfusion to lung
2
perfusion, you get a qualitative sense during this
3
time resolved injection about the perfusion to both
4
lungs, right and left. Here you
can see how
5
symmetrical they are. Whereas
here, in this
6
patient with left pulmonary artery stenosis you can
7
see, one, how dilated this pulmonary artery is and,
8
secondly, look at the perfusion to the lungs
9
through the generation branch, and how little you
10 can
see over here with the left pulmonary artery
11
stenosis which is right over there.
12
[Slide]
13
The types of patients one uses this for,
14 of
course, the myocardial perfusion would be useful
15 in
patients with coronary artery diseases, like
16
anomalous left coronary arteries from the pulmonary
17
artery; patients with other coronary artery
18
anomalies like the right coronary coming from the
19
left cusp; hypertrophic coronary myopathy; or
20
patients who are postoperative who have had
21
coronary artery manipulation, like patients after
22
arterial switch procedure or patients with a Ross
23
procedure. Of course, for the
lung perfusion one
24 can
use it for pulmonary artery or vein stenosis
25 for
example like in tetralogy.
167
1
How does that help us clinically?
We can
2
identify myocardium at risk and also for the lung
3
perfusion it contributes to physiological
4
information for the branch pulmonary artery
5
stenosis and decrease in lung perfusion, basically
6
confirming other types of imaging that we would do
7
within MRI such as velocity mapping.
8
[Slide]
9
Finally, the third use of MRI is in tissue
10
characterization, also called delayed enhancement.
11 How
does that work? We inject contrast right
over
12
here and the time clock starts.
At approximately
13 one
minute or up until one minute is what we call
14 the
first pass technique. Then, greater than
five
15
minutes is the delayed enhancement technique. What
16
happens is these curves represent the signal
17
intensity or the contrast concentration within
18
various types of myocardium. The
normal, in white,
19
rises during the first pass and then gets washed
20 out
by blood that didn't have gadolinium in it.
21
Ischemic myocardium, in yellow, the same thing--it
22
rises, not as high as the normal myocardium, and
23
then gets washed out. But infarcted
myocardium
24
could do one of two things, in both of which after
25
five minutes the infarcted myocardium or scarred
168
1
myocardium has much more contrast agent in it than
2
does the ischemic myocardium because there is not
3
that normal blood flow to wash it out.
So, that is
4 how
that works.
5
The first pass, as I said, just comes by
6 in
the first minute and that is what we see.
After
7
five minutes is the delayed enhancement where we
8 can
actually identify scarred myocardial tissue
9
that takes up the gadolinium, this infarcted region
10
right over here.
11
[Slide]
12
How do we do this with MRI? This
is the
13
ECG, right up here. At the R wave
we put in a
14
trigger delay and then we do a non-selective 180
15
degree pulse, which means we flip all the protons
16
negative so that nothing has any signal intensity
17 at
all. Then, as we watch them relax, what
happens
18 is
the normal myocardium starts recovering and the
19
infarcted myocardium starts recovering too but they
20
recover at different rates. What
we do is we try
21 to
aim for hitting it right here where the normal
22
myocardium is just about to cross the zero line
23
where it starts to give off signal, and that
24
maximizes the difference between the contrast of
25
normal myocardium and the contrast of the infarcted
169
1
myocardium.
2
[Slide]
3
This is an example of a patient after an
4
endocardial cushion defect. You
can see here that
5
this brightness represents scar tissue, fibrous
6
tissue that has accumulated over the ventricular
7
septal defect patch. In short
axis you can see it
8
right here as well.
9
[Slide]
10
Not only can we look at scarred
11
myocardium, myocardial tumors also take up
12
gadolinium, different kinds of myocardial tumors
13
take up gadolinium differently.
This, for example,
14 is
a patient who had a right ventricular mass right
15
over here, in the apex. This is a
four-chamber
16
view. We injected gadolinium and
you can see in
17 the
four-chamber view how the outside gets more
18
perfused than the inside. The
short axis,
19
unfortunately, didn't help us too much.
But then
20
when you look at the delayed enhancement images you
21 can
see that this is the gadolinium accumulating an
22
incredible amount compared to the rest of the
23
myocardium in the tumor itself in the apex of the
24
right ventricle. That is in short
axis and this is
25 in
the apical four-chamber view and you can see it
170
1
right here.
2
[Slide]
3
What kind of cardiac masses enhance or
4
don't enhance? Hyperenhancement,
tumors such as
5
myomas, hemangiomas, angiosarcomas.
Thrombus does
6 not
enhance. Then there are a couple that
are
7
non-specific as well as some that we haven't seen
8 in
published literature yet.
9
[Slide]
10
The types of patients we use tissue
11
characterization for are, of course, those patients
12 who
have myocardial scarring; patients who have
13
potential for that, patients with coronary artery
14
disease or patients who are postop and, of course,
15 as
I mentioned, patients with myocardial tumors or
16
masses.
17
So, how does it help? It
identifies
18
scarred myocardium and also can contribute to the
19
prognosis in patients with tumors.
20
[Slide]
21
How do we dose gadolinium? The
freeze
22
frame people do it anywhere between a single or
23
double dose. This reference that
is right
24
underneath is actually a reference from Journal of
25
Magnetic Resonance Imaging in 1999 that actually
171
1
recommends double dose for all great artery
2
injections of gadolinium. For the
time resolved
3
ones we can use anywhere between a quarter to half
4 a
dose as a minimum.
5
When it comes to blood flow, that is
6
either the myocardial perfusion or the lung
7
perfusion, we use about half a dose of gadolinium.
8
Finally, with the tissue characterization we will
9
just use a single dose of gadolinium.
People do
10
anything from power injectors to hand
11
administration of the gadolinium itself.
12
[Slide]
13
What does the future hold for gadolinium
14
enhanced cardiac MRI? Newer first
pass agents that
15
have a high relaxivity. A lot of
them are higher
16
concentrations instead of 0.5 mmo.
It is 1.0 mmo
17
solutions. Also, they can have a
higher relaxivity
18 for
either one of two reasons, either increased
19
protein interaction or an inherent increase in
20
relaxivity depending on the chelator that one uses.
21
The blood pool agents, as I mentioned,
22
remain in intravascular space and have more robust
23 imaging
of blood vessels and that could be useful
24 in
coronary imaging.
25
The superparamagnetic iron oxide agents,
172
1
which are not really used in cardiac but there are
2
some studies that are being done, they do have an
3
advantage of having a long intravascular half-life
4
which would be useful for coronary imaging.
5
A now burgeoning field is molecular
6
imaging where the gadolinium is tagged to
7
antibodies or other agents that are directed
8
against receptors and antigens.
Now the 3T
9
systems, the ones with the higher magnetic fields
10 are
now coming on line. They have improved
signal
11 to
noise and better resolution types of sequences.
12
[Slide]
13
Whenever I talk about the future, I always
14
temper that by quoting Yogi Berra who said "it's
15
hard to make predictions, especially about the
16
future." With that, the talk
is over so thank you
17
very much.
18
DR. CHESNEY: Thank you very
much. We
19
will have questions and answers for all the
20
speakers at the end of this session.
Our next
21
speaker is Dr. Marilyn Siegel who is going to talk
22
about contrast enhanced cardiac computed
23
tomography.
24
Contrast Enhanced Cardiac Computed Tomography
25
DR. SIEGEL: While we are bringing
this
173
1 up,
I will just say who I am. I am from
2
Mallinckrodt Institute of Radiology, which is the
3
imaging department for Washington University School
4 of
Medicine. I am a pediatric
radiologist. My
5
areas of interest are cross-sectional imaging and
6
particularly CT, MRI and ultrasound.
I also do a
7
little bit of work in adult imaging, particularly
8 in
chest and cardiac abnormalities.
9
[Slide]
10
This is the list of questions that we were
11
sent by e-mail and I am going to address these
12
individually but, before we do that, let's get a
13
little background information on CT and cardiac
14
imaging, the basic facts.
15
If you are doing this you really need a
16
multidetector CT scanner. What
does that mean?
17
That means with each rotation of the tube we get
18
multiple images. When we first
started CT we were
19
getting a single image, now we can get multiple
20
images. That means that we have
more data and we
21 get
better resolution and image quality. We
get
22
faster imaging times with multidetector CT. I can
23 do
a cardiac study in 20 seconds or less so we are
24
moving patients through.
25 Faster imaging time means fewer
artifacts
174
1 in
children who can't hold their breath. We
get
2
better spatial resolution from 0.5 to 1.25 mm. We
3 can
get superb 3D images and we are getting better
4
contrast enhancement and that is what we need to
5
address, and the use of CT is increasing.
6
[Slide]
7
Contrast using cardiac CT--it is across
8 the
board 100 percent. If we can't get
contrast we
9 are
not doing this study. There are problems
in
10
children which demand the use of contrast--small
11
patient size. They have little
fat which means we
12
can't see structures as well and contrast helps us
13 see
those structures better. Then,
intrinsically
14
there is just poor differentiation of soft tissues
15 on
non-contrast enhanced CT. You can't see
the
16
various chambers and it is hard to see some of the
17
vessels. So, we have to use
contrast.
18
[Slide]
19
Let's start with the first question, the
20
indications for cardiac CT in the pediatric
21
population. Two-fold basically,
first of all to
22
make a diagnosis. Is there
disease or pathology or
23 is
there not? Secondly, to aid in clinical
24
decision-making. Is there a need
for another
25
diagnostic test? Should
angiography be done?
175
1
Should MRI be done? Or, should
there be some type
2 of
intervention? We do not use CT for
defining
3
normal anatomy. We don't use it
for assessing
4
function just yet. It can assess
ventricular
5
function and size and output but there is a problem
6
currently with radiation dose. It
increases when
7 we
look at the heart in different phases such as
8
systole and diastole. It is not a
screening tool.
9 We
have an issue of radiation, which has been
10
brought up and which I will address later.
11
[Slide]
12
What can we use it for? We can
divide
13
this into a couple of categories, extracardiac
14
great vessel anomalies, intracardiac shunt lesions
15 and
then some postoperative anatomy. In
children
16 CT
is performed most often for congenital diseases;
17 in
adults it is usually for acquired disease,
18
although we are seeing more of this use in adults
19 for
congenital diseases or living longer. We
now
20
have an adult cardiac clinic which has about 1,200
21
adults currently with congenital heart disease who
22
have survived infancy. So, I
think we will see
23
more of that.
24
[Slide]
25
The extracardiac lesions--you have seen
176
1
them displayed quite well on MR.
We see the same
2
things--aortic arch anomalies, coarctations or
3
narrowing, complete interruption of the arch, other
4
anomalies such as a patent ductus arteriosus and
5
pulmonary artery sling, these are the more common
6
ones. There are other ones that
aren't as common
7
that I am not going to review now.
8
[Slide]
9
I just want to show you some examples.
We
10 are
using more CT. The reason is that we can
make
11
many diagnoses and obviate angiography which is
12
longer, needs more sedation and has a higher
13
radiation dose. This is equal to
MR but the
14 advantage
of CT is, again, the fast time. I can do
15
this in 20 seconds or less. That
means I don't
16
have to use sedation. Sedation is
required for MR.
17 Of
course, CT has the radiation risks so, as we
18
have heard this morning, it really is a
19
risk/benefit analysis. Some
patients who are
20
critically ill can't have MR and we need to do CT.
21
[Slide]
22
Just to give you a couple of examples, on
23 the
left-hand side we have a neonate with right
24
arch. There was some widening on
the chest X-ray.
25
This clearly shows the right arch.
We don't need
177
1 to
go further.
2
This is an adolescent. We have a
double
3
arch. Here is the right arch,
here is our left
4
arch. This was an incidental
finding. This
5
patient doesn't need additional study.
6
[Slide]
7
Pulmonary sling is an anomaly where the
8 left
pulmonary artery arises from the right
9
pulmonary artery. This is a
neonate, not sedated.
10
Here is the pulmonary artery. Here is the right
11
pulmonary artery and here is the left pulmonary
12
artery arising from the right, crossing behind the
13
trachea to go to the left hilum.
14
[Slide]
15
I mentioned aortic coarctation.
This is
16 one
of the lesions that we see--sorry, we will go
17 to
patent ductus arteriosus next. Patent
ductus
18 arteriosus
is a communication between aorta and the
19
pulmonary artery, short tubular structure
20
connecting them. This is a 3D
CT. Here is the
21
aorta, pulmonary artery and this patent ductus in a
22
very young patient. We can see
similar findings on
23
MR. So, we really are equivalent
and can provide a
24
diagnosis quickly.
25
[Slide]
178
1
The other indications for pediatric
2 cardiac CT, diagnosis of shunts at the atrial
level
3 or
the ventricular level, and then we are using it
4 to
evaluate some postoperative anatomy, usually in
5
very complex cyanotic heart disease.
6
[Slide]
7
This case is a one-year old, no sedation,
8
about 10 multi-center of contrast.
There is a
9
communication between the right atrium and left
10
atrium, atrial septal defect and, similarly, a
11
ventricular septal defect. This
patient had
12
tricuspid atresia and has a graft in place, and
13
they wanted to evaluate residual anatomic
14
abnormalities.
15
[Slide]
16
This is another patient who had a murmur.
17
They thought it was an atrial septal defect. We
18 did
a CT as a first examination--we were beginning
19 to
use CT more. We have contrast going
between two
20
atrial chambers. Here is the
right ventricle, left
21
ventricle. You can see normal
tissue between the
22
two. This is following repair,
right atrium, left
23
atrium and there is no contrast flow; there is no
24
residual septal defect. By the
say, you can see
25
valvular anatomy quite well.
There is the aortic
179
1
valve and you can see the three leaflets here.
2
[Slide]
3
Other postoperative evaluations, this is a
4
patient who had tetralogy of flow, had bilateral
5
Blalock shunts from subclavian artery to pulmonary
6
artery. Here is one; here is the
other. We can
7 see
that they are present and evaluate patency.
8
This is a patient with tricuspid atresia
9 who
had a graft from the right atrium to the
10 pulmonary
artery. That was the purpose of this
11
study, to evaluate the graft.
12
I have only shown you selected cases, just
13 to
show you that we are able to do this study, do
14 it
quickly and do it without sedation in our
15
younger population.
16
[Slide]
17
Next, the impact of CT then on
18
diagnosis--we can make a diagnosis with CT. We can
19
predict whether patients should undergo further
20
invasive diagnostic testing, such as angiography,
21
with CT. We can clarify equivocal
angiographic
22
finding, and we are using it to predict whether a
23
patient might need additional surgery.
24
[Slide]
25
Just to give you a couple more examples,
180
1
this is a patient who had Mustard procedure for
2
repair of transposition of the great vessels. This
3
patient is about 19, comes in with some increasing
4
cyanosis. Contrast is going in
the superior vena
5
cava, coming into the right atrium and going across
6 the
baffle Mustard into the left atrium and there
7 is
a leak here in the conduit which is abnormal and
8
probably accounting for the cyanosis.
9
This patient had a coarctation repair.
A
10
stent was placed and you can see that a
11
pseudoaneurysm has developed and has broken through
12 the
stent. So, we are using this again to
make a
13
decision whether we should go on to angiography or
14
whether there should be a need for additional
15
surgery or intervention.
16
[Slide]
17
Let's get to the contrast specific
18
questions and look at how we do CT, some of the
19
doses, some of the limitations and how we monitor
20
safety.
21
[Slide]
22
Contrast dosing, the contrast volume is
23
simply determined empirically based on patient
24
weight. So, we are giving 2
mL/kg, maximum of 4
25
mL/kg or 125 mL. We are using
nonionic contrast
181
1
medium. This is just standard now
I think across
2 the
country in pediatric divisions, radiology
3
divisions. We are using 280-320
mg of iodine
4
concentration.
5
[Slide]
6
There are two ways of giving this
7
contrast. One is by power
injector, the other is
8
simply pushing by hand. Power
injector is really
9
desired if it can be done, and it requires a
10
catheter in the antecubital region.
The flow rate
11
depends on the size of the catheter in place. If
12 it
is a 22 gauge we are going to use a slower flow
13
rate, about 1.5-2.0 mL/sec. If it
is a 20 gauge we
14 can
use 2-3 mL/sec. I have even used higher
rates
15 of
4 mL/sec and in adults they will go up to 5
16
mL/sec. A 24 gauge central line
can be injected.
17 It
is determined to be safe but you need to use a
18
lower flow rate. If you have a
catheter in the
19
dorsum of the hand or the foot, you have to inject
20 the
contrast by hand or manually.
21
[Slide]
22
The limitations of contrast enhanced
23
CT--the contrast-related ones are extravasation at
24 the
injection site and adverse contrast reactions.
25
Then, there are some that are device related, and
182
1 the
big one is radiation exposure.
2
[Slide]
3
Extravasation, a study by Kaste, in 1995,
4
looked at extravasation with poorer injectors and
5
manual injection, very small, 0.3 to 0.4 percent.
6
With nonionic contrast, lower osmolar, this is not
7 a
problem. We have put a lot of contrast
8
occasionally into a site where it shouldn't be
9
because the catheter is not well positioned or it
10
leaks and sometimes after 100 mL they may feel some
11
fullness but there has been no really adverse
12
sequelae. The contrast gets
resorbed. There is no
13
sloughing of the skin as there used to be with
14
ionic agents.
15
[Slide]
16
Adverse contrast reactions--this is a new
17 one
that I added to the slide set. This was
sort
18 of
a meta-analysis of low osmolar and nonionic
19
contrast media. Looking at a
number of
20
institutions, overall the incidence of all
21
reactions was 1-3 percent minor reactions, meaning
22 no
treatment necessary, maybe minimal rash or
23
itching, or minimal vomiting--the incidence was
24
near 1 percent. Major or severe,
meaning intensive
25
treatment necessary and maybe some life-threatening
183
1
issues such as hypotension or cardiac arrhythmia,
2 is
about 0.4 percent or 1/10,000. Most of
these
3
reactions occurred immediately at the time of
4
injection. Five percent occurred
late, after the
5
time of injection and up to 24 hours.
Mortality
6
rate in series looked at since about 1980 with low
7
osmolar contrast medium, 1/100,000.
That is
8
overall all-comers.
9
[Slide]
10
Now, if we look at children, and this is
11
from a study in Finland and is one of the few I
12
could find that has a larger number of patients and
13
this was a questionnaire study so we have some
14
limitation there. There was a 73
percent return
15
rate. They used Omnipaque. Acute reactions, 1.9
16
percent, so in line with the larger meta-analysis I
17
showed you, and all of them were minor or mild.
18
They usually involved larger patients, older
19
patients who weighed more than 24 kg.
20
Late reactions after the
injection or up
21 to
24 hours were about 6.2 percent of the
22
population, again consistent with the larger series
23
meta-analysis I showed you. These
were mild. Some
24
were intermediate. Intermediate
means some
25
treatment necessary but they are not
184
1
life-threatening. So more severe
vomiting and
2
large amount of urticaria is defined as
3
intermediate. This affected the
younger
4
population.
5
[Slide]
6
There is one more series. This
was one of
7 the
larger ones that had children and adults.
They
8
looked at the overall prevalence of adverse
9
reactions. They found it was
about 3 percent.
10
Severe, 0.04 percent; deaths, 0.004 percent.
11
Seventy percent of the reactions were within 5
12
minutes, the remainder later.
They didn't quite
13
define "later" but I guess 24 hours or maybe even
14
later than 24 hours. But if we
look by age again,
15 for
less than 10 years the overall prevalence was
16 0.4
percent; 10-19 years, 2.52 percent. Once
you
17 get
to adults you get a higher prevalence and then
18
over 50 years it decreases. So, that
is just to
19
give you a handle on how frequently adverse
20
reactions to contrast occur.
21
[Slide]
22
This is the other issue. It is
device
23
related; it is technique related.
It is radiation
24
exposure. This is one of the
headlines in 2001 and
25 we
are still dealing with this. There are a
lot of
185
1
articles that have come out.
There was another one
2
that came out last week. This is
an issue that we
3
need to face when we do these studies.
4
[Slide]
5
So, CT accounts for about 10 percent of
6 all
our X-ray procedures but 65 percent of all the
7
dose we give from diagnostic medical X-rays. Chest
8
X-ray gives us about 0.1 mSv. A
pediatric chest CT
9
ranges between 1-10 mSv. With the
current
10
technology available we are able to do a scan and
11
immediately know how much dose you are giving.
12
This requires a 16-row detector.
The first
13
generation multidetector CTs were 4 rows. We were
14
getting 4 images. Now, with 16
rows this is
15
automatically on the scanner so you know what you
16 are
getting at that time. I have done neonates
and
17 I
have gotten down as low as 1 mSv. I can
get very
18 low
doses, as I will show you in a moment, by
19
adjusting certain parameters.
Adult chest CT, 7-15
20
mSv. Cardiac cath--this is
something given to me
21 by
one of the cardiologists and there may be
22
different numbers available but 20-30 mSv. So, if
23 we
can do multidetector CT well we can reduce this
24
radiation dose if we can obviate cardiac
25
catheterization.
