Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter.
Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 177
7
Outcomes of Written Requests,
Requirements, Studies, and
Labeling Changes
O
ne measure of the accomplishments that have been achieved under
the Best Pharmaceuticals with Children Act (BPCA) and the Pedi-
atric Research Equity Act (PREA) is simply the number of labeling
changes attributed to these policies since they or their predecessor policies
went into effect. From July 1, 1998, through October 25, 2011, the Food
and Drug Administration (FDA) approved 425 labeling changes attributed
to studies or analyses requested under BPCA or required under PREA.1
FDA attributed approximately half (54 percent) of the changes to studies
required under PREA and approximately one third (35 percent) to studies
requested under BPCA; the remaining changes (11 percent) were attributed
to both laws.2 Almost 10 percent (n = 39) of the changes were not based
on data from new pediatric studies. For example, for a 2009 change in
pediatric dosing for zidovudine (Retrovir) for the treatment of HIV infec-
tion, FDA approved the change on the basis of the sponsor’s reanalysis of
existing data (Alivisatos, 2008).
1 One further labeling change was posted for December 2011. As of the end of January
2012, FDA indicated that eight more changes for 2011 were yet to be posted (personal com-
munication, Catherine Lee, Office of Pediatric Therapeutics, FDA, January 27, 2012). The first
labeling change in FDA’s listing (February 10, 1998, for naratriptan [Amerge]) is attributed to
the 1994 Pediatric Rule (personal communication, Robert Nelson, Office of Pediatric Thera-
peutics, FDA, January 10, 2011).
2 Some products have had more than one labeling change (e.g., for different indications or
additional pediatric age groups). A labeling change can involve either the addition of pediatric
information to the existing label for a previously approved product or the new labeling of a
product not previously approved.
177
OCR for page 178
178 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
As described in Appendix A, FDA’s listings of labeling changes related
to BCPA and PREA (and their predecessor policies) are not complete. Spe-
cifically, they do not include changes for biologics that were made prior to
September 27, 2007. FDA could not supply the committee with the miss-
ing information. Thus, the list provided to the committee understates to an
unknown extent the number of labeling changes made as a result of studies
of biologics that were required under PREA or the Pediatric Rule.
Figure 7-1 shows the time trend of labeling changes attributed by FDA
to BPCA and PREA through October 25, 2011. From 1998 through 2004,
the general pattern is one of yearly increases in the number of changes
70
60
50
Label Changes
40
30
20
10
0
Ju 998
99
0
1
2
3
4
5
6
7
8
9
10
ct 11
0
0
0
0
0
0
0
0
0
0
20
20
20
)
.)
19
20
20
20
20
20
20
20
20
20
ly
1
O
m
h
ug
ro
o
(f
hr
(t
BPCA BPCA and PREA Pediatric Rule and PREA
FIGURE 7-1 Changes in drug labeling associated with BPCA, PREA (including
the Pediatric Rule), or both, July 1998 through October 2011. The figure excludes
changes for biologics regulated under the Public Health Service Act that were ap-
Figure 7-1
proved before September 27, 2007. It includes changes for some products (e.g.,
contraceptives) that were excluded from the committee’s analysis as well as one
change that is attributed to the 1994 Pediatric Rule.
SOURCE: Compiled from information periodically updated in an Excel file
downloadable at http://www.accessdata.fda.gov/scripts/sda/sdNavigation.cfm?sd=
labelingdatabase.
OCR for page 179
179
OUTCOMES
attributed to BPCA. Although the pattern after that is uneven, most subse-
quent years show a decrease in changes attributable to BPCA alone. Since
2005, pediatric studies required under PREA have accounted for most
labeling changes. Some of these changes are for products studied from the
outset in at least one pediatric age group.