186
1
[Slide]
2
The relative risks to the individual--this
3 is
something given to me by Jim Brink from Yale who
4
looked at a number of articles out there and the
5
lifetime risk of cancer is 20-25 percent or 1
6
person in 4 or 5. Added risk of
CT, 0.05 percent,
7
1/2,000, not statistically significant.
In the
8
population as a whole, there will be about 600,000
9
pediatric CTs in the U.S. per year, and probably
10
increasing. Without CT, 135,000
will die; with CT,
11
135,300 will die, again, not significant to the
12
population but for each individual it is because
13 you
fear one of the children will get that cancer
14 and
that becomes a problem.
15
[Slide]
16
How do we monitor the safety? How
can we
17
have an impact on these risks?
Well, obviously we
18
don't want to overdose. We don't
want to have too
19
much contrast. That leads to a
problem with renal
20
failure, perhaps arrhythmias.
21
[Slide]
22
So, contrast is usually drawn up perhaps
23 by
a technologist at our place, but we always
24
verify the dose prior to injection and contrast is
25
administered by a radiologist or trained personnel.
187
1
Procedurally, we watch the catheter site. We
2
actually feel the catheter site where the contrast
3 is
going in.
4 [Slide]
5
We try to identify patients at risk.
Have
6
they had prior moderate or severe contrast
7
reaction? We are going to try to
get another
8
examination. Medically treated
asthma is a risk.
9 We
heard about deaths this morning and if I am
10
correct one of them did have asthma.
Then, in
11
patients who have had contrast reactions we may
12
premedicate them with corticosteroids.
13
[Slide]
14
Again, the problem is the radiation dose.
15
That is a harder one to deal with.
The dose is
16
directly proportional to several factors: Tube
17
current, the amount of energy that is going into
18 the
patient; the voltage of the equipment; the scan
19
time; the slice thickness; and the total number of
20
slices. If we want to reduce dose
we have to pay
21
attention to each of these factors.
22
[Slide]
23
So, how do we do t? We reduce
dose by
24
optimizing those factors. We use
a lower tube
25
current. For quite a while, if
you look at the
188
1
studies, 200 milliamperage was used in chest and
2
abdominal CTs and in some places it still is. In
3
pediatric radiology now, if we have an infant we
4
decrease it to 25-30 milliamperage.
In an
5
adolescent we might use 80 milliamperage. So, by
6
reducing that we can reduce the dose by half. We
7
reduce the voltage. It is called
kilo voltage. We
8
used 120 for a long time now I am using 80
9
milliamperage or current.
Reducing the kilo
10
voltage will decrease the radiation dose by 30
11
percent.
12
Limited number of scans--in adult cardiac
13
work and liver or pancreas we are using multiple
14
phases, non-contrast, earlier arterial, later
15
arterial, venous delayed. If you
scan 4 or 5 times
16 you
are getting a lot of radiation. Our goal
is to
17 do
it once and, hopefully, get it right and,
18
therefore, minimize some of the radiation.
19
The newer equipment also has automatic
20
dose reduction technology and they will tell you
21 how
low you can go. Of course, if there is
another
22 study
that can be used and the patient is a
23
candidate and can tolerate that study, then that
24
ought to be used.
25
[Slide]
189
1
How successfully are we using CT?
Well,
2 as
you heard this morning there are not a lot of
3
studies out there that address that point. In
4
adults we do have data related to CT angiography of
5 the
coronary arteries and we have dissection
6
information available and aneurysms.
In children
7
there is overall paucity of data.
There are some
8
data available on aortic imaging.
CT in children
9 and
in cardiac work really has just developed
10
within the past two to three years so there are few
11
studies out there. It also is
difficult to get a
12
prospective study because we are dealing with
13
radiation issues. So, designing a
study like that
14 is
going to be a little bit more difficult to do so
15 a
lot of what we are going to see is probably going
16 to
be retrospective analysis looking at series,
17
meta-analysis. There are several
review articles
18
but, again, there is not any type of bench science
19
looking at results.
20
[Slide]
21
In adults, to show you this one slide on
22
coronary artery disease, it can be done quite well.
23 I
have compiled two studies, 95 percent
24
sensitivity, 86 percent specificity detecting
25
cyanosis greater than 50 percent.
The key point
190
1
here is these are small vessels.
We are seeing
2
vessels and stenoses 2-4 mm in diameter.
Given
3
that, we ought to be able to do this in children
4
and, from my experience, we can.
5
[Slide]
6
This is a series that we recently
7
reported. It came out in The
American Journal of
8
Radiology. It was
retrospective. We looked at 22
9
pediatric patients with some type of aortic
10
anomaly, whether it was right arch, double arch,
11
coarctation, patent ductus arteriosus.
All of them
12 did
have some type of confirmatory study to confirm
13 our
findings. We were 96 percent correct and
we
14
could see stenotic vessels, areas of coarct, down
15 to
2 mm. So, again, I think we can do
it. It is
16
going to be a little difficult to prove though at
17
times.
18
[Slide]
19
Direction for CT as far as drug
20
development or utilization of contrast agents,
21
well, the goal of CT is to get the highest contrast
22
enhancement with the least amount of contrast
23
agent. So, when we do contrast
enhanced CT we want
24 a
high level of contrast enhancement and a smaller
25
amount of contrast agent. What
affects contrast
191
1
enhancement? The flow rate and
iodine
2
concentration. Let me show you
that.
3 [Slide]
4
If we look at different injection rates
5
keeping everything else stable and we use a 5 mL
6
flow rate, 3 mL and 1 mL, with faster flow rates we
7 get
higher enhancement, higher density, higher
8 attenuation. Increasing the injection rate
9
increases contrast enhancement.
Theoretically, if
10 we
increase the contrast enhancement by increasing
11 the
injection rate we should be able to use a
12
smaller volume of contrast. If it
goes in quicker
13 we
get a higher contrast enhancement with smaller
14
volume.
15
[Slide]
16
There is a problem in children.
Because
17 we
have smaller catheters, sometimes we can't use
18
that fast flow rate. In our
adolescent population
19 we
can but not necessarily in our neonates.
20
[Slide]
21
The next thing we can look at is what
22
about the concentration? In this
model where they
23
looked at different concentrations but keeping the
24
iodine mass and flow rate constant, you can see
25
that as the iodine concentration increased, 400 mg
192
1 of
iodine, 350 mg and 300 mg, you got better
2
contrast enhancement. So, perhaps
we can get more
3
iodine concentration in there and get better
4
enhancement.
5
[Slide]
6
If we did that, we should be able to
7
decrease the volume. Well, Becker
looked at
8
concentration and actually looked at flow rates and
9
looked at left ventricular density in adults, and
10 he
used 300 mg iodine/mL and a flow rate of 2.5
11
mL/sec and he also used a higher concentration of
12 400
mg at 2.5 mL/sec. He found that if he
used the
13 low
concentration and high flow rate he got the
14
same result as a high concentration and a lower
15
flow rate. So, higher
concentrations work.
16
[Slide]
17
Implication in children--if we can use
18
higher concentrations, as I mentioned, we may get
19
smaller contrast volumes. This is
the problem, the
20
viscosity. Once you get out to
400 or more you
21
can't push it through a smaller catheter. So, the
22 challenge
perhaps for manufacturers is can we get
23
that high contrast or concentration out there and
24 can
we inject it?
25
[Slide]
193
1
What about future clinical utilization?
I
2
think we are going to see some ventricular function
3
studies based on images in systole and diastole.
4 As
soon as we learn how to keep the radiation dose
5
down that is a potential. But I
think we will see
6
more perfusion studies, pulmonary perfusion
7
studies. Basically, what we are
looking at here is
8
measuring density or attenuation value, peak
9
attenuation and time to peak attenuation.
10
[Slide]
11
Just one more example here. In this case
12 we
segment out part of the lung. By
computer we
13 are
able to subtract the lung, remove the soft
14
tissues and remove the heart because now we want to
15
look at the perfusion going to the lungs. We can
16
look at one lung. The right is in
blue, the left
17 in
green. Or, we can look at both lungs and
we can
18
look at the densities of the whole lung, which I am
19
showing you in this case. I can
segment out and
20
look at one lung. I can look at a
part of a lung.
21 I
can do measurements or attenuation value on this.
22 I
can also apply color to this and look at
23
perfusion to the lung.
24
This work has not been done in children
25
yet. We are probably going to
start this with some
194
1 of
our lung transplants to look at perfusion to the
2
lung. This has been done in
adults. They have
3
looked at perfusion in patients with pulmonary
4
emboli but I think this has the potential to look
5 at
perfusion abnormalities associated with heart
6
disease as well, how much blood supply is there
7
really going to the lungs.
8
[Slide]
9 In summary, we are going to be
seeing more
10
CT. It is out there. It is being used more and it
11
certainly can provide a diagnosis impact here. The
12
challenge as far as the contrast medium goes is can
13 we
optimize contrast enhancement? We have
14
discussed that. The other
challenge for us is can
15 we
lower the radiation dose? At that point,
I will
16
stop and thank you.
17
DR. CHESNEY: Thank you. Our next speaker
18 is
Dr. Phillip Moore who is going to speak about
19
contrast enhanced invasive cardiac imaging.
20
Contrast Enhanced Invasive Cardiac Imaging
21
DR. MOORE: While the computer is
being
22
switched over, I will introduce myself.
I am an
23
Associate Professor of Pediatrics at University of
24
California San Francisco and I run the congenital
25
cardiac catheterization laboratory there.
195
1
I was asked to give an overview of
2
interventional catheterization and its current
3
relationship to imaging modalities and some of the
4
imaging agents. So, i will try to
do that for you
5 in
the next little bit. Tom, I am either
going to
6
need your password or need your help, one or the
7
other. I will take either. He chose the less
8
interesting option, at least for us!
9
[Slide]
10
The role of interventional catheterization
11 has
changed over the years since the early '80s
12
when it initially developed from basically blowing
13 a
balloon up into a clogged artery to a variety of
14
things. With respect to
congenital cardiology, if
15 you
look at the history surgery really developed in
16 the
1940s with initiation of PDA ligation and BT
17
shunt, with a huge explosion in the 1950s with the
18
development of cardiopulmonary bypass, allowing
19
application to complex disease.
Then, in the '60s,
20
'70s and into the '80s really the application of
21
newer techniques and to younger and younger
22
patients with congenital heart disease.
23
Surgery now has settled down a little bit
24 in
terms of its development, other than some of the
25
newer issues that Tal mentioned.
Interventional
196
1
catheterization, on the other hand, is tracking
2
this to some degree but starting not until the late
3
'50s, early '60s with initially balloon septostomy;
4
then an attempt at PDA and ASD closure in the '70s
5
that really got rolling in the '80s and the '90s.
6
Now, in the 2000 decade we are starting to see
7
application of some of these more simple procedures
8 to
more complex disease, such as hypoplastic left
9
heart and the initiation of pulmonary valve and
10
aortic valve implants.
11
[Slide]
12
The cath lab nowadays however still
13
consists primarily of angiography and radiography
14 and
the contrast agents that go with it, although
15
that is changing and I will take you through that a
16
little bit. If you look at the
impact of
17
interventional catheterization on congenital heart
18
disease, it is starting to become relatively
19
significant. This is a slide that
was shown
20
earlier in the day, just looking at the incidence
21 of
different types of congenital heart disease
22
lesions. You can see the common
ones, VSD, PDA,
23
ASD, pulmonary stenosis, coarctation.
I have
24
highlighted in yellow those that are now primarily
25
treated in the interventional cath lab.
Those in
197
1 red
are lesions that are really shifting nowadays
2 and
we will have to see what happens over the next
3 ten
years, but from surgery to the cath lab.
Even
4
those more complex lesions, in green, often utilize
5
interventional techniques in association with
6
surgical treatment.
7
[Slide]
8
So, it is really becoming quite
9
significant. This is the data
from UCSF which is
10 not
unlike the data from Boston. We have had
a
11 steady
increase in the number of patients we see a
12
year in the cath lab, and some of that is
13
significantly related to adult congenital heart
14
disease. But you can see--in
yellow is diagnostic
15 and
in red is interventional--that there really is
16 a
dramatic shift over the last ten years to
17
treatment modalities in the cath lab rather than
18
just diagnostic.
19
[Slide]
20
The impact, if you look at it globally, is
21
quite significant. You have seen
some of these
22
numbers already and 32,000 to 40,000 infants a year
23 are
born in the U.S. with congenital heart disease.
24 In
fact, about 60 percent of those will require
25
treatment at some point during their lifetime.
198
1
Right now about a third of those patients can be
2
treated in the cath lab and with advancing
3
modalities, both in interventional technique and
4
imaging, as well as imaging drugs, the potential
5 for
up to two-thirds of these patients for
6
treatment in the cath lab may be possible.
7
[Slide]
8
There are a variety of approved procedures
9
already that are listed up here.
They are not all
10
that important to this discussion but they
11
encompass a variety of different techniques and
12
devices for a variety of different lesions that are
13
currently performed.
14
[Slide]
15
There are some very interesting and
16
exciting investigational procedures that are being
17
developed, including valve stent implantation for
18
both pulmonary insufficiency and aortic
19
insufficiency, the latter of which might have quite
20 a
substantial impact on adult acquired disease;
21
covered stent implantation in more complex lesions
22
such as Fontan completion and shunt palliation in
23
infants; internal vessel banding for hypoplastic
24
left heart palliation; and intravascular suturing
25
which is just really in its infancy but may have
199
1
some wide-reaching implications.
All of these are
2
going to require very, very specific improvements
3 in
imaging to take these to the next level in the
4
interventional cath lab.
5
[Slide]
6
One of the difficulties, which you have
7
already sort of touched on today, is that the range
8 of
patients is very huge, from premature infants
9
down as low as 600 mg for valvular pulmonary
10
stenosis in some institutions to adolescents, young
11
adults and even nowadays some middle-aged adults
12
with congenital heart disease.
That obviously
13
makes the application to imaging modalities and
14
imaging drugs quite problematic.
15
[Slide]
16
Currently, in the cath lab by far and away
17
radiography or fluoroscopy is the prime imaging
18
modality that is used and nonionic contrast is the
19
drug of choice that is used. In
fact, this really
20 has
been studied quite a bit both in adults and
21
pediatrics with regard to cardiac imaging and
22
probably doesn't warrant a huge amount more issues.
23
We also use echocardiography, both
24
surface, transesophageal and intracardiac imaging
25 in
the cath lab in interventional procedures
200
1
primarily. For contrast, it is
agitated saline
2
although some Optison type contrasts are currently
3
being used.
4
[Slide]
5
This is just to give you an example of
6
angiography. This is a lateral
X-ray or angiogram
7 of
a patient who has had a tetralogy repair and has
8
some compression of the repair site in between the
9
right ventricle and the pulmonary arteries. We use
10
that to define the anatomy, but you can see that
11 you
are quite limited here in terms of
12
intravascular structures. You
obviously don't see
13 the
myocardium; you don't see soft tissue
14
structures around it.
15
Then, we also use this, including nonionic
16
contrast, in some of the tools we use.
This is a
17
stent implantation to open that up.
Then,
18
afterwards again nonionic contrast angiography to
19
look at the area where we have implanted the stent
20 for
improvement in the stenosis.
21
[Slide]
22 This is just an example of an ASD
closure,
23
using fluoroscopy here to define the delivery of
24 the
device. This little tube right here is
25
actually intracardiac ultrasound.
We are getting
201
1
ultrasound pictures while we are implanting. Then,
2
using some nonionic contrast at the end of the
3
procedure to confirm position of the device. But,
4
again, you can see we are quite limited in terms of
5
sort tissue definition here.
6
[Slide]
7
We pick up some of that in the cath lab
8
with the use of echocardiography.
This is an
9
example of an intracardiac echocardiogram. So, the
10
right atrial space is up here; the left atrial
11
space is up here. We are
evaluating the defect.
12
This is a balloon that is passed through the wall
13
here that has a hole in it. Now
we are getting
14
ready to deploy a device. This is
a CardioSeal
15
type device that has been opened in the left atrium
16 and
now we are bringing it back against the atrial
17
septum. As I mentioned, we do
occasionally use
18
some contrast with regards to echo in the cath lab
19 to
assess position of devices. Again, this
is the
20
atrial septum hole; the device being positioned;
21 the
other side of the device has been deployed.
22 Now
the device has ben released and you can see we
23 get
much better soft tissue definition here.
We
24
will sometimes use, obviously, colored Doppler but
25 you
will see some injection of some agitated salien
202
1
contrast up here to look for any residual leak.
2 But
there are limitations to that technique in
3
terms of some of the modalities we use.
4
[Slide]
5
How significant are complications or
6
problems with currently used nonionic contrasts?
7
They are really fairly limited.
If you look at
8
just all complications associated with
9
catheterization in children, particularly
10
interventional caths, you find that major
11
complications are quite rare, less than 2 percent;
12
minor complications less than 10 percent. In fact,
13 the
risk factors for complications are really
14
related to age, less than a couple of years, and
15
interventional procedures. If you
look at the
16
larger series the use of contrast and types of
17
contrast do not really fall out in terms of major
18
issues for risk factors.
19
There are, however, well-known and well
20
described risk factors associated with contrast
21
that is currently used. Transient
renal failure
22 occurs,
is dose dependent, and there are allergic
23
reactions that I think have been discussed. That
24
being said, we are becoming more and more specific
25
with the use of some of these additional imaging
203
1
modalities in terms of our judicious use of
2
contrast in the cath lab, and these complications
3 or
side effects are being reduced.
4
[Slide]
5
What adjunct imaging modalities are
6
currently used and associated with interventional
7
treatment? The one that is most
common at our
8
institution would be MRI or magnetic resonance
9
angiography, particularly as it pertains to arch
10
abnormalities, coarctation, pre and post anatomy
11
evaluation, as well as flow determination and
12
patients who have right ventricular dysfunction,
13
pulmonary insufficiency, particularly tetralogy or
14
flow patients. I should add that
at other
15
institutions CT might, in fact, be the imaging
16
modality of choice in this setting but in our
17
institution it tends to be MRI.
18
[Slide]
19
You have seen some beautiful examples of
20
that so I won't belabor this. This
is an example
21 of
an MRI image of coarctation. The way we
use
22
that in interventional is we obviously can get very
23
detailed anatomic definition of how big the vessel
24 is,
how long the stenosis is, and what tools we are
25 going
to need during the procedure to then address
204
1
that.
2
[Slide]
3
This is just an angiogram of a coarctation
4
that we would then bring to the cath lab, evaluate
5
prior with angiography with a nonionic contrast and
6
then repair with a stent implantation--I apologize,
7 I
gave you two pre's and one post. It
looked
8
great, trust me!
9
[Laughter]
10
[Slide]
11
There are limitations currently with the
12 use
of some of these additional modalities and the
13
tools we currently have in intervention.
This is
14 an
example of an MRI after we implanted a stent.
15
Right now, currently available stents are all
16
stainless steel based.
17
[Slide]
18
This is the image artifact you get on
19
implantation of a stainless steel image in an MRI.
20 So,
we have this beautiful arch. This is where
the
21
stent is and we see nothing in and around the area
22
because of artifact. So, there
still is a
23
disconnect. All our tools are
really based in
24
fluoroscopy angiography at this point wo we do need
25
some work in that area certainly.
205
1
[Slide]
2
Nuclear medicine perfusion scan,
3
particularly as it relates to lung perfusion, is an
4
adjunct modality we use quite a bit with respect to
5
interventional treatment, particularly as it
6
evaluates branch pulmonary artery stenosis in a
7
large number of patients who have had surgical
8
repair.
9
[Slide]
10
This is just an example. This is
an
11
infant with a complex congenital heart lesion
12
called pulmonary atresia, VSD, and these patients
13 are
born with no central or true pulmonary
14
arteries. Their arteries come off
abnormal blood
15
vessels arising from the aorta, which you can see
16
here. The surgeon can do a
remarkable job of
17
recreating lung arteries by sewing them together
18 and
bringing them back together but, in fact, these
19
children are left, as I think Tal Geva mentioned in
20 his
presentation, with significant abnormalities to
21
their blood vessels afterwards.
They do quite well
22 and
yet have very abnormal blood vessels.
So, we
23
need some method of assessing how abnormal those
24
different areas of the lung are and nuclear
25
medicine is quite effectiveness at looking at those
206
1
areas where there is too much flow and areas where
2
there is too little flow so when we take that
3
patient to the cath lab we can address our
4
attention to those vessels that most need it.
5
[Slide]
6
This is just an example of a patient who
7 has
had this type of repair. You can see in
this
8
right lower pulmonary artery that there is quite a
9 bit
of narrowing, as well as the right middle
10
pulmonary artery. That patient
had limited flow to
11
those areas. So, we can bring
them to the cath lab
12 and
can use balloons to work on those arteries and
13
afterwards assess with angiography to show that we
14
have had quite an effect on those areas.
Then we
15
follow-up with additional pulmonary flow scans,
16
nuclear medicine scans, to look at the effect and
17 to
follow those patients long-term.
18
[Slide]
19
As I have hinted at, there are significant
20
limitations to angiography and radiography, the
21
most significant of which is anatomic soft tissue
22 detail. In addition, as has been mentioned for CT,
23
there is radiation exposure which is quite dramatic
24 in
these patients. Then, this is a very
expensive
25
technique and non-portable so that makes quite a
207
1 bit
of limitations, particularly with application
2
worldwide in small centers.
3
[Slide]
4
To just give you a glimpse of what the
5
future of interventional may hold, it is going to
6 be
directly related to what you are talking about
7
today and that is the use of additional imaging
8
modalities and the development of better imaging
9
drugs. Certainly, MRI/MRA is the
area that has the
10
most activity and interest in terms of use for
11
interventional cath. CT is a
definite possibility.
12 Not
much work has been done yet. Then, 3D
echo, if
13
that modality continues to develop, may have some
14
application.
15
[Slide]
16
Let me just talk for a minute about what
17 has
been done in the MRI area. That is the
one
18
that I am the most familiar with and which has had
19 the
most activity. Obviously, MRI is an
excellent
20
diagnostic and imaging tool and over the last
21
number of years the magnets have gotten small
22
enough that we can now get to the patients when the
23
patients are in the magnets. In
addition, the
24
speed at which the images can be obtained has
25 improved
enough so that we can actually get
208
1
real-time imaging of the heart as it beats. So,
2
that has opened the door for us to now consider
3
using the cath imaging modality as a direction for
4
interventional techniques.
5
[Slide]
6
In fact, there are a number of combined
7 MRI
fluoroscopy interventional labs that have been
8 put
in place, a few in the United States and a
9
number around the world, that really consist of an
10
angiography suite and an MRI suite that are
11
connected by an interconnecting table that can
12
slide a patient from one to the other, with a set
13 of
doors that slide in between that allow isolation
14 of
the magnet from all the metal in the fluoroscopy
15
area.
16
[Slide]
17
This is just a picture of the suite we
18
have at UCSF. This is a 1.5 tesla
short-bore
19
magnet and a Phillips C-arm rotating angiography
20
suite. It is separated by these
isolating doors.
21
This table slides between the two so you can work
22 in
one room or the other and move the patient back
23 and
forth.
24
[Slide]
25
This is just an example of moving
the
209
1
patient from the MR scanner back across to the
2
angiography suite.
3
[Slide]
4
This is just showing that with these
5
short-bore magnets you can actually get to the
6
patient, either their head for neck vessel access
7 or
to the other side to their groin for leg access
8 so
that we can do some of these interventional
9
procedures right in the scanner.
In fact, you have
10 an
image monitor there that you can look at in live
11
image and that can be swung all around the room in
12
front of the operator so they can watch what they
13 are
doing while they are moving.
14
[Slide]
15
This is just an example of a
16
catheterization in the MRI scanner.
This is
17
something that we have been working on.
This is a
18
prototype catheter that allows you to detect the
19 tip
of the catheter very obviously. You can
see
20 the
soft tissue images nicely as the catheter moves
21 up
and around the aortic arch towards the left
22
ventricle. So, this is opening up
the potential
23 for
use of this modality for catheterization and,
24 in
fact, last year there was nice work done by a
25
group in Germany, developing a device specific for
210
1 the
atrial septum that can be used in the MRI
2
scanner.
3
[Slide]
4
We have done some work at our institution
5
that shows that even with currently approved
6
devices they can be used. This is
an animal model
7
closing an ASD, which is seen right here. This is
8 an
Amplatzer device being deployed, the left atrial
9
side of it being deployed in the left atrium. This
10 is
live MR fluoro. Here is the right atrial
side
11 of
the device being deployed and then the device
12
being released. Obviously, the
potential advantage
13
here is that instead of just seeing the
14
intravascular space we can see soft tissue around
15 as
well and help guide our interventions.
16
[Slide]
17
This is just showing what you can do in
18
terms of a soft tissue look at a variety of
19
different types of stents that are currently
20
available. This is some work we
did in the
21
pulmonary arteries. You can see
that the image
22
quality can, in fact, get quite good if you can
23
match some of the tools with the imaging modality.
24 You
can see the chain-link fence of the stent
25
sitting in the right ventricular outflow track
211
1
pulmonary artery in this model.
2
[Slide]
3
This is just an example of a stent being
4
deployed in the right ventricular outflow track in
5 an
animal model that really shows us that we can
6 use
these images to guide some of these techniques.
7
[Slide]
8
Just to sort of summarize for you, I would
9 say
that the current radiography or angiography
10
techniques that we use and the agents that we use
11
really are quite safe and useful for pediatric
12
interventional catheterization, and it is not clear
13 to
me that there needs to be a whole lot of study
14 in
that area.
15
But advances in interventional cardiology
16 are
really going to come from advances in 3D
17
imaging in these other modalities, MRI, CT or
18
3-dimensional echo. In fact, safe
and effective
19
contrast agents will be key to allowing these
20
interventional advances because our image quality
21
will need to increase substantially.
22
[Slide]
23
The challenges for this include faster
24
acquisition time, which we are getting towards and
25
which no doubt will come in the next few years.
212
1 But
the other issue is image resolution. We
really
2
need to be able to define images down to 1-2 mm in
3
size for pediatric work in some of these
4
procedures. Right now, that is
going to depend
5
primarily on improved contrast agents.
6
[Slide]
7
I would just say my view of the future is
8 the
combination of real-time 3D imaging with some
9
improved contrast agents for the use of
10
interventional cath to really bring interventional
11
repair to a new level, both improved accuracy but,
12
more importantly, the ability to repair complex
13
congenital heart disease in the cath lab. Thank
14 you
very much.
15
DR. CHESNEY: Thank you very
much. It has
16
been suggested by our colleagues at the FDA that
17
maybe we need to take a break at this point. I
18
don't know who has shown that they are not totally
19
alert but somebody picked up on it.
20
[Laughter]
21
So, maybe we could take a ten-minute break
22 now
and come back at 2:55 for our next speaker.
23
Thank you.
24
[Brief recess]
25
DR. CHESNEY: Thank you, all. Just a
213
1
business issue, we, as in the proverbial "we," have
2
made a decision not to try to finish tonight. I
3
think for many of us for whom this information is
4
very new, very interesting but, as a result of all
5 the
time and work that has gone into preparing for
6
this meeting, I think that we probably will need
7
time to do a little more thinking and absorbing all
8 the
material that you all have given us. I
9 understand
that all of our consultants are going to
10 be
back here in the morning so we will try to
11
finish on time tonight and reassemble in the
12
morning. That means that we need
to have
13
transportation back to the hotel.
So, I wonder if
14
everybody who would like a ride in a van from here
15 to
the hotel at the end of this session would
16
please raise their hands. Dr.
Santana is going to
17
stay here for the night!
18
Thank you for bearing with us.
Our next
19
speaker is Dr. Craig Sable who is going to speak to
20 us
on contrast enhanced cardiac ultrasound.
21 Contrast Enhanced Cardiac
Ultrasound
22
DR. SABLE: Thank you. I would like to
23
thank the FDA for inviting me to speak.
I am the
24
Director of Echocardiography at Children's National
25
Medical Center.
214
1
[Slide]
2
The topic I have, contrast use in
3
echocardiography, is a little bit of a dichotomy in
4
that by far and away of all the imaging modalities
5 we
are discussing today echocardiography is the
6
most common. About 18 million per
year are
7
performed in the United States.
With that number
8
ever increasing, especially as the machines become
9
more and more portable, probably a conservative
10
estimate, although there are no data, is that about
11 one
million of these are performed in children.
12
It is done in real time. It is
low cost.
13 it
is portable. It is very widely
available.
14
There is almost no discomfort.
There is no
15
radiation. It is primarily used
for cardiac
16
structure and cardiac function, both systolic and
17
diastolic. It gives us
considerable information
18
about hemodynamics. It helps us
with regional wall
19
motion, both at rest and during exercise where the
20
imaging is more difficult.
21
The dichotomy is that even though echo is
22 the
most widely used, if you look at the data that
23 Dr.
Geva presented earlier, probably ten-fold more
24
than all the other modalities combined but there is
25 the
least amount of information on contrast in
215
1
echo, especially in children.
2
[Slide]
3
There are some limitations to
4
echocardiography that contrast has the potential to
5
overcome. Many patients have poor
acoustic windows
6
which may make it difficult to look at structure,
7 the
endocardial border, regional wall motion and
8
Doppler signals. Patients at
particular risk for
9
this include those with pulmonary disease, obesity,
10
chest wall deformity, postoperative patients and
11
after exercise. The consequences
of these
12
suboptimal images include misdiagnosis, low
13
diagnostic confidence, need for additional tests
14 and
higher inter-observer variability.
15
Finally, echo without contrast does not
16
help us very much with coronary perfusion.
17
Probably a conservative estimate is that up to 5
18
percent of all the pediatric patients, probably
19
tens of thousands per year, could benefit from
20
contrast echo.
21
[Slide]
22
Well, what can contrast echo do for us?
23 Why
use it? These agents are intravenously
24
injected and may enhance the echogenicity of blood.
25 The
goal would be to delineate the echocardiogram
216
1 by
opacifying the cavity, enhancing Doppler signals
2 and
allowing us to image perfusion of the
3
myocardium. This would increase
the sensitivity of
4 the
test, heighten the diagnostic confidence,
5
improve the accuracy and reproducibility and
6
enhance clinical utility.
7
[Slide]
8
This is not an uncommon example of an
9 older
patient, trying to see the endocardial
10
border. This is after contrast
echo and the
11
endocardial border can be shown right here. It is
12
much better seen with contrast echo.
I will show
13 you
some more examples as we go through.
14
[Slide]
15
The desired contrast agent properties are
16
that they are non-toxic. They can
be intravenously
17
injectable either as a bolus or continuous
18
infusion. They are stable both
during passage
19 through the heart and the lungs. They remain in
20 the
blood pool or have a well specified tissue
21
distribution. The duration of the
effect will be
22
comparable to the study itself, and they will be
23
very small size.
24
[Slide]
25
To give you some historical perspective,
217
1 the
original contrast agent was agitated saline.
2
Agitated means that we literally put it in a
3 syringe
and we shake it up, mix it up with a little
4 bit
of air. It is very helpful to identify
shunts,
5
particularly atrial septal defect shunts. But the
6
limitations are the bubbles are too big so if you
7
inject it in the right side of the heart and it
8
goes through the lungs you won't see it very well
9 on
the left side of the heart, and the bubbles
10
dissolve very quickly.
11
You can inject directly into the heart
12
with agitated saline or into the coronary arteries
13
but, again, that definitely has some limitations.
14 The
size itself can cause complications and it is
15
invasive and impractical.
16
[Slide]
17
There have been newer generations of
18 contrast
agents that have come out in recent years
19
that have tried to overcome some of these problems
20
with agitated saline. Albunex was
the first agent
21
that came out. It is highly
echogenic on the left
22
side; It is only 2-4 micrometers,
which is about a
23
third of the size of the red blood cell, but it is
24
only effective for about 2 minutes.
25
So, second generation agents use gas
218
1 instead of air, and the two that are most
commonly
2
used and are FDA approved are Optison and Definity.
3
These either have perfluoropropane or carbon or
4
other gases. These act for a
longer time. There
5 are
even third generation agents with newer gases
6 and
different shells that have even more exciting
7
properties that I will touch on as we go through.
8
[Slide]
9
Air is highly soluble but it has low
10
persistence and stability and diffuses rapidly
11
versus some of the gases that are in the agents
12
like Definity and Optison that have higher
13
molecular weight, low solubility and are very
14
persistent and stable.
15
[Slide]
16
This is just a cartoon on the left of
17
Levovist, showing the contrast agent as it kind of
18
adheres to the blood cells, and then an electron
19
micrograph reproduction of Optison in the blood
20
stream next to the red blood cells.
21
[Slide]
22
This is a list from the article that I put
23 in
your handout from 2000. There are newer
lists
24 but
this is just an example. This is in the
latest
25
statement by the American Society of Echo on
219
1
contrast echocardiography listing some of the
2
agents out there.
3
[Slide]
4
Just to kind of summarize, Albunex is FDA
5
approved but not very commonly used.
Optison and
6
Definity--I believe there is one agent out there
7
that is also approved that isn't used very
8
frequently but Optison and Definity are the two FDA
9
approved contrast agents that are most commonly
10
used. Then, Levovist and Echovist
are approved in
11
Europe. There are several other
contrast agents
12
that are likely to be approved in the near future.
13
[Slide]
14
For us to understand how contrast agents
15 are
useful in ultrasound we need to know a little
16 bit
about how the ultrasound and contrast interact
17
because that will become very important in
18
understanding how these agents are used and how the
19
machine is used with the agents.
The bubbles
20
themselves, in addition to reflecting the
21
ultrasound, are actually resonating with the
22
frequency of the ultrasound beam.
23
Just to review, with ultrasound we are
24
sending ultrasound waves at a frequency much higher
25
than human sound, anywhere from 1-7 MHz, even up to
220
1 12
MHz. The ultrasound bubbles actually
resonate
2 at
the same frequency as the ultrasound beam.
The
3
key, as someone mentioned earlier, that we need to
4
have our echo machines do is differentiate the echo
5
from the contrast from the ordinary tissue.
6
[Slide]
7
But it is not quite that simple, and to
8
understand this a little bit further there is the
9
principle called the mechanical index, which is
10
essentially a measure of the energy at which we
11
expose the tissue and ultrasound bubbles when we
12 are
doing an echo and it is displayed on the
13
ultrasound machine. All the
ranges I am going to
14
display are proven to be very safe.
At less than
15 0.1
mechanical index the bubbles oscillate, just as
16 I
told you. At higher power they actually
17
oscillate at several different frequencies, and
18
higher still they actually break.
19
[Slide]
20
This is just a cartoon kind of showing
21
that at low power they resonate in a linear
22
pattern. At higher power they
resonate in a
23 harmonic
manner, which I will talk about in a
24
second. This is the way that most
echo machines
25
function. Then at a higher power
still the bubbles
221
1
will disrupt, which is very important for perfusion
2
imaging.
3
[Slide]
4
Just to briefly review the principle of
5
harmonic imaging, normally bubbles resonate at the
6
frequency of ultrasound but at higher MI bubbles
7
will have multiple different frequencies, the
8
loudest being twice the normal frequency, or the
9
second harmonic. The resolution
of ultrasound is
10
higher at higher frequencies. So,
the fact that
11
these bubbles can resonate at twice the normal
12
frequency means we can significantly improve the
13
resolution and that is a huge advantage of contrast
14
echo.
15
However, there is a caveat.
Tissue also
16 has
second harmonic imaging and the good news is
17 that, just in a happenstance way, contrast
echo
18
allows us to have this new way to image tissue with
19
much better image quality. The
bad news is that
20
turning on the second harmonics of the echo machine
21
doesn't necessarily completely distinguish the
22
tissue from the ultrasound bubbles.
But just keep
23 in
mind that for purposes of this talk we are
24
generally using second harmonic imaging to image
25
contrast.
222
1
[Slide]
2
This just shows the first harmonic and
3
second harmonic peak.
4
[Slide]
5
This is just an example. The time
at
6
which you can image is much greater using second
7
harmonic imaging. This is an
image without
8
anything. This is harmonic
imaging without
9
contrast, and the best image of all is harmonic
10
imaging with contrast.
11
[Slide]
12
There are even some more higher grade
13
technologies that I won't get into in detail, but
14
they allow the bubbles to actually break and help
15 us
with perfusion. So, harmonic imaging is
best
16 for
tissue opacification and breaking the bubbles
17 is
best for looking at perfusion.
18
[Slide]
19
With left ventricular opacification, as I
20
said before, it helps with poor windows; left
21
ventricular systolic function; stroke volume
22
calculations; space occupying masses such as clots
23 and
tumors; and regional wall motion both at rest
24 and
stress, both with exercise and drugs.
25
[Slide]
223
1
This is just an example of a four-chamber
2 and
two-chamber view with and without enhancement.
3
With using contrast agents, a multi-center study,
4
published in The American Journal of Cardiology,
5
showed that 91 percent of patients got adequate
6
enhancement using contrast.
7
[Slide]
8
This is just another example of an
9
unenhanced image, and then with Definity the
10
endocardial border is much better defined. You can
11 see
a little bit of hypertrophy here. If
this were
12 a
clot or something like that, again, that would be
13
much better defined with contrast.
14
[Slide]
15
Another study done in AJC using Definity
16
looking at patients that had terrible images or
17
non-diagnostic exams, the percent of patients with
18
diagnostic exams was increased from zero percent to
19 70
percent with Definity.
20
[Slide]
21
This is opacification, looking at
22
different segments during a stress echo where it is
23
very critical to evaluate wall motion.
This is
24
without contrast and this is with contrast. Again,
25 the
segmental wall is much better seen in four
224
1
views with contrast than without.
Both at rest and
2
exercise contrast echo improves regional wall
3
motion detection and left ventricular
4
opacification.
5
[Slide]
6
It can also help looking at Doppler
7
signals. We use Doppler signals
for a wide variety
8 of
things in echo. One of the things we use
it for
9 is
pulmonary vein Doppler to help with diastolic
10
function. This is just an example
of a pre- and
11 post-injection
of Levovist with contrast echo.
12
Again, the signals are much more clear with the
13
contrast.
14
[Slide]
15
Perfusion, as we have alluded to with MRI
16 and
other modalities, is really what we are moving
17
towards in the field of imaging.
We look for
18
structure. We look for
function. But if we could
19
really get a handle on coronary perfusion the field
20
would be moved tremendously forward.
What we want
21 to
try to do is identify ischemic tissue and viable
22
pericardium and find areas that are at risk. So,
23 we
want to try to get ways to image the
24
microvasculature in a non-invasive way.
25
[Slide]
225
1
This is just an example of a normal
2
perfusion scan. We are actually
breaking the
3
bubbles. This is just the
myocardium here using
4
power imaging. This dark area
here is an area of
5
apical infarction. This is a
similar patient with
6
apical infarction both on contrast echo and on
7
SPECT nuclear scan, and SPECT scans are still the
8
gold standard but there are several adult studies
9
comparing perfusion using contrast echo versus
10
SPECT with very good results.
11
[Slide]
12
This is an image using pulse inversion.
13
First you will see the endocardial border light up
14 and
then after a little bit of time you can
15
actually see the myocardium light up, very similar
16 to
one of the images that was shown in the MRI talk
17
looking at the perfusion of the myocardium
18
itself--incredible potential.
19
[Slide]
20
There are additional applications.
One
21
that we are using in adults is treatment of
22
hypertrophic cardiomyopathy by injecting alcohol
23
direct into the coronary artery of the hypertrophic
24
myocardium. When you are doing
that procedure, you
25
definitely want to make sure that you are injecting
226
1 in
the right part of the heart and contrast echo is
2
used to identify that.
3
The really exciting thing is that these
4
contrast bubbles--and, hopefully in the next five
5 to
ten years we will be back here talking about
6
them for that particular use--can be the magic
7
bullet for treating things like clots, injecting
8
genes in certain parts of the heart or other parts
9 of
the body, doing some interventional things like
10
opening up ASDs or dilating valves and even
11
treating cancer. As the field of
pediatric
12
cardiology has moved from diagnostic caths to
13
interventional caths with less diagnostic caths and
14
more diagnostic echo, hopefully, in the future we
15 are
actually going to move towards therapeutic
16
echo, and using some of these contrast agents of
17 the
future could definitely get us there.
18
[Slide]
19
This is just an example of using contrast
20 to
identify the area of a ventricular septum that
21 has
hypertrophied and injecting ethanol to ablate
22
that area and treat hypertrophic cardiomyopathy.
23
[Slide]
24
Well, we are really here to talk about
25
safety. Hopefully, I have given
you an idea of
227
1
what contrast echo can do in adults.
The safety
2 has
been established and there are two ways to
3
think of contrast. It is a drug
and, as a drug, it
4 has
been very well established, using very
5
stringent criteria, that there are very minimal
6
side effects. I will show you a
few examples.
7
There is only one study in pediatrics.
8
But then there is the ultrasound-contrast
9
interaction where there are some biological effects
10 of
the sound waves and the bubbles working
11
together. In terms of a drug,
there have been very
12 few
side effects.
13
[Slide]
14
There is only one study I could find that
15 had
any substantial amount of side effects.
This
16 had
Optison being used at 100 times the current
17
recommended dose and only 70 percent of patients
18 had
side effects, only one of whom needed to be
19
treated. Those included headache,
nausea,
20
vomiting, flushing and dizziness.
Again, this is
21 at
100 times the dose. There were no side
effects
22 in
an interoperative study when it was given in 57
23
patients.
24
[Slide]
25
In terms of the ultrasound-contrast
228
1 interaction,
at exposure levels well above clinical
2 use
and clinical power of ultrasound, there could
3 be
bioeffects in the tissue itself. You
could
4
actually heat up the blood to the point where there
5
could be potential problems, but using ultrasound
6
levels identical to normal exams that is unlikely
7 to
happen and has been shown in repeated animal
8
studies to have no bioeffects even though
9
ultrasound is disrupting the bubbles and this could
10
theoretically lead to cavitation at very high
11
temperatures. But it is something
that has not
12
been shown to happen in animals, but as we go
13
forward it may be the basis for some studies.