A few labeling changes that are attributed to PREA might be more ap-
propriately linked to other policies. One such policy is FDA’s unapproved
drugs initiative (FDA, 2006a). That initiative has led to pediatric studies
and the approval of three previously unapproved but long-marketed pan-
creatic enzyme replacement products for use by children and adults (see,
e.g., Giuliano et al., 2011). When it approved these products in 2009 and
2010, FDA imposed a deferred PREA requirement for the development of
a formulation suitable for the youngest and lowest-weight patients (see,
e.g., Beitz, 2009a).
To cite a different example, the labeling of pralidoxime chloride
(Protopam) for pediatric use in 2010 (attributed to PREA, with no new
studies submitted) might be credited to efforts of the child health advocates
and others concerned about children’s emergency access to this treatment
for exposure to organophosphate pesticides and chemicals (e.g., nerve
agents) (Krug et al., 2011). The drug was originally approved in 1964 and
was listed by the National Institute of Child Health and Human Develop-
ment as a priority for a systematic literature review in 2006 (71 FR 23931).
The 2011 to 2016 strategic plan of the Biomedical Advanced Research
and Development Authority (a unit within the Department of Health and
Human Services) includes “supporting the development of medical counter-
measures suitable for use in special populations such as children” (BARDA,
2011, p. 11).
Some pediatric studies conducted and submitted to FDA under BPCA
have not yielded labeling changes. With its list of products with labeling
changes related to BPCA and PREA, FDA also supplied the committee with
a list of 14 active moieties for which requested studies were conducted and
exclusivity was granted without information from the studies being added
to the label. In addition, it is possible that some requests have led to studies
for which FDA neither approved a labeling change nor granted exclusivity.
FDA may deny exclusivity if submitted studies do not meet the terms of
the written request.
Twelve of the 14 grants of exclusivity without labeling changes were
approved before September 2007. For five of these, no information about
the study results is posted. For the remaining seven, short summaries are
available (consistent with the requirements of BPCA of 2002). Some of
these summaries reveal that FDA concluded that no labeling change was
necessary because the studies had not demonstrated efficacy but did not
raise new safety signals. In one case, a summary reveals FDA’s concern that
inclusion of any information from a requested study (of the pharmacoki-
OCR for page 180
180 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
netics of topotecan [Hycamtin]) could be interpreted to imply approval for
pediatric use even if the label noted that safety and efficacy had not been
established (Hirschfeld, 2003).
In addition, in a presentation to the Institute of Medicine (IOM) com-
mittee, FDA staff explained that in the early years of BPCA and PREA (and
the Pediatric Rule), pediatric studies were sometimes submitted in sponsor’s
annual reports (not as part of a New Drug Application [NDA] or Biolog-
ics License Application [BLA]), were not reviewed, and did not lead to
labeling changes (Mathis and Jain, 2011). Moreover, the Center for Drug
Evaluation and Research (CDER) and the Center for Biologics Evaluation
and Research (CBER) traditionally did not amend labels to reflect efficacy
findings that did not support pediatric use. For example, the labeling for
fluconazole (Diflucan) still does not note that the product was studied (by
request) for the treatment of tinea capitis in children and that the studies
found that the product did not work better than an already approved prod-
uct (griseofulvin) (Mathis and Jain, 2011).
In the Food and Drug Administration Amendments Act of 2007
(FDAAA), Congress required that information from studies conducted un-
der PREA and BPCA be incorporated in product labeling, whether or not
the results supported pediatric use or raised new safety signals. Congress
also directed that FDA post the clinical, clinical pharmacology, and statisti-
cal reviews for these studies. Both actions have increased the value to the
public of the studies requested under BPCA or required under PREA.
Nevertheless, two products (bivalirudin [Angiomax] and gatifloxacin
[Zymar]) that were granted exclusivity after the passage of FDAAA did not
have associated labeling changes. The clinical reviews for these products are
posted, but the recommendations on regulatory action and all or part of the
risk-benefit assessments are redacted (Ayache, 2009; Nevitt, 2009), making
it difficult to assess why no labeling change was made.
The rest of this chapter starts with a discussion of written requests and
PREA requirements. Later sections discuss the committee’s assessment of
pediatric studies (as reviewed by FDA staff) and labeling changes. (Appen-
dix A discusses how the committee selected its sample.)