14
[Slide]
15
As I said, there is no evidence of
16
bioeffects at conventional imaging with normal
17
hematocrits, a mechanical index at 1.9 which is
18
higher than we ever use, and agent concentration
19
less than 0.2 percent which, again, is much higher
20
than we ever use. At very high
concentrations,
21
high ultrasound energy and very low hematocrits
22
there have been reports in animal models of
23
hemolysis, platelet lysis and pulmonary hemorrhage.
24
[Slide]
25
There are alternatives to contrast echo,
229
1
including transesophageal echo, MRI, nuclear
2
studies and angiography, but contrast echo has the
3
advantage that it is not invasive; it can be widely
4
available; and it can be done at the bedside.
5
[Slide]
6
So, based on all of this data, the
7
American Society of Echo recommended in their 2000
8
statement that physician and sonographer competence
9 is
critical, but any echo, either standard or
10
stress, that has suboptimal views, meaning that you
11
can't see 2/6 apical segments, and/or there is
12
inadequate Doppler, contrast echo could be
13
considered to be indicated.
However, your lab has
14 to
have the ability to have the highest quality
15
standard equipment before you move to contrast echo
16 for
left ventricular opacification. For
myocardial
17
perfusion it is still considered investigational.
18
[Slide]
19
To summarize the adult data before I get
20
into the pediatric data, in the past we have used
21 it
to identify intracardiac structures and shunts.
22
Presently, we can do intracoronary myocardial
23 contrast.
We can enhance endocardial borders and
24 do
Doppler. In the near future--myocardial
25
perfusion, stress perfusion and viability, and in
230
1 the
far future drug gene delivery and clot lysis.
2
[Slide]
3
Pediatrics is a little bit different.
We
4
look at structure more than function.
We have less
5
experience with wall motion assessment.
We do have
6
better windows because the heart is closer to the
7
chest, and we have higher frequency transducers.
8
There aren't very many large multi-center trials,
9 and
we do use drugs in an off-label manner quite a
10
bit.
11
[Slide]
12
But there are many potential
uses for
13
contrast. We use it for
shunts. We have about
14
three-quarter of a million adults with congenital
15
heart disease in this country.
That number is
16
going way up. Many of them have
complex disease,
17 or
are in the postop setting or have single
18
ventricles. There is a large
number of pediatric
19
patients with coronary disease, maybe not typical
20
atherosclerosis but we have a huge population of
21
children and adults with Kawasaki disease. We have
22 a
large transplant population. And, some
of the
23
diseases, such as transposition of the great
24
arteries, are at risk for coronary artery disease.
25
And, there is a growing field of stress echo in
231
1
kids.
2
Some limitations--putting an IV in a
3
little baby is kind of a big deal but in an older
4
child it really isn't. There is
very little data
5 and
it is a little harder for us to get the volume
6
needed to have competence.
Coronary artery disease
7 is
somewhat uncommon and in many of our patients
8
image quality is satisfactory so getting an
9
appropriate volume to have competence is somewhat
10 of
a limitation. And, contrast agents are
11
relatively expensive.
12
[Slide]
13
There is one study in pediatrics.
When I
14
first thought about this talk I thought I would
15
just show you this study and let you all think
16
about it. But, clearly, this is
an issue because
17 we
have a long way to go. Dr. Kimball, in
18
Cincinnati, published this study in 2003 looking at
19
patients referred for stress echo, Kawasaki
20 disease, transplant postoperative patients and
21
atypical chest pain.
22
[Slide]
23
Here is the stress echo protocol using
24
dobutamine or bicycle. They used
0.1 mL to 0.2 mL
25
kind of empirically for the contrast protocol of
232
1
Optison, using 25 mg as a cutoff.
The adult dose
2 is
0.5 mL.
3
[Slide]
4
They followed by a saline flush and
5
monitored saturation heart rate and blood pressure
6 for
45 minutes after the injection. They got
7
standard parasternal and apical views using
8
harmonic imaging with a mechanical index of 0.4.
9
[Slide]
10
They tried to look at 16 myocardial
11
segments. Six are seen in two
views. There was a
12
total of 22 segments that were graded on a scale
13
from 0-3 by one blinded pediatric cardiologist,
14
both with and without contrast.
15
[Slide]
16
They looked at 22 children over a 14-month
17
period. Their diagnoses are shown
here, 19 were
18
dobutamine studies and 3 were exercise.
The
19
smallest patient was 8 months old.
20
[Slide]
21
They had no hemodynamic changes or
22
complaints. Image quality was
improved in 21/22
23
studies, especially in the apical segments. When
24
talking to Dr. Kimball recently, he said that they
25
have since done about 20 more patients, again, with
233
1
zero side effects reported.
2
[Slide]
3
In summary, contrast echo has been proved
4 to
be safe in adult patients. It has been
endorsed
5 by
the American Society of Echo for left
6
ventricular opacification studies at rest and
7
exercise. There are important
additional uses for
8
contrast that are likely to be developed and
9
approved in the near future, including myocardial
10
perfusion and tissue specific delivery.
11
[Slide]
12
In pediatrics we are a little bit behind
13 the
adults, but echo is the most commonly used
14
diagnostic modality in children with cardiovascular
15 disease and there are important potential
uses for
16
contrast echo, as I said earlier, probably tens of
17
thousands of patients per year.
Based on Dr.
18
Kimball's study we can begin to conclude, from his
19
study at least, that contrast echo can safely be
20
performed in children and it improves the quality
21 of
stress echo. But there are obviously
limited
22
data. We are only looking at 22
patients published
23
evaluating the use of contrast echo in children.
24 [Slide]
25
My recommendations would be that we need,
234
1 as
a pediatric cardiology community with the
2
support of the FDA, to develop dosing for
3 pediatrics,
assess safety and establish specific
4
indications. Hopefully, we can
get together with
5 the
American Society of Echocardiography and
6
develop specific guidelines that will serve as a
7
resource for additional pediatric cardiologists to
8 use
contrast echo.
9
Finally, I would like to acknowledge Dr.
10
Weissman, Dr. Rychik, Dr. Kimball and the American
11
Society of Echo for contributing to some of the
12
content of this talk. Thank you.
13
DR. CHESNEY: Thank you very
much. Our
14
last speaker for the afternoon session is Dr.
15
Dilsizian who is going to speak to us on
16
radiopharmaceuticals in nuclear cardiac imaging.
17
Radiopharmaceuticals in Nuclear Cardiac Imaging
18
DR. DILSIZIAN: Thank you very
much. I
19
appreciate the invitation to be part of this panel.
20 My
background is that I am an adult cardiologist
21 who
is also double-boarded in nuclear medicine.
I
22 have
spent the last 13 years at the NIH doing work
23 in
hypertrophic cardiomyopathy involving also the
24
pediatric population. Currently,
I am the Director
25 of
the Cardiovascular Nuclear Medicine at the
235
1
University of Maryland. I have
been there now for
2 a
couple of years.
3
[Slide]
4
We have heard about a lot of technologies
5 and
if I were sitting in the audience I would say
6 it
seems like everybody is showing function,
7
perfusion and all this nice stuff and you say why
8
would I even want to use nuclear?
Just the name
9
itself is scary and why would we even bother with
10
this?
11
So, what I would like to do is I would
12
like to say to my colleagues that as far as anatomy
13 is
concerned, echo, CT, MRI--it is great.
Any time
14 you
think about nuclear you have to think about the
15
physiology and metabolism.
Okay? So for anatomy,
16
nuclear has no business. Whenever
we think about
17 the
physiology or metabolism we should be thinking
18
about nuclear medicine.
19
Why? It is because unlike some of
the
20
flow tracers that they have mentioned so far, the
21
beauty of nuclear cardiology--which, although the
22
field was back in the 1940s the real perfusion
23
imaging began in mid-1970s--because of the fact
24
that it has been used for the last three decades in
25 the
adult population to detect coronary artery
236
1
disease, it has passed the test of time and we
2
respect that field.
3
Now, there is something about perfusion
4 imaging
in nuclear that has to be important and
5
unique. What is it about it? It is because when
6 we
inject a tracer like thallium-201, technetium,
7
maybe tetrofosmin, rubidium-82 with PET and N-13
8
ammonia with PET we are not only looking at flow,
9 we
are looking at retention of that radiotracer in
10 the
cell. It is a very, very unique
characteristic
11 of
nuclear medicine. The isotope that you
inject
12 and
that is attached to a radio ligand is being
13 actually
intercepted and retained in the cell. No
14
other technology can do that.
With tetrofosmin
15
they will enter the mitochondria and, therefore,
16
they tell you about the intactness of the
17
mitochondria where ATP is formed and no other
18
technology can do that.
19
SPECT imaging stands for single photon
20
emission computer tomography, while PET is positron
21
emission tomography. The only
difference between
22
these two terms is the P and the SP, which means
23
that what differentiates these two technologies is
24 the
radiotracer. Radiotracers used with PET
are
25
positron emission radiotracers.
The tracers used
237
1 with SPECT are single photon emission
radiotracers.
2 I
don't want to get into that detail. All
that you
3
need to know is why do we need to move into the PET
4
technology which has also been around for a couple
5 of
decades. It is because as we move from
6
thallium-201 to technetium perfusion to PET what we
7 are
trying to do is we are trying to get the same
8
biological/physiological behavior, yet reduce
9
radiation exposure.
10
So, this is a very important concept.
11
Thallium-201 is an elegant biological tracer, a
12
potassium analog injected as a salt.
What is the
13
problem? Physical properties, low
energy, high
14
physical half-time of 72 hours, long physical
15
half-time. Therefore, we are
limited by the
16
dosimetry, 5 mCi is all we can get.
That limits
17 our
quality of images and diagnostic capabilities,
18
especially in large patients. It
may not apply to
19
kids but in kids we are not talking about large
20
size, we are talking about the long physical
21
half-life and, therefore, we want to limit the body
22
distribution, limit exposure to the kids.
23
Moving to technetium-labeled perfusion
24
tracers, its physical properties are 6-hour
25
half-life, 140 K energy. Again,
why do I need to
238
1
know that? It is because the
energy is much more
2
appropriate for the current gamma cameras that are
3
available. Tomorrow, if we change
the sodium
4
iodide crystal we may choose another radiolabel but
5 the
ligand remains the same. So, short
half-life
6
and, therefore, we can give 25-30 mCi.
Suddenly,
7 we
have been able to get similar information, if
8 you
will, but getting a higher count and that
9
allows us to not only get myocardial perfusion but
10
also function with the same setting--very important
11
concept.
Where does PET come in?
12
Well, let's push the envelope further.
Now we are
13
going to use radiotracers that, because of the
14
energy characteristics, you are going to have much
15
better, higher count rates. In
addition to that,
16 you
can have attenuation correction. It may
not be
17
important for kids again. But
more importantly,
18
what is important is that rubidium-82 has a very
19
short half-life, 32 seconds.
Ammonia N-13,
20
ten-minute half-lie. So, now we
are talking about
21 not
only physical properties that are shorter and
22
shorter but biological properties.
Rubidium goes
23 in
and goes out 32 seconds later.
Therefore, the
24
radiation exposure to the kids will be limited and
25 now
we can concentrate on the physiology.
That is
239
1
what is exciting about nuclear medicine.
2
In the era of genomics and proteomics you
3
understand that we are really in the field that is
4
becoming the molecular imagers.
So, now let me go
5
into clinical applications based on this
6
background.
7
[Slide]
8
The main applications will be congenital
9
heart disease, diagnosing coronary circulation
10
anomalies. We have heard all of
that and I don't
11
want to show you any images; Kawasaki disease,
12
hypertrophic cardiomyopathy or monitoring
13
chemotherapy which can be done with
14
echocardiography or MRI, but in some patients you
15
actually want to know reproducibility with very
16
accurate numbers.
17
[Slide]
18
I want to pick specifically hypertrophic
19
cardiomyopathy. That hasn't been
discussed much
20 and
I want to tell you why. One is because I
have
21
done a lot of research on this but the other thing
22 is
that it exemplifies where perfusion imaging with
23
nuclear has an advantage over other technologies.
24
We have learned in the last several years
25
that with hypertrophic cardiomyopathy, which is
240
1
really thickening of the heart and it can be
2
asymmetric septal hypertrophy or concentric
3
hypertrophy, there are some genetic diverse
4
features. When I was in medical
school I was
5
taught that the prevalence of hypertrophic
6
cardiomyopathy in the general population was 3
7
percent. That was my education
and that is based
8 on what?
That wasn't based on genetic studies.
9
Those were just learned recently.
That was based
10 on
echocardiographic or abnormal EKG findings.
The
11
prevalence, therefore, actually may be higher.
12
And, I am going to show you that now that we are
13
getting into genetic identification we can identify
14
that there is a higher prevalence perhaps in the
15
general population than 3 percent.
16
What is important here in kids is that
17
sudden death, unfortunately, occurs commonly in
18
young patients. What do I mean by
that? If you
19
diagnose hypertrophic cardiomyopathy in a child
20
between ages 1 and 14, 50 percent of those kids
21
after diagnosis will die in 9 years.
That is
22
scary. Okay? Therefore, everything that I am
23
going to say now about radiation exposure you have
24 to
put in perspective of what we are talking about
25 and
what we are identifying because I think at the
241
1 end
of this we have to say what is the added
2
potential fatal cancer in these kids versus their
3
survival. Again, this is one
subset of patients
4
that exemplifies how we have to think about nuclear
5
imaging.
6
[Slide]
7
I mentioned to you that there was recently
8 an
elegant publication in The New England Journal
9
that told you the prevalence of where some of the
10
genetic abnormalities can be in patient
11
populations. Now you can screen
them, especially
12 if
there are increased sudden deaths in those
13
patients.
14
This is one pathologic slide from a young
15
patient who died suddenly with cardiac arrest.
16
This is the septum and you can see all of this red
17
stuff is scarring. You see these
small vessels
18
here. They are thickened. This is a young patient
19
that has an unusual interstitial structure and
20
coronary arteries that causes these kids to die.
21 You
have heard about these athletes playing
22
basketball and dying suddenly.
This is the same
23
patient population.
24
How do I identify these? In the
25
traditional way we say, well, you know, I will do
242
1 CT
angiography. Guess what, the coronaries
are
2
normal. So, CT angiography is not
going to give
3 you
the information. Now, what is it that I
am
4
going to do? What I would like to
do is identify
5
ischemia. Right? Ischemia is a supply-demand
6
mismatch. Even though the vessels
may be normal,
7 the
demand component may be abnormal because it is
8 a
thickened heart.
9
Now, one of the strengths of
nuclear
10
medicine is that we are going to put patients on
11 the
treadmill. All of the other fun stuff we
have
12
heard is pharmacologic stress. It
is not what
13
patients actually do. We are
looking if someone is
14
running on the basketball court--running--is he
15
going to have arrhythmias, is he going to die?
16
That is what I want to know and, therefore, I am
17
going to reproduce that on the treadmill and inject
18 a
nice radiotracer which will tell me if that
19
patient is ischemic or not.
20
[Slide]
21
We did this study at the NIH and here is a
22
very nice example. This is a
young kid, 8 years
23
old. Obviously, the dark area
would be lack of
24
blood flow. This patient has no
coronary disease.
25 We
are talking about ischemia based on a
243
1
supply-demand mismatch that is completely
2
reversible. So you say, well, why
is this
3
important? Why do I need to know
that? Is there
4 any
relationship between ischemia and sudden
5
cardiac death?
6
[Slide]
7
Again, what I want to show you is that
8
even though that is done with thallium, you can get
9 the
same information with Sestamibi or tetrofosmin.
10
Again, it is flow tracers. If the
body
11
distribution is such that the kids are getting less
12
radiation exposure, obviously you will be moving in
13
this direction and perhaps PET in some direction.
14 I
just want you to have that in mind, that we are
15 not
just stuck in the 1970s. We could
actually be
16 in
the 21st century as the technology moves with
17 the
radiotracers as long as we are getting the
18
signal that we need for a patient.
19
[Slide]
20
Again, this is patient before and
after
21
treatment with verapamil. You can
see that the
22
extent of ischemia is actually better, just medical
23
therapy. Therefore, now I can
follow the patient
24 and
say that by treatment with a beta-blocker and
25
verapamil am I really impacting ischemia and am I
244
1
going to impact sudden cardiac death?
2
[Slide]
3
Again some pathological--these are
4
thickened arterial walls.
5
[Slide]
6
This is the data that I want to share with
7 you
which I published in 1993. So, 23
patients
8
presenting to NIH--these are kids.
They presented
9
either with symptoms of cardiac arrest or syncope
10 and
they obviously survived a syncope episode, or
11 had
a very strong family history of cardiac arrest.
12 So,
now these patients were being evaluated with EP
13
studies looking at arrhythmogenicity and you can
14 see
that by doing EP studies, inducible VT was only
15 27
percent of these cardiac arrest or syncope kids,
16 and
none in those who had family history of cardiac
17
disease.
18
On the other hand, the thallium SPECT
19
study showed all of these guys who had syncope or
20
cardiac arrest actually had ischemia, and 3/8 with
21 the
family history also had ischemia. Now
you
22
would say, well, how do I know this is not--you
23
know, is it too sensitive; it may not be specific?
24 On
the other hand, you are seeing more kids than
25 you
would. They didn't have any symptoms; they
245
1
didn't arrest. And, the follow-up
is very
2
interesting. All of these kids
obviously had AICD
3
placed and were treated with verapamil and beta
4
blockers. You treat them
medically and you also
5
have a backup. You know, these
are kids. They may
6 not
take their medication. Four out of the
15
7
patients with cardiac arrest had further episodes
8 on
anti-ischemic therapy. Three of the 4 events
9
were temporally related to discontinuation of the
10
medication. The kids didn't take
it.
11
How do we know the patient was going to
12
have an arrest? AICD fired which
could capture it.
13 You
know these three patients here, this is
14
one-year follow-up. One of the
kids was playing
15
basketball and had sudden cardiac arrest. So, not
16
only were we right, we actually predicted it.
17
[Slide]
18
So, I want us to think about radiotracers
19 and
what decision we are going to make regarding
20
research or clinical indication vis-a-vis
21
risk/benefit of radiation. Coming
to the bread and
22
butter of our meeting here, how do I look at
23
radiotracers and how do I decide?
How do I
24
translate an adult dose to a pediatric dose?
25
What did we do? What we did was
simple.
246
1 In
the 1980s we just dosed the thallium based on
2 the
kid's weight. That is all we did. So, that is
3 one
way to do it. The other way is to do it
on
4
body surface area. Right?
5
Well, one interesting approach would be
6 why
don't we just look at the relative dose based
7 on
radiation exposure? That is, can we take
a
8
millicurie administered to a child and decide that
9
dose based on the same absorbed radiation of 1 mCi
10
administered in adults, that is, the radiation
11
exposure translated into millicuries rather than
12
some body weight or body surface area?
13
[Slide]
14
Let me emphasize two points. One
is what
15 I
would like to do is whatever patient population I
16 am
studying. As you know, no kid is going
to
17
undergo nuclear study unless there is a real
18
diagnostic dilemma or question.
Right? So, the
19
last thing I want to do is inject the radiotracer
20 in
a kid and get non-diagnostic, poor quality
21
images because I didn't give enough dose. So, I
22
have wasted a dose. I don't have
any information
23 or,
worse yet, I don't have the right information
24
because the images were of poor quality.
Okay?
25
That is critical.
247
1
The next question is everything is
2
risk/benefit, not just imaging.
Forget about
3
nuclear, everything we talked about, everything is
4
risk/benefit ratio. That is part
of medical
5
decision-making. So, hopefully,
today and tomorrow
6 we
are going to have to decide what is it that we
7 are
talking about. I mean, obviously we
should not
8 be
studying kids unless they are going to be
9
benefiting from that technology.
Therefore, we
10
have to put into perspective how much risk are we
11
willing to take based on that technique versus the
12
benefit.
13
[Slide]
14
What is different about
15
radiopharmaceuticals versus X-rays or CT? The
16
difference is that when you inject a
17
radiopharmaceutical it is not a total body
18
exposure; it is a non-homogeneous exposure because
19
these are targeted agents.
Hopefully, we are
20
targeting the liver; we are targeting the heart.
21
That is the goal. If we just went
equally
22
everywhere, then we would not be doing the right
23
thing. So, we are creating
radiotracers to target
24
specific organs to do the right thing.
If that is
25
what we are doing, therefore, you understand that
248
1 it
is not one number. It is an uneven
distribution
2 and
each tracer has its own distribution.
3
[Slide]
4
How do we go about deciding what is
5
exposure? A couple of ways have
been done. As you
6
know, one is to look at the total-body or
7
whole-body dose. That is the
total energy
8
deposited in the body divided by the mass of the
9
body. This approach assumes a
uniform whole-body
10
exposure to radiation. We just
discussed that that
11 is
not the case in nuclear medicine.
12
What is the other approach? Well,
the
13
other approach is a very clever approach I think
14
which is the effective dose or the effective dose
15
equivalent. That is, you say, you
know, here are
16
multiple organs, the top nine or ten most commonly
17
involved in the radiotracer you are using and you
18 use
weighting factors and summing the individual
19
contributions of the single dose organ to come up
20
with a number. When you inject
thallium or
21
rubidium or tetrofosmin or FDG, this is the body
22
exposure and these are the weighying factors. What
23 are
weighting factors?
24
The tissue weighting factors we are going
25 to
use for different organs--very nice in that each
249
1 of
these account for fatal cancers or risk of
2
disease above the normal incidence per unit of
3
ionizing radiation for each organ system. Okay?
4 In
essence, we are taking each organ system and we
5 are
saying what is the potential risk and weighing
6
each and coming up with a number.
I think it seems
7 to
be the most logical thing to do, at least at the
8
present time.
9
[Slide]
10
Now you take that and you sum it up for
11
patients and it is going to give you some
12
tabulation. These are the
weighting factors or the
13
risk that I just mentioned for each of these organ
14
systems. The remaining organs you
can estimate to
15 be
about 0.5.
16
[Slide]
17 Let's take a patient example. I just
18
picked an adult, 10 mCi FDG which is a
19
fluorodeoxyglucose. It is a PET
agent that is
20
commonly used. Now you use the
weighted factors
21 and
the 10 mCi dose. This is the body
distribution
22 and
you come up with an effective dose for the
23
total body, which is unity.
Right? If you add up
24 all
these weighted, it should be 1 and it is 0.68.
25 Now
you take that number and you say if 0.68 is my
250
1
effective dose I need to know what is the incidence
2 of
fatal cancer per rem for that age group.
We
3
have different age groups and we have the risks for
4
each age group.
5
[Slide]
6
These are the nominal probability
7
coefficients for stochastic effects.
This is
8
detriment times 10
-4 per rem in ICLP. Just quickly,
9 I
think it is an important thing to look at.
Here
10 are
the children. Fatal cancer is 8 times
10
-4
per
11
rem; non-fatal cancer, 1.6; severe hereditary
12
effects, 1.6; the total is 11.2.
13
Adults, you can see fatal cancer is 4;
14
total, 5.6. Geriatric--I didn't
think over 50 is
15
geriatric but I am approaching geriatric age, I am
16
afraid--total cancer is 1 and fatal cancer is much
17
lower even compared to the rest of the population.
18
So, what is the point here? For
children,
19 you
can see that the risk is two to three times
20
greater than for adults--cancer or total. Okay?
21 For
individuals over 50 years of age the risk is
22
one-fifth or one-tenth. So, you
know, when we are
23
making decisions about radioisotopes you can say,
24
well, based on these we should, again, optimize the
25
dose and for this patient population, the so-called
251
1
geriatrics, who cares? Because we
are really not
2
having much effect here.
3
[Slide]
4
Here is the calculation. We take that
10
5 mCi
FDG for adult patients and we come up with a
6
number of 0.68 effective dose. We
multiply that by
7 the
fatal cancer rate, which is 4 times 10
-4 and we
8
come up with the probability. The
probability is
9
what? It is 0.27 percent. As was brought up
10
before, we know that the natural incidence for
11
fatal cancer is 25 percent over someone's lifetime.
12 Right?
So, now we say what is the added
13
incremental fatal risk of doing this procedure with
14
FDG? You say, well, it is 0.3.
15
Now, in kids I am going to make it very
16
simple. It is double. Right?
It is 0.06. So,
17 all
we are saying is that this is going to be
18
25.0-something, 24.06, 24.09 but that is the risk
19 for
that procedure. Now we have to make a
20
decision, is this worth the procedure versus the
21
benefit? That is really what we
are discussing
22
here.
23
[Slide]
24
So, NIH--and I am proud to have been there
25 and
I consider the Radiation Safety Committee a
252
1
pretty bright group of individuals--have recently
2
changed their requirements for research.
Perhaps
3 we
should take guidance from this. As you
know,
4
previous guidelines said organ dose--organ, not
5
total-body effective dose--should be 3 rem
6
quarterly or per injection or 5 rem annually. That
7 is
what we have been doing all this time.
They
8
have decided that, you know what, that is too
9
conservative and, therefore, for research subjects
10 at
the NIH the total effective dose now is 5 rem
11 and
that is a significant drop. The
guidelines,
12
again, for pediatrics were to do one-tenth of the
13
dose. Now they are saying
one-tenth of the dose of
14 the
total effective which is much, much better.
15 So, just food for thought, I mean,
we
16
don't have to reinvent the wheel.
We can always
17
kind of look at how NIH came to this conclusion.
18
Perhaps we can take it from there and move forward.
19
Thank you very much.
20 DR. CHESNEY: Thank you.
If you ever
21
wondered what your classmates in medical school who
22
majored in physics as undergraduates did when they
23 got
out of medical school, I think we now know!
24
[Laughter]
25 It is pretty overwhelming to some of
the
253
1
rest of us! I wanted first of all
to say that the
2
handout you received during this talk, reducing
3
radiation risk from computer tomograph for
4
pediatric and small adult patients, came from Dr.
5
Andrew Kang who is with the Center for Devices and
6
Radiological Health. Questions
for the speakers?
7 Dr.
Fink?
8 Q&A for the Speakers
9
DR. FINK: Just to try and put
things in
10
perspective, I read a long time ago that air flight
11 at
35,000 ft gave you an exposure of about 0.01 rem
12 per
hour. Is that still an accurate figure
for air
13
flight?
14
DR. DILSIZIAN: I am not exactly
sure
15
about the number but it is equivalent to about a
16
chest X-ray or so, yes, just going, say, from
17
Boston to California.
18
DR. CHESNEY: Dr. Nelson?
19
DR. NELSON: I guess I would
suggest we
20
reserve radiation risk as its own particular
21
discussion. The question I would
like to ask is
22
throughout the presentations at times I didn't get
23 a
very clear sense about where in the development
24 of
some of these agents you would need to use
25
children, as opposed to where you would be able to
254
1 get
the answers from using adults. For
example, if
2 the
question is the accuracy of imaging at 1-2
3
mm--likely adult vessels that are 1-2 mm or in the
4
breakdown, for example, of a chelation compound
5
what is different about the milieu of the pediatric
6
patients' blood stream as opposed to adult blood
7
stream and clearance. I mean,
what is it that we
8
need to use children for, not in terms of what we
9 can
use it for diagnostically because, obviously,
10
that is very impressive, but what do we need to use
11 them
for in terms of development of new products as
12 far
as testing to get them to a point where they
13 can
be used safely and effectively?
14
I didn't hear that specific question come
15
out. Because in research the
principle is you use
16 the
adult first if you don't need to use the child
17 to
get the information. Once you have it,
then you
18 can
use it clinically. So, I am just curious
both
19 in
terms of imaging capabilities but also
20
metabolism and excretion for compounds such as the
21
chelated gadolinium compounds.
What do you really
22
need to use kids for, for research?
23
DR. CHESNEY: Dr. Geva?
24
DR. GEVA: I am not sure about the
25
radiopharmaceuticals but as far as, certainly, MRI
255
1 and
echocardiography and perhaps CT, as well as in
2 the
catheterization laboratory, I would say that as
3 a
rule extrapolation of data from adults to
4
pediatrics is fraught with potential danger.
5
Just to give you an example, if you are
6
looking at gadolinium dosage and use of contrast
7
agents in MRI, there are considerations that come
8
into play that the adult imaging folks do not have
9 to
contend with, such as small body size, signal to
10
noise ratio, fast heart rates and things of that
11
nature that all impact on what we do and the kind
12 of
data that we get and the type of contrast agents
13
that we have to use.
14
DR. SABLE: I think I can just add
from
15 the
echo perspective. I do agree that the
16
indications are clearly a different issue. I think
17
there is some needed information for safety, not
18
because the adult data isn't very clear but because
19
some of the data needs to be obtained to make kind
20 of
a segue into the pediatric community. I
think
21 not
having any pediatric studies definitely hurts
22 the
perception that these drugs can be used at all
23 in
pediatrics.
24
I think the main role I would think of in
25
studies for contrast echo would be to establish
256
1
minimal dosing guidelines that may be efficacious.
2 One
study just randomly picked a dose between 20-40
3
percent the adult dose but it would be very
4
important to establish specific dosing that would
5 be
acceptable. I think that is probably the
main
6
role that I see.
7
DR. CHESNEY: Dr. Fogel?
8
DR. FOGEL: Yes, I think that with
trying
9 to
extrapolate adult data down to kids, I have to
10
agree with Tal that it is fraught with danger in
11
terms of being able to know exactly what you are
12
dealing with, especially with the small size. When
13 you
inject, for example, gadolinium in a baby it
14
reaches the heart in, like, 2, 3, 4 seconds,
15
whereas in an adolescent or an adult it make take
16 10,
15 seconds before it gets there. There
are all
17
these differences in kids versus adults and, as you
18
alluded to, metabolism. I think
in kids we really
19
have to get a handle that we potentially don't have
20 if
we try to extrapolate it from adults.
So, I
21
would strongly recommend that children be studied.
22
DR. CHESNEY: Dr. Fost?
23
DR. FOST: It is on a different
subject.
24 Are
we still on this one? I wanted to change
the
25
subject. So.
257
1
DR. CHESNEY: Dr. Moore?
2
DR. MOORE: I would just take a
little
3
different approach I guess, and that would be that
4
obviously there are limited resources and that we
5 do
get an awful lot of information from the adult
6
studies that is applicable. But
the specific areas
7
that probably vary quite dramatically, as I think
8
both the previous speakers hinted at, are the
9
smaller children and infants. In
particular, I
10
think that extrapolation is a bit much.
So, if one
11 had
to focus one's resources in the pediatric
12
population for these agents, I would say that the
13
dramatic differences are down in the younger age
14
groups because of the difference in metabolism, the
15
faster heart rates in particular, and the smaller
16
body and image size that you need to detail that
17
makes dramatic differences.
18
DR. CHESNEY: Dr. Nelson?
19
DR. NELSON: Just as a clarifying
question
20 so
I understand, a lot of the need out here is in
21
terms of the ability to capture effective images
22 and
to accomplish what you, indeed, want to get but
23
does that also translate into what I would consider
24
sort of basic metabolism issues?
Do they break
25
down the gadolinium? Do they
chelate and do they
258
1
disassociate any differently? If
you know the GFR
2 of
a neonate, do you really need to know what your
3
clearance of the drug is, etc.?
That is very
4
different from imaging modalities related to heart
5
rate and, you know, when do you start turning on
6 the
scanner, etc. I just want to get clear
about
7
where the differences are. Is it
in the imaging
8
areas or is it in the actual basic metabolism and
9
dosing?
10
DR. CHESNEY: Dr. Fogel?
11
DR. FOGEL: Well, I don't think we
know
12
that. I mean, I don't think we
have the data in
13
terms of metabolism and safety in kids to be able
14 to
extrapolate that from adults. We have
seen a
15
number of presentations today already that showed
16
that the cancer risk and other things are dependent
17 on
the age at which you are actually doing the
18
study. I mean, we don't
know. If we are injecting
19
gadolinium in kids how do we know that when they
20 are
age 40 that those people who had that long-term
21
effect many, many years ago are all of a sudden
22
going to start turning up with cancer of some organ
23
system? The fact that we don't
know this, and that
24 we
don't know what the long-term effects are, and
25 we
don't have as much of a handle on the metabolism
259
1 and
how the body handles gadolinium or other
2
contrast agents make it important that we, one,
3
start doing the testing now; two, we start making a
4 log
of the people we are testing; and, three,
5
hopefully in the future we will be able to get
6
follow-up studies 10, 15, 20 years down the road to
7 be
able to say, yes, we did this kid a service or
8
maybe we did the kid a disservice by doing it. I
9
don't know.
10
DR. CHESNEY: Let's see, Dr.
Sable, Dr.
11
Moore, Dr. Geva and Dr. Siegel.
12
DR. SABLE: I think it is tempting
to
13
start to separate out the device from the agent
14
but, especially with ultrasound, you really can't
15
because the agent reacts to the ultrasound and in
16
children, especially small infants, the
17
transmission is closer to the chest and you are
18
using different frequency transducers so it is
19
almost impossible to separate out the device
20
because the device determines the actual properties
21 of
the agents and they could be different with
22 different types of devices and different
heart
23
rates. So, it is all kind of
intertwined together.
24
DR. CHESNEY: Dr. Siegel?
25
DR. SIEGEL: With ionated or
contrast
260
1
agents there is a lot of experience out there so,
2 to
address your first question, is there a need for
3
doing this in children per se, if we look at the
4
reactions to contrast agents for CT, the reaction
5 types
are different. In adults they are more
6
severe type of reactions; in children they are
7
usually milder or intermediate.
That is important
8 if
you are going to talk to a parent and say we are
9
giving a contrast agent but in children we will
10
expect this, and you can tell them that the
11
reactions will be minimal rather than that there is
12 a
great risk that you are going to have some type
13 of
severe reaction. So, I think based on
that,
14
there is a need to look at children.
15
As far as your second one goes on
16
metabolism, I believe even with the contrast agents
17 for
CT we are still not sure at this point why it
18
happens. We think there are
obviously two types of
19
reactions, either direct drug toxicity or something
20 due
to their idiosyncratic reaction. I am
not sure
21 we
will ever be able to work that out but I think
22 it
is important to know what the risk is in
23
children per se and, based on that previous
24
evidence, it does differ.
25
DR. CHESNEY: Dr. Moore?
261
1
DR. MOORE: I would just make the
argument
2
that I think there are precedents set in other
3
pharmacotherapeutic areas where the metabolism and
4
response in small children is different, and I
5
would be quite concerned, particularly with some of
6 the
new MRI agents and the blood pool agents that
7 are
going to spend a lot of time in the circulation
8 and
are cleared by a variety of mechanisms
9
including hepatic mechanisms, that their response
10 may
be different to the younger age group, and
11
those probably should be looked at a priori as
12
opposed to after the fact.
13
DR. CHESNEY: Thank you. Dr. Geva?
14
DR. GEVA: I would just make a
distinction
15
between the metabolism and the behavior of
16
gadolinium agents. There is
actually a fair body
17 of
knowledge, including pediatrics and including
18
infants. That literature goes
back to the late
19
'80s and early '90s. But what is
unique and hasn't
20
been discussed in great detail is the clinical
21
indications. If there is any
discussion about
22
labeling for use of these contrast agents for
23
specific diagnostic indications, then there are
24
gaps in knowledge. Otherwise,
gadolinium is being
25
used or has been used on a large scale for many
262
1
years for known cardiac indications and information
2 can
be used from that experience.
3
DR. CHESNEY: Dr. Fost, Dr. Gorman
and
4
then Dr. D'Agostino.
5 DR. FOST: This is directed to anybody in
6 the
room, the experts, the FDA or anyone else who
7 can
answer. How close are we to
nanotechnology
8
becoming part of this whole question--devices,
9
coding of devices, drug delivery devices? My
10
understanding is that the EPA for example is still
11
stuck in thinking of a chemical as a chemical. It
12 is
benzine and we have rules about that, and the
13
notion that it might be in a much smaller particle
14
size and different format has not yet penetrated.
15 The
developmental effects of these devices or
16
particles might be, obviously, much more worrisome
17 for
children than for elderly adults. What
does
18
anybody know about that? Is
anybody yet
19
manufacturing things? Is it in
the pipeline? Is
20 it
a year away or ten years away? And, how
will we
21
react to that? That skips the
question of whether
22 you
would want different studies. I think
you
23
would have to have very different studies for
24
developmental studies for children than adults in
25 the
early phases of that. Does anybody know
263
1
anything about that?
2
DR. CHESNEY: Nanotechnology for
our
3
experts? Dr. Fogel?
4
DR. FOGEL: I have read a little
bit about
5 it
in terms of reviews and my understanding, both
6
from a medical standpoint as well as an
7
electronic/technology standpoint, is that that is
8
like 10, 15 years down the road at a minimum,
9
although they are making large advances every
10
single day and I will probably have to eat my words
11 in
5 years. But I think at least the
estimates
12 from
the people who are really into it are that it
13 is
at least 10, 15 years down the road before we
14 see
anything.
15
DR. CHESNEY: Any other consultant
want to
16
speak to that issue? Dr. Gorman?
17
DR. LOEWKE: Dr. Chesney, I am
sorry, I
18
wanted to follow up on the last topic.
Before we
19 get
too far away from it I just wanted to ask a
20
question. Most of the comments
about extrapolation
21
appear to be from a safety standpoint and I was
22
wondering how you feel about efficacy from the
23
adult population and extrapolating that to the
24
pediatric population.
25
DR. GEVA: I think there is an
easy
264
1
answer. I think it is a big
no-no. I think you
2
simply cannot do that. It is just
a different
3
animal.
4
DR. CHESNEY: Dr. Fogel?
5
DR. FOGEL: Yes, I mean I think we
are,
6
one, dealing with different disease processes; two,
7 we
are dealing, as we all mentioned before, with
8
kids who are very small, with very tiny blood
9
vessels and that can make a real big difference,
10 and
I don't think you can extrapolate one from the
11
other.
12
DR. CHESNEY: Dr. Sable and Dr.
Siegel.
13
DR. SABLE: I would agree with
those
14
comments. There may be some
diseases that have a
15 few
exceptions--adolescents with heart transplants
16
versus adults. But I think the
vast majority of
17 the
diseases we do see in pediatric cardiology are
18
different though there are some that have enough of
19 an
overlap that would be a starting point to use
20
adult studies.
21
DR. CHESNEY: Dr. Siegel?
22
DR. SIEGEL: I am going to agree
with the
23
rest of the panel. Children are
different. Their
24
heart rates are faster. They are
not going to
25
cooperate. They can't hold their
breath,
265
1
particularly if we are talking under five or six.
2
They have less fat. So, you can't
really
3
extrapolate the efficacy from the adult studies. I
4
think when you get to the adolescent population you
5
probably can but in the younger population it is
6
going to be very difficult.
7
DR. CHESNEY: On the same
issue? Dr.
8
Danford?
9
DR. DANFORD: We are going to be
asked to
10
discuss what specific kinds of heart lesions might
11 be
special categories that warrant special
12
investigation. I am going to
throw that out for
13 the
panel of experts but I am going to ask you
14
about a specific one, and that is shunt lesions and
15 do
you find that you need to dose your contrast
16
material differently for any of these modalities in
17 the
setting of a shunt where your contrast might go
18
places that you don't necessarily want it right
19
away?
20 DR. CHESNEY: Dr. Siegel?
21
DR. SIEGEL: Well, for CT the
dosing will
22 not
change with the lesion. We have a set
23
technique and that is what we use.
What might
24
change is the timing of the study, whether I do it
25
during an earlier arterial phase or perhaps later
266
1 in
a venous phase, or trying to do just one phase
2 if
it is possible. But we use a very
standard
3
dose.
4 DR. CHESNEY: Dr. Fogel?
5
DR. FOGEL: When we inject
gadolinium what
6 we
do is we actually watch the gadolinium flow
7
through the body before we put our foot on the
8
pedal, if you will, to start the imaging for freeze
9
frame or, if it is time resolved gadolinium we
10
always have basically our imaging so we can tell
11
where the opacification is going to happen and then
12
start the imaging. So, in terms
of not seeing
13
things when we want to because of the shunt itself,
14 we
can time exactly when we start the imaging to
15 see
when we want to actually grab that freeze frame
16 to
do it.
17
I have to say that in all the gadolinium
18
studies that we have done, even those with shunt
19
lesions, we have never really had a problem in
20
terms of opacification. Now, if
we had studies
21
that were done that would decrease the dose and
22
keep ratcheting down the dose to its minimum
23
effective dose to decrease whatever safety issues
24
there might be, then, yes, I think that we might
25
have to take into account shunt lesions versus
267
1
non-shunt lesions. But at the
doses that we are
2
giving, at least with MRI gadolinium, we don't
3
really see any difference in terms of
4
opacification.
5
DR. CHESNEY: Dr. Sable?
6
DR. SABLE: With echo contrast
there is
7
absolutely no data.
Theoretically, if you are
8
trying to light up the left ventricle, if you had a
9
right to left shunt, you may actually have to use
10
less but it obviously depends on where the shunt is
11 and
the size of the shunt. It would be an
12
interesting thing to study but since all the
13
studies are done on opacification of patients
14
without shunts there really is no precedent for
15
even trying to answer the question accurately.
16
DR. CHESNEY: Same subject or a
different
17
one?
18
DR. NELSON: The same one.
19
DR. CHESNEY: Go ahead, Dr.
Nelson.
20
DR. NELSON: I am trying to figure
out why
21 I
am confused, and it may be because I am a simple
22
critical care medicine doctor.
23
[Laughter]
24
I mean, if a company hands me a catheter I
25
decide if I am going to be able to stick it in a
268
1 vessel
or not as long as they tell me the catheter
2 is
safe. And, I am trying to figure out
what is it
3
that we are going to ask--since ultimately I am
4
assuming that this kind of conversation would find
5 its
way into written requests, etc.--what is it
6
that we are going to ask the sponsor to do versus
7
what it is we are going to then do with whatever
8
tool they give us.