WRITTEN REQUESTS AND PREA REQUIREMENTS
Written Requests
Status of Written Requests
By October 2011, FDA had issued 340 written requests since BPCA
became effective on July 1, 1998.3 Of these requests, FDA had subse-
3Unless otherwise noted, the data discussed in this section are compiled from FDA sources.
The agency lists the moieties for which requests have been made and posts statistics on written
OCR for page 181
181
OUTCOMES
quently granted exclusivity for 176 active moieties (and 185 products).4
Thus, roughly half of the written requests issued to date have led to the
submission of pediatric studies for which exclusivity was granted, although
at least 14 of these did not lead to changes in product labeling. Although
FDA does not identify them, some written requests have been declined by
sponsors and other requests are still open, with studies planned, under way,
or submitted but not yet evaluated. Some sponsors have submitted some
of the requested studies, but an exclusivity determination will not be made
until all the requested studies are submitted and evaluated. More grants of
exclusivity and labeling changes can therefore be expected for previously
issued written requests.
Figure 7-2 shows trends in the issuing of written requests and the grant-
ing of exclusivity. Written requests peaked in 1999 and then dropped off
sharply, with a relative leveling off more recently. Although FDA sometimes
issues written requests for studies that are under way or already completed
(see discussion of nitric oxide in Chapter 6), studies initiated in response
to written requests usually take years to plan, conduct, complete, analyze,
and submit. Thus, the peak in grants of exclusivity comes in 2008, several
years after the peak for written requests.
The early surge in written requests is not surprising, given that neither
incentives nor requirements for pediatric studies had previously been in
place and that a substantial number of already approved drugs had not
been studied in children (see Table 1-1 in Chapter 1). Once FDA had issued
requests for many obvious candidates (e.g., drugs widely used off-label by
children, blockbuster drugs with possible pediatric use, and drugs with
pediatric studies already planned or under way), a subsequent decline is
likewise not surprising. Also, with the passage of time, a reduction in writ-
ten requests could be expected in part because of the growth in the number
of products for which studies had been required under PREA and in part
because of the loss of eligibility for popular older products as existing pat-
ents or other exclusivity expired.
Despite its declining role, BPCA has continuing value because its incen-
tives are not limited to the indications covered by an application for the ap-
proval of a new drug. For example, written requests may take into account
advances in knowledge since a determination about required studies of an
indication was made under PREA. New data may show that a condition
requests and exclusivity at http://www.fda.gov/Drugs/DevelopmentApprovalProcess/Develop
mentResources/ucm049867.htm.
4 Of the 138 earliest written requests that were issued as of September 2000 (Appendix B in
FDA, 2001a), FDA had approved labeling changes for 78 (56 percent) of the active moieties
by October 2011. Some of these written requests may still be open. For example, in 2010, an
FDA advisory committee was asked for its views on the advisability of an amendment to a
2001 written request for studies of the drug sildenafil (Revatio) for treatment of pulmonary
arterial hypertension (Temple, 2010).
OCR for page 182
182 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
100
90
80
70
60
50
40
30
20
10
0
11
98
99
0
1
2
3
4
5
6
07
8
9
10
0
0
0
0
0
0
0
0
0
20
20
20
20
19
19
20
20
20
20
20
20
20
20
Number of exclusivity grants Number of WR issued by FDA
FIGURE 7-2 Number of written requests (WR) issued and number of grants of
exclusivity, by year, July 1998 through September 2011.
SOURCE: Personal communication, Catherine Lee, Office of Pediatric Therapeutics,
Figure 7-2.eps
FDA, November 3, 2011.
is more common in children than previously believed or new research may
suggest a promising new use in children.