9
So, it is unclear to me if what we would
10
want them to have to do in order to fulfill the
11
requirement of the written request is to
12
demonstrate that it is better to image this lesion
13
doing it this way versus that way, using all the
14
different modalities, and the like, that have been
15
beautifully demonstrated. It is
clear that you
16
can't be a cardiologist unless you have very good
17
computer skills in imaging, and the like. So, it
18 is
unclear to me that you would expect them to do
19
that as opposed to give you tools that are safe,
20
that have been demonstrated that you can put into
21
someone at a certain dose. Then,
from there, it is
22 up
to the field to then do those kinds of studies.
23
So, that is where I am getting a little
24 bit
confused about a discussion of safety versus
25
efficacy. It is not that it
doesn't have to be
269
1
done but in my mind it is a question of who does
2
what. What do you expect to be
done in the
3
development of the product before the trials are
4
done to show whether it is better to do it by
5
contrast echo versus MRI or combination modalities,
6
etc.? It is not clear to me that
that would be
7
part of the actual agent development program.
8
DR. CHESNEY: Dr. Hudak?
9
DR. HUDAK: I am glad you said
that
10
because I am just a simple neonatologist and I am
11
having the same confusion. I
mean, you are the
12
experts. You have brought all
these techniques
13
forward. You showed marvelous
pictures. You have
14
shown us lots of different ways that these methods
15
sort of amplify the diagnostic abilities and
16
amplify your physiological understanding of
17
different situations. So, in
terms of efficacy I
18
have the same confusion. I mean,
you are the
19
experts; you know if this works or not; you know if
20 you
are seeing what you want to see; and you are
21 the
ones really to tell us. I don't know
that
22
there is a role for requesting studies that
23
demonstrate efficacy.
24
With respect to the echo, I have one
25
particular question and that is what sort of a time
270
1
window do you have after giving the injection to be
2
able to conduct your study?
3
DR. SABLE: I will answer your
second
4
specific question and then comment on your first
5
comment. There have been several
adult studies
6
looking at this. If you do a
bolus injection you
7
probably have 5-7 minutes. So, if
you are doing a
8
stress study you would probably give 2 boluses. I
9
didn't show this in my slides but I have several
10
slides on this topic. If you do a
continuous
11
infusion with a very low dose you can probably do
12 it
for 20 or 30 minutes. A typical echo
without
13
exercise, just looking at functional wall motion,
14
probably can be done in 10 minutes.
So, a single
15
bolus--the goal of it is to last the length of the
16
study.
17
In terms of your first question and
18
comment, I think echo is much more immature in
19
terms of how contrast echo has been used in
20
children than the other modalities here.
So, from
21 my
own field I would make a plea that we definitely
22 do
need help in trying to get some pediatric
23
studies off the ground looking at efficacy.
24
DR. HUDAK: I guess with regard to
that I
25 am
not sure what the role of the FDA or this
271
1
committee is with respect to that issue.
I mean, I
2
think that the way these technologies have
3
progressed--I mean, they are out of the box and
4
going forward before agencies like this even get a
5
chance to get a handle on what is going on.
6
With regard to the other issue, the safety
7
issue, I couldn't agree more with the safety
8
concerns. I think, again, the
critical issue, as
9 Dr.
Fost suggested, are the things that we are not
10
going to necessarily know for years to half a
11
lifetime. I think that certainly
with any of these
12 new
agents or new technologies or sonicating funny
13
bubbles in the blood and in the organs, one needs
14 to
carefully consider what registries or long-term
15
follow-up one needs to establish on these patients
16 to
have some sort of mechanism to see exactly what
17
happens to these patients. It is
certainly not
18
going to be a randomized, controlled study but I
19
think that, you know, 20 years from now we
20
certainly want to know if there are any major
21
complications from some of these techniques.
22
DR. CHESNEY: Can I just ask--and
I don't
23
know how good an analogy this is but we have been
24
using antibiotics for years and not knowing dosing;
25 not
knowing really precise efficacy. We knew
they
272
1
worked in adults. We extrapolated
to children. We
2
didn't know about the metabolism.
And, I think
3
that is what I am hearing from our colleagues here,
4
which is that maybe they would get better pictures
5 if
they had a different concentration of the drug
6 or
understood its metabolism better. Dr.
Fogel?
7
DR. FOGEL: I am just a simple
8
cardiologist; let me say that.
9
[Laughter]
10
For me, I think that when one looks at a
11
drug one not only has to consider--I keep getting
12 the
sense that a lot of people are trying to
13
separate the efficacy and the safety.
We have
14
always been taught, you know, that it is a
15
risk/benefit. For example, we may
be using 0.1
16
mM/kg in kids and seeing that things are fine but
17 how
do I know that with 0.5 mM/kg I couldn't see
18
something just as fine. You know,
in general,
19
although it is not a general rule, one thinks that
20 the
lower the dose you give the better the safety
21
profile of the drug would be, and that is not
22
necessarily the case in every single drug but in a
23
substantial portion of the drugs that are out there
24 you
would think that common sense would tell you
25
that that would be better.
273
1
So, for me, I would want to see, one,
2
clinical trials, controlled clinical trials not
3
open-label Phase IV reporting, rigorous controlled
4
clinical trials looking at various doses and dose
5
response, and then safety and then having a log of
6
patients who are getting it and, hopefully they
7
would consent to it to be able to follow them up in
8 10
years.
9
DR. CHESNEY: Dr. Loewke?
10
DR. LOEWKE: I agree. We definitely look
11 at
things from a risk/benefit perspective and we
12
look at the safety of the product and the efficacy
13 of
the product in the patient
population. Much of
14
what we are talking about here, and the drugs that
15 are
being used, and for the purposes for which they
16 are
being used are not approved in kids. So,
we
17
don't have knowledge that these products, when used
18 in
kids, would give us the right information to go
19
forward with. We don't have that
information and
20
that is why we are also talking about efficacy
21
here.
22
If you talk about extrapolation and
23
extrapolating efficacy data from adults to kids, we
24
stated that then you could use the adult data as
25
your basis for efficacy to support efficacy in kids
274
1 and
then you would do additional studies,
2
pharmacokinetic studies and safety studies in
3
pediatrics. But here I am
hearing, if I am
4
correct, that you feel we need to pursue efficacy
5 as
well as safety in the pediatric population.
6
DR. CHESNEY: Can I just make a
comment?
7 Is
it safe to say that efficacy in your world is a
8
better image? Is that a fair
statement or not?
9 What
is efficacy as you see it? Dr. Siegel?
10
DR. SIEGEL: Well, I don't think
efficacy
11 is
a better image. We would love to have
that.
12 But
is it an image that provides useful clinical
13
information? Does it get it
right? Is it
14
accurate? Can you make a
diagnosis with it? That
15 is
efficacy. I mean, is it an accurate
imaging
16
test, whatever we use it for--for diagnosis or
17
improving patient management? We
like pretty
18
pictures. Of course, we would
like them to look
19
better but when we are talking about efficacy I
20
think that is it. Safety is
obviously its own
21
issue.
22
I am not sure where dose is falling into,
23 if
it is safety or if it is for efficacy to get
24
prettier pictures. But I think we
all work under
25 the
assumption that less is better. If we
can give
275
1
less contrast, that would be better for the
2
patient, although we don't know that and we really
3
don't know what dose works out there and what the
4
risk factors are. When we report
the adverse
5
reactions we never say really what the dose was
6
that was given. We presume it was
just a standard
7
dose. So, dosing would be
important, if we could
8
have a study that would say at different doses we
9 get
different outcomes or reactions--safety; and
10
also different diagnostic quality.
11
DR. CHESNEY: Thank you. I think that is
12
similar for almost every drug we use, the lower the
13
dose, the better. In my world of
antibiotics if
14 you
give less, wouldn't that be better in the long
15
run? Dr. D'Agostino and then Dr.
Sable.
16 DR. D'AGOSTINO: It would help me greatly,
17 and
I also am simple-minded--it would help me
18
greatly if I could have some discussion from the
19
experts on what is, in fact, the indication. We
20
have been told by the FDA there are four
21
indications that they are interested in--structure
22
delineation, disease detection,
23
functional/physiological assessment and diagnostic.
24
Like, in the MRI it seemed like you could do
25
everything. In the CT it seemed
like it was only
276
1
diagnostic. It would help me very
much when we
2
come to these questions if I sort of knew what
3
these were aiming at, and what is it that this
4
population should look like, what the sample should
5
look like. Is that reasonable to
ask of the
6
speakers, if they could just sort of rattle off
7
what they think their modalities are aiming at?
8
DR. CHESNEY: Dr. Loewke?
9
DR. LOEWKE: I think that is one
of the
10
major questions to the panel for the discussion
11
that we have planned.
12
DR. D'AGOSTINO: Well, the
speakers didn't
13
necessarily present their material in that way. If
14 we
could start having the speakers tell us what
15
they think is going on, then I think we could agree
16 or
disagree with them. I mean, they use
quite
17
different vocabulary.
18
DR. CHESNEY: Did you want to propose
a
19
vocabulary that we should ask them to use?
20
DR. D'AGOSTINO: Well, we have
been given
21 the
vocabulary by the FDA and the speakers didn't
22
necessarily use that vocabulary.
So, if you could
23
just rattle off, each of the speakers saying is it
24 a
diagnostic tool that they have; is it a
25
structural delineation tool that they have?
277
1
DR. LOEWKE: It may vary depending
on the
2
population you are studying, what your endpoints
3
would be and what type of indication a manufacturer
4
would seek. So, I think if we
have our discussion
5
about what populations you feel need additional
6
study for the drug classes some of that is going to
7
come out as we go through the questions tomorrow.
8
DR. D'AGOSTINO: Why do you not
want to
9
have the speakers tell us what they think--is there
10 any
reason?
11
DR. LOEWKE: Time-wise--
12
DR. D'AGOSTINO: I am talking
about
13
something that would take two minutes at most on
14 the
part of the speakers. I mean, the one
for CT
15
said it is for diagnostics. Does
that exclude
16
others? It would help us I think
in terms of
17
answering the questions.
18
DR. CHESNEY: As long as it only
takes two
19
minutes for each speaker and each speaker
20
understands what you are asking for because I am
21 not
quite sure I do. But if you all are
clear,
22
then let's go ahead.
23
DR. D'AGOSTINO: The FDA said
there are
24
four indications. I am not asking
something
25
profound. The FDA said there are
four indications,
278
1
structural delineation, disease assessment,
2
functional assessment, diagnostic.
When Dr. Fogel
3
made his presentation he chose to use the
4
words--let me see if I can fish it out--anatomy,
5 blood flow, tissue characteristics. Are they all
6
structural? Are they
different? I mean, it is a
7
different vocabulary.
8
DR. CHESNEY: Dr. Geva, Tom has
singled
9 you
out.
10
DR. GEVA: I think it is actually
quite
11
complicated and perhaps one can differentiate
12
between an outcome variable for a trial as opposed
13 to
what is clinical reality. In clinical
reality I
14
think that in most cases what we are being asked to
15 do
is to evaluate a set of clinical questions and
16 it
depends on the imaging modality that you are
17
using, but it is rare to really draw these concrete
18
boundaries between structure, anatomy--this is
19
somewhat artificial.
20
DR. D'AGOSTINO: But we are asked
to
21
design or help them design clinical trials so you
22 are
going to have to do that.
23
DR. GEVA: Exactly, I agree. As I said,
24 it
is useful perhaps to distinguish between
25
defining endpoints for clinical trials and to try
279
1 and
formulate the indications for the use of
2
specific contrast agents sort of in an
3
all-inclusive fashion. I do think
that we need to
4
make that effort and one of my hopes for all of
5
these discussions is to be able to come to a
6
conclusion about indications for use of, let's say,
7
contrast agents in pediatric cardiac imaging.
8
DR. CHESNEY: I think maybe Dr.
Loewke was
9
referring to this, that maybe this is something we
10
should address in the morning with respect to
11
specific endpoints, specific studies, specific
12
conditions and so on, whereas now I think we are
13
more asking questions of the presentations that
14
were given, although Dr. D'Agostino's is a broader
15
question. Dr. Loewke, did you
want to comment? I
16
have a whole list of questions still here that
17
people are asking.
18 DR. LOEWKE: I agree.
I think, as we talk
19
more about the populations that need additional
20
study and what endpoints you would recommend, we
21
will be able to figure out from that what types of
22
indications could be sought based on the population
23
studies and the clinical value of the information
24 you
are going to obtain.
25
DR. D'AGOSTINO: Why wouldn't you
ask the
280
1
question the other way around? If
you want
2
structural delineation, then what type of
3
population and what type of study would you run, as
4
opposed to a diffuse question--well, here is a
5
population, what kind of indication do I want? Why
6
aren't you addressing it the other way around?
7
DR. LOEWKE: We are trying to
assess how
8
these products are being used out there, and that
9 is
the information--
10
DR. D'AGOSTINO: That is what I am
asking,
11 how
are they being used, and then that will tell us
12 how
to, hopefully, put studies together.
13
DR. CHESNEY: Can we tackle this
long list
14
here? I have Sable, Ebert, Fogel,
Nelson, Fink,
15
Moore. So, Dr. Sable, you are
first on the list.
16
DR. SABLE: In terms to referring
to his
17
comment or just previous questions?
18
[Laughter]
19
DR. CHESNEY: Whatever!
20
DR. SABLE: I just wanted to add
one thing
21 to
the efficacy/safety issue. In many cases
we
22
move from one modality to the other as we get
23
better at them. If echo is the
least invasive and
24
safest thing to do, if we find new reasons to do
25
echo it may lead to safer management of our
281
1
patients overall. So, I think,
again, it is almost
2
impossible to separate safety and efficacy because
3 we
are really trying to do both with everything we
4
do. If I come up with new ways of
keeping kids out
5 of
the cath lab, if Dr. Moore comes up with ways
6 for
keeping patients out of the operating room,
7
then we have achieved both and I don't see any way
8 to
separate them.
9
DR. CHESNEY: Thank you. Dr. Ebert?
10
DR. EBERT: I don't want to
belabor the
11
point on dosing but I would like perhaps some of
12 the
experts to address the issue of dose ranging
13 and
how well that has really been established in
14
adults. We are talking about dose
ranging of these
15
agents in pediatrics but my impression from some of
16 the
presentations is that we may not even have the
17
dose ranging established for these agents in the
18
adult population. There was some
mention of
19
different infusion rates for example, but there may
20 be
some benefits of trying to do this in adults so
21 it
is not an extrapolation per se but if we can
22
show that this is a relatively flat dose-response
23
relationship or a steeper curve, does that give us
24
some information in the pediatric population?
25
DR. CHESNEY: Dr. Dilsizian?
282
1
DR. DILSIZIAN: I can answer that
from the
2
nuclear perspective. For example,
if you take a
3
traditional thallium stress study and go back to
4 the
literature, the usual dose of injection is 2
5 mCi
for adults, but the range is up to 5 mCi.
With
6
time it has gone up to 3 mCi, 3.5 mCi.
Now we
7
double the dose and the reason for that is,
8
obviously, the quality of the images or maybe the
9
obesity population. Maybe the
weight change also
10 dictates the dose. But we have a range and the
11
range is pretty large. Also, even
with technetium
12
perfusion tracers, although the package insert will
13 say
8 mCi at rest and 22 mCi with stress, if the
14
patient is large we can give up to 30, 35, 40 mCi.
15 So,
we do have a range.
16
How do we decide that? Well, it
has been
17
more anecdotal. It hasn't been a
series of
18
patients, for example, with 20, to 30, to 40 to
19
say, you know, well, if you are above 100 kg, which
20 is
what I do in my lab now--I say above 100 kg I
21
want to do two large dose technetium studies
22
because in my experience that is what is shown.
23 But
no one has shown that 100 kg is the cutting
24
edge. Maybe you would like to
have that type of
25
study, maybe some dose escalation with some
283
1
methodology to say, well, what is the optimum dose
2 and
what is the range.
3 DR. CHESNEY: Dr. Fogel, Dr. Nelson, Dr.
4
Fink and Dr. Moore.
5
DR. FOGEL: At least with the
6
gadolinium--and I have to say I am not as familiar
7
with the adult dose ranging trials and I don't even
8
know if there were any--I know in children, for
9
example, when not as much gadolinium got in as was
10
intended we have had less opacification and less
11
diagnostic imaging than we would like.
I would
12
personally like to know what the minimum dosage
13
would be in the various age ranges that I could use
14 to
get a diagnostic study but I have to say, from
15 an
anecdotal standpoint, there must be some dose
16
response and it is probably steep in the small dose
17
ranges and that is where I want to be.
18
DR. CHESNEY: I think this is
fascinating.
19 I
am glad you brought this to us because I think
20
most of us just assumed that this had all been
21
worked out; you know exactly what you are giving
22 and
why; and when we send a patient down for an
23
X-ray it is guarantied safe and effective, and now
24 we
are discovering that it has never been done.
25
This is very interesting--at least in children.
284
1 Dr.
Nelson?
2
DR. NELSON: I would like to
change the
3
topic to one that I notice isn't on our questions
4 for
tomorrow but it might become a part of the
5
discussion of CT scans and the nuclear area, and
6
that is the radiation risk that was mentioned by a
7
couple of speakers.
8
I guess my question is to what extent,
9
other than the one study that was quoted which I
10
have not looked at, to what extent are a lot of the
11
figures about radiation risk extrapolated based on
12 a
linear theory of risk? I will say that
at least
13 in
my institution we have deviated from that and
14
have, in one case, approved up to 2 rem on a SPECT
15
scan for a non-therapeutic, non-direct benefit
16
procedure on the argument that there is, in fact,
17 no
documented risk of any radiation and that most
18 of
this is all just linear extrapolation.
Except
19 for
that one study, which I would have to look at
20 and
see where that would fit in with all the data,
21
some of the other studies that have looked at
22
epidemiololy have shown no evidence of radiation
23
risk at low levels. So, we
concluded in looking at
24 it
that one couldn't say there was any risk below 5
25 rem
and then felt that under those circumstances it
285
1
might be appropriate to go forward.
2
So, I just put that on the table because 5
3 rem
strikes me as an exceedingly low number if, in
4
fact, you are going to be doing studies that are
5
outside of the potential for benefit.
Now, if you
6 are
doing studies under that rubric you are not as
7
limited to the risk, thinking of the IRB
8
categories, but I just wanted to get that on the
9
table to have some conversation about that, whether
10
that will be a backdrop for discussions of those
11 two
development plans tomorrow or not.
12
DR. DILSIZIAN: I am glad you
brought this
13
up. Obviously, that was my
conclusion in that that
14 is
very low. If you look at even PET
radiotracers
15
with short half-lives, if you look at the body
16
distribution even in research in kids to make some
17 new
diagnostic metabolic finding in cardiomyopathy,
18 we
are not allowed to if we follow the 0.5 rem
19
rule. So, we need to, in essence,
come up with a
20
better endpoint. I agree.
21 DR. CHESNEY: Dr. Fink, Dr. Moore and Dr.
22
Siegel.
23
DR. FINK: Yes, as the discussion
24
progresses I guess one of the questions that occurs
25 to
me is have we done our homework? We
don't have
286
1 a
lot of background data and if we are going to
2
study these agents in kids, don't we really need
3
some of the background data?
Particularly in the
4
imaging field it would seem that this is an arena
5
that is particularly well suited to going back to
6
animal models; that animal models could answer many
7 of
the technical questions in terms of dye dosage.
8 You
have a range of different sizes you can look
9 at; different heart rates. You can even answer
10
some of the questions of pulmonary capillary
11
toxicity to particulates. You can
put in
12
catheters. You can measure
minimal changes in
13
oxygenation. And, should we be
discussing human
14
studies and using children as guinea pigs when we
15
have guinea pigs?
16
DR. CHESNEY: I am going to think
that was
17
rhetorical.
18
[Laughter]
19
Point well taken. Drs. Moore,
Siegel and
20
Gorman.
21
DR. MOORE: Just a follow-up to
the
22
radiation comment, that is the one thing I think
23 you
do have to keep in perspective with these
24
patients is that these procedures are repetitive
25
diagnostic follow-up procedures on these patients.
287
1 So,
the exposures you are talking about acutely
2
certainly are relevant but many of these patients
3
start in infancy and continue throughout their life
4 and
throughout their adult life to go ahead and
5
accumulate these radiation exposures.
So, I think
6
that just needs to be considered in that particular
7
issue with this patient category.
It is very
8
different than some other areas.
9
DR. CHESNEY: Dr. Siegel?
10
DR. SIEGEL: To respond to a few
of the
11
points, first of all, dose. The
dose that I stated
12 was
2 mL/kg. We use that; we know that it is
safe.
13 I
mean, the contrast agents are sort of maturing
14 and
I think it is an issue of the safety there; we
15
have been there. But when it
comes to dose, that
16 is
an area that can be investigated. CT has
much
17
better resolution. That is why we
like it. We get
18
thinner sections; we are able to see more anatomy.
19 We
should be able to do less in the way of dose and
20
volume. I do but I am one person
and it works if I
21 get
down to 1 mL/kg. I know it does. I can't
22 necessarily
get in the full volume but I can't
23
prove that to anybody unless we do the research
24
with that.
25
Let me just get to the animal models and I
288
1 will go back to radiation. We are using animal
2
models actually. We are doing
research now on
3
animal models that are closer to adults I think,
4
looking at the amount of contrast we need to get a
5
diagnostic examination--the amount of the
6
concentration that I talked about, all the
7
parameters. I have looked at the
issue of doing
8
this also on animals that would be similar to
9
infants. It is difficult. Of course, doing animal
10
research is even becoming more difficult than doing
11
human research so nothing sounds that easy in this
12
world. But I think if we can get
the support out
13
there, that is what we need to be able to do, to
14 get
back to the basics and show it there.
15 Radiation dose--most of the
radiation dose
16
that we are talking about with CT, this being, you
17
know, a new use for CT now, a lot of it is going
18
back to the atomic bomb and, you know, doing this
19
extrapolation. We have no data on
CT and we are
20
talking about different types of, you know
21
radiation and different exposure times and
22
different intensities in any one moment in time.
23 So,
there is a lot of work to be done out there to
24
look at this dose factor, this radiation dose
25
factor and then the diagnostic or efficacy ability.
289
1
DR. CHESNEY: Dr. Fogel and then
Dr.
2
Gorman.
3
DR. FOGEL: In terms of the
4
radiopharmaceuticals and the radiation exposure, I
5
guess I am not 100 percent clear that I am totally
6
sanguine with the notion of the effective dose and
7
tissue weighting factor. I guess
if you read the
8
definition correctly it takes into account fatal
9
cancers and the risk of hereditary disease. So,
10
that means that non-fatal cancers, ones that we
11
have 90 percent cure rates for, are not taken into
12
account when we look at the total effective dose.
13 So,
I guess I am wondering doesn't that minimize
14
what the risk is? What if it
induces cancers that
15
have a 90 percent cure rate and that doesn't count
16 in
the total effective dose? What if the
radiation
17
induces cardiomyopathy in children?
That doesn't
18 get
factored into this total effective dose.
So, I
19
guess I am not 100 percent happy with using total
20
effective dose as a number by which one can then
21
hang their hat on, saying this is a safe dose or
22
this is not a safe dose. I am
wondering if there
23 is
any comment.
24
DR. CHESNEY: Not this late in the
25
day--hold back until tomorrow morning!
Dr. Gorman,
290
1 and
we do have two speakers for our open public
2
hearing today--three? I am sorry.
3
DR. GORMAN: One of the issues
that is
4
becoming increasingly clear to me as I have
5
listened to you talk is that we have at least three
6
different technologies and at least three different
7
maturities of the contrast agents we are talking
8
about. I think when we talk about
ionizing
9
radiation, whether the cath lab or CT, we have a
10 lot
of information. When we go to the MRI we
have
11
less and when we go to echocardiography we have
12
even less. I would like our
experts to postulate,
13
looking into the future, is there going to be
14
enhancement of the technology of the device or
15
enhancement of the contrast agents that are going
16 to
lead to increasing diagnostic ability of your
17
technology?
18
DR. CHESNEY: Dr. Sable?
19
DR. SABLE: I think with regard to
20
ultrasound it is probably going to be both but
21
probably more with the agents themselves as we
22
begin to think about therapeutic ultrasound. As I
23
said in my talk, I think the biggest gap is between
24
volume and use of contrast.
Pretty much every
25
catheterization uses contrast and most MRIs and all
291
1 CTs
usw contrast, and echo. We are using it
in
2
zero percent of our studies; we probably should be
3
using it in some number far greater than that.
4
But, clearly, the agents have to get a little bit
5
better. The machines are pretty
much there for us
6 to
use it in their current state but the potential
7 to
go much further is certainly there.
8
DR. CHESNEY: Dr. Fogel?
9
DR. FOGEL: I think with
gadolinium
10
agents, just like echo, it is probably both, again,
11
more weighted towards the agent itself.
I am
12
thinking more along the lines of the blood pool
13
agents and molecular imaging. I
would also have to
14 say
that with 3 tesla machines coming on line and
15
with the software always becoming better and faster
16
scans we will be able to do more and more with the
17
agents we already have and, hopefully, more and
18
more with the agents that are coming.
19
DR. GORMAN: When you talk about
20
increasing the magnetic strength of the coil, what
21
does that do for you? Does that give
you increased
22
resolution or increased speed or both?
23
DR. FOGEL: Both.
24
DR. CHESNEY: Dr. Siegel?
25
DR. SIEGEL: As you stated, the CT
is more
292
1
mature so I think the advances we will see there
2
will be more in the device, basically how fast we
3 can
give it and the time to start scan. The
only
4
thing in the contrast agents, as I mentioned, might
5 be the
concentration. It is already out there,
the
6 400
mg of iodine. The question is can we
change
7 the
viscosity. Most of the advancements at
this
8
point will be in the new technology that is coming
9 out
in the device.
10
DR. CHESNEY: Yes?
11
DR. LOEWKE: Dr. Chesney, can I
ask a
12
question? As you mentioned, many
of these
13
modalities can be used without the contrast agent.
14 As
Dr. Siegel pointed out, she is not doing cardiac
15 CT
unless she is using a contrast agent. I
would
16
like to know, in your routine clinical practices
17 for
the patients you see, do you do non-contrast
18
images? They are not effective
and then you move
19 on
to contrast? Do you automatically start
with
20
contrast enhanced images? Then,
and I don't know
21 if
you can do this, what is the first-line
22
diagnostic? Is it ultrasound and
if ultrasound
23
doesn't give you the answer do you go to MR? Is
24
there a hierarchy or a path you follow?
And, are
25
there certain patient populations where, if this is
293
1
non-diagnostic, you move to this test, if that is
2 not
diagnostic--if you could give some input.
3
DR. CHESNEY: Dr. Sable?
4
DR. SABLE: In our practice and I
think
5
most pediatric cardiology practices ultrasound is
6
definitely the first-line of imaging modalities.
7
Then you kind of take your pick as to what comes
8
next.
9
DR. LOEWKE: That is non-contrast?
10
DR. SABLE: In our practice we
don't use
11
contrast yet. As I said, there is
one group out
12
there--a few places are using it a little bit but
13
there is only one group that has done enough to
14
publish. So, unlike all my other
colleagues, we
15
would almost never--we are thinking about starting
16 a
contrast program but we haven't done so yet.
17
There is a small percentage of our patients that we
18
think clearly would benefit from contrast echo.
19
Those patients are now getting sent to MRI, CT or
20
angiography. So.
21
DR. CHESNEY: Dr. Siegel and Dr.
Fogel.
22
DR. SIEGEL: As far as
non-contrast goes,
23 we
don't use it. If you are doing cardiac
it
24
really is contrast. There will be
an occasional
25
exception. If you are looking for
calcification
294
1 you
might do it but the contrast resolution is so
2
poor that all you are doing is wasting radiation.
3 In
that instance we will go for the contrast
4
enhanced because of that issue.
5
As far as first-line of imaging, I totally
6
agree that if it is cardiac or intracardiac related
7 we
will be using echo. But our approach if
it is
8
extracardiac where we are wondering about
9
mediastinal great vessels is, if there is a
10
vascular ring or abnormal arch, then we are going
11 to
CT. So, we sort of will do
stratification based
12 on
the lesion that we are interested in.
13
DR. CHESNEY: Dr. Fogel and then
Dr. Geva.
14
DR. FOGEL: In terms of MRI and contrast
15
versus non-contrast, we actually view contrast as
16 an
adjunct to the non-contrast images. We
will
17
always get the non-contrast images first unless we
18 are
doing viability and perfusion, in which case we
19 do
contrast very early on in the study. For
the
20
most part we will do the non-contrast ones first.
21
That is because if you do the contrast ones first
22 you
can't get good dark blood images if that is
23
what you are trying to do. Plus,
we feel that in
24
terms of it being an imaging modality, in and of
25
itself it is more of an adjunct with some rare
295
1
exceptions, like viability and perfusion. It adds
2 to
the diagnostic information but we always get the
3
non-contrast ones as well.
4
In terms of the order in which one gets
5
imaging studies or the protocol by which one gets
6
imaging with relation to a specific disease or
7 specific clinical syndrome, I think we do echo
8
before we do something like MRI or cath.
I have to
9 say
that there is some good justification for it.
10
There are times when that is done because the
11
people who are managing the patient's course aren't
12
necessarily educated enough in terms of all the
13
diagnostic imaging modalities to tell which one is
14 the
optimal one to do first, and because echo, as
15 Tal
said, is being used almost like a stethoscope
16 it
almost comes like a knee-jerk reaction, "let's
17 get
an echo first and then whatever we can't do we
18
will get by another non-invasive imaging modality."
19 But
there are certain things that have been shown
20 to
be nearly gold standards like vascular ring
21
anatomy by MRI, ventricular function parameters by
22 MRI
that are clearly better than echo but, yet, we
23
will generally see an echo always being performed
24
first. I think that is because of
the education of
25 our
colleagues rather than the fact that it is a
296
1
better imaging modality for those specific types of
2
disease entities.
3
DR. CHESNEY: Dr. Geva?
4 DR. GEVA: I agree with what Mark has just
5
said. Just to add, I think that
what you are
6
hearing here is a little bit a reflection of
7
variations in access to technology and expertise
8
around the country in various centers.
That all
9
comes after the echocardiogram.
As far as use of
10
contrast agents in pediatric ultrasound, I agree
11
with Craig, at this point in time it is esoteric;
12 it
is rare. It is being used in very, very
small
13
numbers.
14
DR. CHESNEY: Dr. Sable?
15
DR. SABLE: The other thing I
think to
16
keep in mind when you listen to us speak, we are
17
somewhat of a biased group when you have MRI and CT
18 and
cath experts from around the country. If
you
19 go
out into the community in small pediatric
20
cardiology practices I think it is even more
21
weighted toward echo because of the availability
22 and
the portability, not that it is a better
23
technique. It is just so easily
obtainable.
24 Open Public Hearing
25
DR. CHESNEY: I think maybe we
should move
297
1 on
to the open public hearing. We do have
2 something
that I have to read but my understanding
3 is
that Dr. Gelfand and Dr. Duffy, on Dr.
4
Gardiner's behalf, will be making presentations,
5 and
the other two speakers are just going to
6
provide us with handouts. Am I
correct about that?
7
MR. PEREZ: No, there is one
additional
8
handout and two statements. The
handout is in your
9
packets.
10
DR. CHESNEY: So, we have three
11
altogether, people who are going to speak--four
12 people
who are going to speak.
13
This has to be read before the open public
14
hearing. Both the Food and Drug
Administration and
15 the
public believe in a transparent process for
16
information gathering and decision-making. To
17
ensure such transparency at the open public hearing
18
session of the advisory committee meeting, FDA
19
believes that it is important to understand the
20
context of an individual's presentation.
For this
21
reason, FDA encourages you, the open public hearing
22
speaker, at the beginning of your written or oral
23
statement to advise the committee of any financial
24
relationship that you may have with any company or
25 any
group that is likely to be impacted by the
298
1
topic of this meeting. For
example, the financial
2
information may include a company's or a group's
3
payment of your travel, lodging or other expenses
4 in
connection with your attendance at the meeting.
5
Likewise, FDA encourages you at the beginning of
6
your statement to advise the committee if you do
7 not
have any such financial relationships.
If you
8
choose not to address this issue of financial
9
relationships at the beginning of your statement it
10
will not preclude you from speaking.
11
Our first open public hearing speaker is
12 Dr.
Michael Gelfand.
13
DR. GELFAND: I am Dr. Michael
Gelfand.
14
[Slide]
15
I am the immediate past president of the
16
Society of Nuclear Medicine. My
trip was funded by
17 the
Society of Nuclear Medicine, which is the large
18
professional organization in nuclear medicine, a
19
scientific organization. I have
no current
20
relationships with any of the manufacturers in the
21
drug field. I have never been a
consultant for any
22 of
them, nor have I ever received any honoraria
23
from them. I am Professor of
Radiology and
24
Pediatrics at the University of Cincinnati and the
25
head of Nuclear Medicine at Children's Hospital.
299
1
[Slide]
2
I basically want to point out the context
3 of
pediatric nuclear medicine with some reference
4 to
cardiac imaging. There is going to be
some
5
deviation from that but, basically, the pediatric
6
nuclear medicine is alive and well and growing.
7 [Slide]
8
The number of nuclear medicine procedures
9
done in children's hospitals--I was able to get the
10
figures from Boston and Philadelphia.
This is the
11
annual volume in 2003. These are
hospitals that do
12
about 150,000 total imaging procedures per year in
13
each case. So, it runs to 3, 4, 5
percent of the
14
total imaging.
15
[Slide]
16
The distribution of studies is quite
17
different from adult nuclear medicine and varies a
18 lot
from hospital to hospital. What studies
are
19
being performed?
20
[Slide]
21
It turns out that the largest percentage
22 of
what we do is GU studies. We do tumor
imaging,
23 GI
imaging, bone imaging which is a fair component
24 of
it, and others.
25
[Slide]
300
1
To break that down further, GU cases
2
include cystography in our institution.
Just to
3
give you an idea of the radiopharmaceuticals that
4 we
are using, some of these are heritage
5
radiopharmaceuticals that go back many, many years;
6
some of them are more recent.
7
[Slide]
8
We do tumor imaging. I might point out
9
that half of our tumor volume is with agents that
10 are
either in a gray area or are fully approved by
11 the
FDA. This is an IND agent. FDG-PET is sort of
12 in
a gray area. There is a special NDA type
of
13
situation for FDG right now which will change
14
according to congressional mandate at some point.
15
Actually, a lot of cardiac imaging is lung imaging,
16 as
was pointed out at the University of California
17 at
San Francisco, probably two-thirds of this, and
18
this is done with technetium-MAA.
19
I might point out here is an example where
20
safety is not in the package insert.
If my
21
technologist were to mix 1 mCi of technetium or 5
22 mCi
with the kit and make it up and then I was to
23
give this dose in an appropriate amount to an
24
infant, we would have a problem.
We would have a
25
clinical adverse effect because, in fact, this
301
1
infant may be getting 30, 50 times as many
2
particles as an adult would get, perhaps even more.
3
This kind of information is often not in package
4
inserts.
5
We do brain perfusion, endocrine and
6
mostly thyroid, and we do heart imaging at our
7
hospital but in our particular case we do not do as
8
much as, say, Boston or Philadelphia where they do
9
substantial amounts.
10
[Slide]
11
At Cincinnati Children's Hospital we have
12
experienced continued growth in nuclear medicine
13
volumes, but at a somewhat slower rate than the
14
total number of imaging exams.
15
[Slide]
16
We have been growing at 4.8 percent per
17
year in nuclear medicine. I might
point out that
18
this is the year that I was president of the
19
Society of Nuclear Medicine and half of this year,
20 and
when I came back and paid attention to what I
21 did
for a living we had the best half year we have
22
ever had. We have been having a
7.5 percent
23
increase a year in the radiology department.
24
[Slide]
25
Boston and Philadelphia, according to the
302
1
information I was given by the department chiefs in
2
those areas, are also reporting increasing volumes
3
from year to year. Pediatric
nuclear medicine case
4
volumes are dependent on having an imaging
5
physician who is interested in pediatric nuclear
6
medicine. If the staff imaging
physicians in a
7
hospital are disinterested or believe that nuclear
8
medicine is likely to disappear or pediatric
9
nuclear medicine is likely to disappear, this
10
becomes a self-fulfilling prophesy.
11
[Slide]
12
The numbers of myocardial perfusion
13
imaging studies, according to manufacturers' data,
14
were about 4,000 per year in the U.S. in 2002. It
15 may
actually be slightly more if you brought in
16
another brand. Boston does about
100 per year or
17
over 1 percent of their nuclear medicine volume.
18
Philadelphia did 224 last year, which is about 3
19
percent of their nuclear medicine volume, and this
20
number is not that far below the number of MR
21
contrast administrations for cardiac imaging
22
according to the information we were given earlier.
23
[Slide]
24
What can you do myocardial perfusion
25
imaging for? In children one is
Kawasaki's
303
1
disease, as was alluded to. In a
study published
2 in
The Journal of the American College of
3
Cardiology, in 46 patients myocardial perfusion
4
defects were present by mibi; in 37 percent of 27
5
patients who had normal coronary arteries by
6
angiography; in 63 percent of 11 who had resolved
7
aneurysms; and in all the patients who still had
8
aneurysms. So, that is one
indication that is
9
solid.
10
Another, we are getting information about
11
hypertrophic cardiomyopathy.
Another possible
12
indication is after the arterial switch operation
13
where there are fixed perfusion defects in a
14
considerable number of children.
In this one study
15
almost all the children had fixed perfusion defects
16 by
mibi imaging after the switch operation a number
17 of
years later.
18
[Slide]
19
Myocardial perfusion imaging in pediatrics
20
with technetium-labeled radiopharmaceuticals--one
21 of
the technetium agents has a shorter half-time
22 and
a considerably lower radiation dose;
23
thallium-201, better image quality, flexible timing
24 of
image acquisition and you can do a gated wall
25
motion study as well as get information about wall
304
1
motion, which may give you some feeling as to what
2 is
working and what is not working.
3
[Slide]
4
Radiation exposure from diagnostic
5
pediatric nuclear medicine procedures is
6
acceptable. Comparisons between
different
7 radiographic
procedures, and between radiographic
8
procedures and nuclear medicine procedures is
9
accomplished by use of effective dose calculations.
10
This is really the industry standard.
It has taken
11
over from whole-body dose. It has
taken over from
12
exposed dose from individual organ doses because of
13 the
weighting. Of course, any weighting
scheme is
14
going to be somewhat imperfect but that is the best
15 we
have and it is the industry standard.
16 [Slide]
17
Effective dose has a weighting factor for
18
each tissue and a calculated dose for each tissue.
19 If
you sum it up across a number of tissues, 10,
20 12,
15 tissues, you have an estimate of the risk to
21 the
patient. Implicit in that radiation dose
it
22
should have a lot to do with what the patient would
23 get
if they just got a whole-body exposure, you
24
know, standing 5 miles from the Hiroshima bomb for
25
example.
305
1
[Slide]
2
To give you an idea of how some of these
3
things fit in, in tumor imaging, CT of the chest,
4
abdomen and pelvis, and this is using the low dose
5 Tc,
as was alluded to by Dr. Siegel. This is
6
probably a third or fourth of what people used to
7 get
in a lot of places--very comparable to what we
8 do
with tumor imaging in PET, and less than gallium
9
which is a long half-life radiopharmaceutical, 2.7
10
days. It turns out that our
neuroblastoma imaging
11
with I-123-MIBG is about half of either of either
12 of
those two.
13
One of the interesting things too is when
14 I
was preparing the article with Mike Staven on
15
pediatric dosimetry, we talked about weight basis.
16 It
turns out that smaller children, if you accept
17 the
Hiroshima Nagasaki data that are presumably
18
more at risk, actually get lower effective doses as
19
they decrease in age for a given
20
radiopharmaceutical that is given on a weight
21
basis. So, generally the infants
are getting about
22
half the effective dose of what teenagers and
23
adults are getting when it is given on a weight
24 basis.
25
[Slide]
306
1
CT of the chest, abdomen and pelvis
2
imaging for infection, white cells--very similar
3
dose of gallium because of the longer half-life.
4 One
of the things again here is you have a target
5
organ. Spleen gets radiation
doses for white cells
6 but
when you factor in the exposure in the
7
effective dose calculation it is not a huge risk.
8 You
get to renal infection only and it turns out
9
that nuclear medicine studies are considerably
10
lower than CT.
11
[Slide]
12
Heart and lung, MAA studies for lung
13
perfusion are low. Technetium
agents are
14
considerably lower than thallium.
We can give
15
extremely low dose when we start doing things like
16
cystograms. You know, we are
talking about flying
17
from here to St. Louis, or something.
18
[Slide]
19
Bone and brain, again low doses.
Renal
20
agents, very low doses. Sometimes
we are a little
21
higher than the equivalent X-ray procedure; often
22 we
are lower; often we are in the same range.
23
[Slide]
24
Why we need implementation of the Best
25
Pharmaceuticals for Children Act, we basically have
307
1
been doing this whole thing off-label for children
2
under 18 years, for 30 years within nuclear
3
medicine off-label. There is a
mandate in the Best
4
Pharmaceuticals for Children Act to look at
5
pediatric data to work with drug manufacturers. To
6 do
so, you know, there is some point to this.
You
7 get
safety data our of it. You may get
8
effectiveness data out of it as well.
9
[Slide]
10
As I pointed out, you can have problems if
11 you
don't use radiopharmaceuticals intelligently in
12
very small children because there may be a
13
non-radioactive component that will cause you a
14
problem when you give 50 times as much on a per
15
kilo basis to the patient. So,
there are reasons
16 to
do this.
17
[Slide]
18
Another thing that Dr. Dilsizian alluded
19 to
was the whole concept of what happens when you
20 try
to do research, and the mechanism in a lot of
21 the
basic research in radiopharmaceuticals is the
22
Radioactive Drug Research Committee and it
23
basically states what he went over, that the
24
radiation dose for an adult subject for a single
25
study conducted with one year--and they have limits
308
1
here--and they say that basically from a single
2
dose the whole body, the blood and the lens of the
3 eye
shouldn't get more than 3 rem and other organs
4
shouldn't get more than 5 rem.
5
[Slide]
6
Then, they say under 18 years of age you
7 have
to cut that to 10 percent. First of all,
we
8 are
talking about whole-body dose which is an
9
obsolete concept and, secondly, it doesn't address
10 the
whole problem that there isn't a
11
radiopharmaceutical around that has a target organ
12
that has only 60 percent more than the whole-body
13
dose. They are all 5, 10 times
higher. But when
14 you
factor back in the effective dose this is not a
15
significant factor.
16
[Slide]
17
For example, fluorodeoxyglucose for
18
myocardial viability and for tumor imaging, for
19
standard adult dose you are looking at an effective
20
dose that is above that 0.3 limit.
You are talking
21
about a bladder dose that is way above that. As
22 you
go down, as effective doses drop a bit as the
23
patients get smaller, if you give it on the same
24
weight basis you still have bladder wall doses and
25
effective doses that are way above those limits.
309
1
[Slide]
2
For a whole series of radiopharmaceuticals
3
that are particularly of interest in tumor imaging
4 at
the moment, again everything is higher.
5
Effective doses are higher. Here
you could
6
probably sneak in with carbon-11 methionine but the
7
bladder doses are higher and it is either the
8
kidney or the bladder that is the target organ in
9
each case. But these doses are
factored into the
10
effective dose and they stand out here but it
11
doesn't mean that there is a huge amount of risk
12
associated with them. What this
means is that the
13
whole area of molecular imaging becomes an area
14
that you can't approach in pediatrics.
15
[Slide]
16
Well, could you use a faster camera?
17
Well, there are some faster cameras but if you drop
18 the
dose 50 percent you are still not there.
Can
19 you
reduce the administered activity another 50
20
percent and double the imaging time?
You are still
21 not
there for most of these agents.
22
[Slide]
23
Basically, effective dose takes into
24
account all these risks. We have
regulations for
25
experimental use of radiopharmaceuticals that have
310
1 an
arbitrary standard that no target dose should
2
exceed the whole-body dose by more than 67 percent.
3 We
don't use whole-body absorbed radiation dose
4
anymore and target organ dose for most
5
radiopharmaceuticals is way above that 67 percent.
6
[Slide]
7
With the current RDRC regulations,
8
molecular imaging technology will not be readily
9
available for the study of pediatric
10
life-threatening diseases, including cancer, but
11
also heart disease. With the
current RDRC
12
regulations you can't evaluate new molecular
13
imaging techniques and we should develop an up to
14
date standard based on effective dose that permits
15 the
study of children with life-threatening
16
diseases including cancer and heart disease.
17
[Slide]
18
Finally, I would just like to point out
19
what others have said, that children and adults may
20
differ in the pharmacokinetics of drugs.
Pediatric
21
disease processes are very different from adult
22
disease processes, and I think you have been
23
getting that kind of information all through this.
24
Finally, pediatric data from adequate and
25
well-controlled clinical trials are better than
311
1
extrapolated adult data. Thank
you.
2 DR. CHESNEY: Dr. Cerqueira is our next
3
speaker.
4
DR. CERQUEIRA: Thank you very
much. It
5 is
a pleasure to be here. My name is Manuel
6
Cerqueira. I am a cardiologist at
Georgetown
7
Hospital here, in D.C., and I am representing the
8
American Society of Nuclear Cardiology.
I drove
9
myself here so they are not paying my expenses in
10 any
way. I am a former president of the
American
11
Society of Nuclear Cardiology.
12
The American Society of Nuclear Cardiology
13 is
pleased to comment on pediatric cardiology and
14 the
use of imaging agents. ASNC is a
professional
15
medical society of more than 4,500 members which
16
provides a variety of continuing medical education
17
programs related to nuclear cardiology.
We develop
18
standards and guidelines for training and practice
19
within nuclear cardiology and we promote laboratory
20
accreditation and physician certification in this
21 sub-specialty
to guarantee overall quality.
22
We are principally an advocate for the use
23 of
nuclear cardiology in both adult and pediatric
24
populations. The Society believes
that the medical
25
necessity for the use of cardiac radionuclide
312
1
imaging in children can really be included in four
2
different areas. There is a
handout which is
3
available at the back of the room.
4
These areas include congenital heart
5
disease, including anomalies of the coronary
6
circulation and the presence of cardiac shunts.
7
Anatomic methods of imaging, which have been
8
described by some of the other presenters, do not
9 always identify the physiological
consequences of
10
abnormal communications between the various
11
chambers of the heart. The
radionuclide
12
techniques, however, are able to adequately
13
describe the passage of the radionuclide throughout
14 the
heart and allow detection of these
15
physiological changes that are present.
16
Another area in which we believe there is
17
value for nuclear cardiology in the pediatric
18
population is Kawasaki's disease, which is a
19
systemic vasculitis syndrome occurring in early
20
childhood which affects the coronary arteries and
21 may
cause aneurysms as well as thrombotic
22
occlusions both at the time of the acute disease,
23 as
well as later on in life. Long-term, it
may
24
affect coronary artery blood flow and the degree of
25
perfusion to the myocardium.
Initial obstructive
313
1
lesions may be difficult to evaluate and long-term
2
there may be formation of aneurysms, and optimal
3
management of these patients should include
4
assessment of cardiac function as well as blood
5
flow at a minimum of one-year intervals.
This was
6
published in the guidelines that were put out by
7 the
American College of Cardiology and the American
8
heart Association for the use of cardiac
9
radionuclide imaging.
10
Risks associated with Kawasaki's disease
11
include subsequent stenosis and thrombosis leading
12 to
myocardial infarction as well as sudden death.
13 The
incidence of Kawasaki's disease in the year
14
2000 requiring hospitalization was 4,248 patients.
15 The
median age of these patients at the time of
16
admission was 2 years old. Again,
many of these
17
children will benefit from subsequent long-term
18
following with radionuclide methods.
19
Another area in which radionuclide
20
techniques can be useful in children is the
21
identification of myocardial ischemia in patients
22
with hypertrophic cardiomyopathy.
23
The fourth area is radionuclide
24
ventriculography or MUGAs, as they are commonly
25
called, to monitor children receiving Adriamycin as
314
1
part of therapy for various tumors.
2
Echocardiography and some other techniques can be
3
used but the reproducibility of measurements has
4 not
been as well established and standardized as we
5
have for the use of radionuclide techniques. For
6
that reason, this will provide a very valuable
7
method.
8
Physicians make medical decisions daily in
9 the
diagnosis and treatment of children.
Within
10 the
practice of medicine, medical judgment has
11
supported use of available radiopharmaceuticals in
12 the
treatment of children. The advantages of
using
13
myocardial perfusion imaging in children include,
14 one,
reducing a potential long period of sedation
15
which may be required in some children; two,
16
reduction of overall radiation exposure associated
17
with conventional angiography; and, three,
18
providing a more accurate diagnosis in many cases.
19
Having affirmed a role for cardiac
20
radionuclide imaging in the pediatric population,
21 the
Society wishes to point out that there is a
22
paucity of clinical studies in this area. Clinical
23
guidelines relative to pediatric populations are
24
estimates based on the best available information.
25
General agreement has been achieved to use as low a
315
1
dose of radiation as possible and to carry out the
2
procedures as quickly as possible.
However, we do
3 not
have criteria for identifying appropriate
4
pediatric referrals, nor do criteria exist to
5
determine optimal protocol or technical settings
6 for
the imaging studies.
7
In approaching the pediatric population we
8
know that children are more sensitive to radiation
9
than adults; the number of radionuclide-enhanced
10
phases must be minimized; and automated dose
11
reduction technology exists; and inappropriate
12
referrals can and should be eliminated in many
13
cases.
14
Several questions remain however.
How
15
little radiation is needed to ensure accurate
16
results? How are dosages for various
ages
17
determined or differentiated? How
can the medical
18
profession develop criteria for appropriate
19
pediatric referrals?
20
The American Society of Nuclear Cardiology
21
looks forward to working with the FDA and with
22
other interested parties and stakeholders to
23
resolve these questions. Thank
you for the
24
opportunity to comment on this important matter.
25
DR. CHESNEY: Thank you very
much. Our
316
1
next speaker is Dr. Peter Gardiner from
2
Bristol-Myers Squibb.
3
DR. GARDINER: Dr. Chesney, thank
you. I
4
will actually be very brief and, in the interests
5 of
disclosure, not only did the company pay for my
6
travel but they pay my salary as well.
7
[Slide]
8
We consider ourselves worldwide leaders in
9
cardiovascular imaging research.
Our current
10
product line includes Cardiolite, which is a
11
technetium-labeled radiopharmaceutical, as well as
12
Definity, the ultrasound contrast agent.
You have
13
heard quite a lot already today about both of these
14
agents in their respective technologies and, in the
15
interest of time, I will really just skip to my
16
summary slide in that basically the points that I
17
would have made have been covered already.
18
[Slide]
19
I would just like to point out that
20
nuclear imaging is the only modality approved by
21 FDA
for the assessment of both myocardial perfusion
22 and
function in adults. There is clearly
extensive
23
experience, and you have heard much of that today,
24 in
the adult population. Again, as you have
heard,
25
there is limited and variable experience in the
317
1
pediatric population. There are
certainly some
2
challenges in terms of how to conduct clinical
3
research in that population and that is certainly
4
something that we look to continue to work with the
5
agency and others, whether it is looking for
6
creative ways to actually gather the information
7
that has been discussed today.
8 Perhaps to Dr. Maldonado's point,
9
certainly as a company we very much support the
10
FDA's initiatives to evaluate nuclear cardiac
11
imaging and, in fact, other cardiac imaging
12
modalities in the pediatric population.
So, thank
13
you.
14
DR. CHESNEY: Although we deeply
15
appreciate your brevity, I wonder if you would want
16 to
comment just a little bit more about how you
17
would support pediatric studies or support the
18
issue today, and in what ways or where do you see
19 the
most important need?
20
DR. GARDINER: I think it is
really in
21
many of the topics that have been discussed in
22
terms of defining the appropriate dosing, the
23
appropriate efficacy and the safety of these
24
agents; the challenges, the size of the population
25 and
the variety of the pediatric population, and
318
1
clearly some modalities are more appropriate than
2
others. But I think the areas
that have been
3
touched on are certainly ones that we would see as
4
being important in terms of the questions to
5
address, the questions that are going to be the
6
subject of discussion tomorrow.
7
DR. CHESNEY: Dr. Maldonado and
then Dr.
8
Gorman.
9
DR. MALDONADO: Actually, I wasn't
even
10
aware of the CFR regulation that Dr. Gelfand
11
presented. I see these CFR
regulations that he
12 said
are obsolete and probably might be an
13
impediment for studies, and I can see your lawyers
14
stopping you from doing the studies although they
15 may
be very good. But if there is another
law in
16 the
Code of Federal Regulations with limits, it may
17 be
problematic. I don't know if there is a
18
solution to this because that can be also an
19
impediment. As obsolete as it is,
it may be an
20
impediment and I think that Dr. Nelson may have the
21
answer.
22 DR. GARDINER: It may be the difference
23
between investigational clinical research and
24
clinical practice that may have some bearing on
25
that question.
319
1 DR. CHESNEY: Dr. Nelson, do you want to
2
address this issue?
3
DR. NELSON: Yes, it might depend
on your
4
RDRC but often if it is an intervention that is
5
designed for the possibility of benefit they won't
6
apply those restrictions to it.
If it is an
7
intervention that is of no benefit but for research
8
purposes only, they would apply those restrictions.
9 So,
it depends then on how you construct the trial
10 and
how it is designed. It sets up a whole
other
11 set
of issues you need to address but it is
12
possible to go above that exposure if it offers the
13
possibility of diagnostic benefit.
Then, how much
14
evidence do you need to establish that would then
15 be the question.
16
DR. GARDINER: Dr. Gelfand I
believe would
17
like to make a comment, if he is allowed to.
18
DR. GELFAND: I don't believe that
the
19
RDRC limitations apply to an IND by an
20
investigator, and an investigator by a company.
21 So,
that would not be a problem in that situation.
22 The
second aspect is I have generally found that
23
many, many RDRCs are terrified of going over those
24
limits, regardless of what has just been said about
25
possible benefit to the patient.
320
1
DR. CHESNEY: Dr. Loewke, would
you like
2 to
comment on this issue?
3
DR. LOEWKE: Basically I wanted to
say
4
that 361.1 is non-IND research.
For these
5
products, if they are administered following the
6
regulation, the research can be conducted and they
7 do
not have to submit an IND.
8
DR. CHESNEY: Dr. Gorman, you had
a
9
question?
10
DR. GORMAN: If you are willing to
share
11
this information, was Bristol-Myers Squibb
12
responsible for the two PPSRs to this division?
13
And, if so, what are you intending to study?
14
DR. GARDINER: That is not
something I am
15
prepared to discuss at this point.
16
[Laughter]
17
DR. CHESNEY: Thank you very
much. Our
18
last speaker in the open public hearing is Dr. Jack
19
Rychik from the American Society of
20
Echocardiography.
21
DR. RYCHIK: Thank you. I will just read
22 a
brief statement. Good afternoon. My name is
23
Jack Rychik. I am a pediatric
cardiologist with a
24
specialty interest in pediatric echocardiography.
25
First of all, I would like to congratulate my
321
1
friends and colleagues here in the field of
2
pediatric cardiology who I think have done a superb
3 job
today in really framing this question very
4
well, and I truly enjoyed your presentations today
5 so
thank you.
6
I am a staff member at the Children's
7
Hospital of Philadelphia. I have
served as
8
director of echocardiography at that institution
9
from 1996 to 2003. Currently, I
am the director of
10 the
fetal heart program at Children's Hospital of
11
Philadelphia. I come before this
committee as a
12
representative of the American Society of
13 Echocardiography
and as chair of the Pediatric
14
Council of the American Society of
15
Echocardiography, and they have paid for my Amtrak
16 to
get down here from Philadelphia.
17
The American Society of Echocardiography
18 is
an organization of nearly 9,000 professionals
19
committed to excellence in cardiovascular
20
ultrasound and its application to patient care
21
through education, advocacy, research, innovation
22 and
service to our members and the public at large.
23 As
a member of this organization and a physician
24
with a strong interest in the clinical application
25 of
non-invasive imaging modalities in children, I
322
1 am
here to advocate for the promotion of the safe
2 and
effective use of ultrasonic contrast agents for
3
cardiovascular imaging in children.
4
Ultrasound imaging of the cardiovascular
5
system, or echocardiography, is, as we have heard,
6 the
most commonly used modality for imaging of the
7
cardiovascular system in infants and children. The
8
application of echocardiography in children has
9
over a 30-year track record of safety; is an
10
imaging modality which is highly reproducible with
11
excellent temporal and spatial resolution; provides
12 for
real-time data on both cardiac structure and
13
function; and is a mobile technology which means it
14 can
be performed repeatedly and serially at the
15
patient beside. As such,
echocardiography has
16
become the first-line modality for imaging in
17
children with cardiovascular disease and has grown
18
tremendously in its use, again as we have heard
19
today.
20 Despite its first-line use, however,
there
21 are
still some limitation, primarily related to
22
difficulties in ability to acquire a complete and
23
satisfactory image in every patient in every
24
specific subtype of lesion. Ultrasound
is
25
dissipated within tissue as it travels through long
323
1
distances and is impaired by bony structures and
2
air. These issues become of
primary importance in
3
older or larger patients, however oftentimes
4
acoustic windows, even in small children, can be
5
poor which can lead to poor image resolution. The
6
usual sharp distinction between the borders of
7
blood and tissue can be blurred, thereby making it
8
difficult to reliably measure cavity volumes and
9
wall thicknesses, and consequentially impairing our
10
ability to measure ventricular ejection and wall
11
motion abnormalities.
12
Hence, for our adult cardiology
13
colleagues, the advent of echo contrast agents has
14
been extremely helpful.
Intravenous injection of
15
ultrasound contrast agents has been documented to
16
improve endocardial border delineation.
Contrast
17
enhancement of the blood-tissue boundary has
18
improved assessment of ventricular wall motion,
19
wall thickness, ejection fraction and delineation
20 of
structural abnormalities.
21
Recent experimental results indicate that
22
echo contrast has the potential to provide
23
qualitative and quantitative assessment of
24
myocardial perfusion and coronary blood flow. This
25
would add tremendously to the diagnostic
324
1
capabilities of echocardiography.
As we have
2
heard, the safety profile of the echo contrast
3
agents in adults has been well defined and there
4 are
currently several third generation products
5
approved for us, but its utility and its safety in
6
children has not been defined.
7
We believe that the time has come for
8
children to reap the potential benefits of this
9
form or cardiovascular imaging.
There are some
10
great potentials for its use and let me give you
11
some examples:
12
One can utilize echo contrast for
13
endocardial border, volume and ejection fraction as
14 we
have talked about. It can be used for
15
evaluation of intracardiac shunts and in
16
particular, for example, in cases of patent foramen
17
ovale in patients who have had stroke.
18
It can be used for visualization of
19
complex baffles and channels.
This is specific for
20
congenital heart disease in cases of Mustard or
21
Senning operation for transposition of the great
22
arteries or in the Fontan operation for single
23
ventricle.
24
Contrast agents could potentially be used
25 to
improve visualization of thrombus in venous
325
1
pathways of patients after Fontan operation for
2
single ventricle. Visualization
of thrombus by
3
conventional surface echocardiography is oftentimes
4 a
difficult task due to the scatter created by the
5
synthetic patch material that is used.
Contrast
6
agents may be extremely helpful in reliably
7
identifying thrombus and avoiding the need for
8
further testing, such as transesophageal
9
echocardiography or more invasive modalities such
10 as
angiography.
11
As well, as we have heard, it can be
12
useful in the assessment of coronary artery flow
13 and
myocardial perfusion. Although coronary
14
atherosclerotic disease in infants and children is
15
rare, there is still a great need to reliably
16
assess coronary blood flow in conditions such as
17
congenital coronary anomalies before and after
18
surgery; Kawasaki disease; after arterial switch
19
operation for transposition; after Ross operation
20 in
which coronary re-implantation is performed; for
21
aortic valve disease and after palliation for
22
hypoplastic left heart syndrome in which aortic
23
reconstruction is undertaken and coronary flow
24
potentially impaired.
25
From personal experience, I can tell you
326
1
that I would conservatively estimate that
2
approximately 5-10 percent of our patients coming
3 to
our echo labs at Children's Hospital of
4
Philadelphia could potentially be candidates who
5
could benefit an in incremental manner from the
6
addition of a contrast evaluation.
At our single
7
center, where close to 15,000 echocardiograms are
8
performed each year, this means that approximately
9
1,000 patients per year could potentially benefit
10
from this additional modality.
11
The American Society of Echocardiography
12 has
in the past taken the lead in providing a
13
synthesis of available evidence justifying the
14
adoption of relevant new technologies in the field
15 of
echocardiography. In addition, the ASE
has
16
played a key role in establishing guidelines for
17
training and experience in these various modalities
18 and
uses of echocardiography. An example is
one
19
that Dr. Sable mentioned early, the position paper
20
that was published in 2000 on the use of contrast
21
echocardiography in adults. I can
tell you that an
22
update is currently being planned for utility,
23
again, in adults. The ASE,
therefore, plans to
24
take an active role in the process of promoting the
25
safe use of contrast echo in children.
327
1
With growing interest in the subject, we
2
have formed an ad hoc committee of the Pediatric
3
Council of the American Society of Echo to look
4
specifically at this issue of safety and utility of
5
contrast echo in children. This
committee is
6
comprised of experts in pediatric echocardiography
7 as
well as adult echocardiography, professionals
8 who
can share their knowledge and experience in the
9 use
of contrast agents. It is the desire of
this
10 ad
hoc committee, the Pediatric Council of the ASE
11 and
the ASE as a whole to promote and advocate the
12
expansion of the safe and effective use of contrast
13
echocardiography in children and to develop
14
guidelines for use and training.
15
We look forward to working with the FDA
16 and
acting as a professional resource to them as
17
they move forward in these endeavors. Thank you
18 all
very much.
19
DR. CHESNEY: Thank you.
20
DR. LOEWKE: Dr. Chesney, may I
just make
21 one
clarification, back again to the CFR 361.1 just
22 so
people fully understand that that applies to
23 basic
research. It is not IND drug development
24
clinical trials where you are actually looking to
25
develop and ultimately manufacture a new drug.
328
1
DR. CHESNEY: I think that brings
our
2
afternoon session to a close. On
behalf of the
3
committee and the FDA, I want to thank our speakers
4
enormously for the incredible expertise you
5
brought, and we look forward to working with you
6 tomorrow
to answer the more specific questions.
7
With respect to administrative issues, the
8 van
will leave the hotel tomorrow morning at 7:15
9 to
bring us here. I understand there is a
van to
10
take us back to the hotel now, those of us who are
11 not
going to the Ritz Carlton--
12
[Laughter]
13
Did the FDA want to make any other closing
14
comments today? I guess not. Thank you all very
15
much.
16
[Whereupon, at 5:20 p.m. the proceedings
17
were recessed, to resume at 8:00 a.m., Wednesday,
18
February 4, 2004.]
19 - - -