Most written requests are proposed by sponsors rather than initiated
by FDA, although FDA may significantly alter those proposals. Overall,
sponsor proposals are associated with approximately 80 percent of the
written requests that FDA had issued as of October 31, 2011. FDA initi-
ated the other 20 percent of requests. By October 2011, FDA had received
approximately 700 sponsor proposals for written requests. The high rate
of requests related to sponsor proposals may explain why almost 90 per-
cent of written requests issued since January 1, 2008, have been accepted
by sponsors (personal communication, Catherine Lee, Office of Pediatric
Therapeutics, FDA, November 30, 2011).
Changes in Written Requests
The committee’s sample of 46 FDA actions included 27 products for
which written requests were issued, including one product for which the
written request covered two indications. Some requests were not amended;
others had four or more amendments. The committee (or consultants) also
reviewed additional written requests for studies of migraine, hypertension,
and gastroesophageal reflux disease (GERD). Most of these additional
requests and any amendments dated from the late 1990s to the mid-2000s
OCR for page 183
183
OUTCOMES
and thus were not reviewed by the Pediatric Review Committee (PeRC),
as provided for in the reauthorization of BPCA in 2007 (see Chapter 3).
The requests that the committee reviewed generally followed a stan-
dard template similar to the one presented in Chapter 3. Depending on the
drug and indication, FDA might also have developed a more substantive,
product-specific template. It has done so, for example, for studies of pedi-
atric hypertension, migraine in adolescents, and GERD. Box 7-1 summa-
rizes major substantive changes in written requests for studies of migraine.
Chapter 6 summarizes the changes in written requests for studies of proton
pump inhibitors to treat GERD in neonates and infants, and Appendix E
provides a more detailed analysis of changes in written requests for studies
of pediatric hypertension.
The substantive details in written requests have tended to become
somewhat more specific over time. Changes in the basic request template
or in specific requests have often
• added precision (e.g., by specifying a period for safety follow-up or
a minimum percentage for age or racial subgroups in a sample);
• required more rigor in trial designs (e.g., by dropping the option for
a trial with no placebo and only alternative doses of the test drug
or by increasing the statistical power of trials to detect a clinically
meaningful effect);
• required more accommodation of developmental variability (e.g.,
by requiring sponsors to try to develop age-appropriate formula-
tions, if needed); or
• incorporated the legislative requirements for greater public access
to information about study results (e.g., by requiring that sponsors
submit NDA supplements to add information from trials—whether
negative or positive—to the label).
Although FDA’s letters that describe amendments in particular written
requests often do not explain the reason for changes, subsequent clinical
reviews of submitted studies suggest that some changes have come after a
sponsor encountered difficulties with conducting the studies as requested.
For example, if studies with an older age group identified serious safety
concerns (e.g., as in studies of the anesthetic desflurane), the agency might
amend a request to eliminate a study with a younger age group. Similarly,
if a sponsor encountered serious difficulties in enrolling children, an amend-
ment might reduce the specified sample size.
Potential of Requests to Generate Useful Information
On the basis of the requests in its sample, the committee concluded that
most written requests had reasonable potential to generate useful informa-
OCR for page 184
184 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
BOX 7-1
Major Amendments to Written Requests for Pediatric
Studies of Drugs for Treatment of Migraine
Zolmitriptan (Zomig)
The written request for zolmitriptan was issued in March 1999. It specified four
studies in adolescents: (1) a short-term safety and tolerability study (sample size
not specified), (2) a pharmacokinetic study (sample size not specified), (3) a well-
powered efficacy study (sample size not specified), and (4) a long-term safety
study with 300 subjects. There were no substantive amendments.
Sumatriptan (Imitrex)
The original written request for sumatriptan was issued in June 1999. It specified
three studies: (1) a single-dose pharmacokinetic study with adolescents (sample
size not specified), (2) a well-powered efficacy study (sample size not specified),
and (3) a long-term safety study with 300 subjects. The amended written request
in May 2000 changed the entry criteria in the efficacy and safety studies from
subjects with an average of one to six migraines per month (with “migraines” de-
fined by the International Headache Society) to subjects with an average of two or
more migraines per month, as requested by the sponsor. The amendments also
allowed the sponsor to compare pharmacokinetic results from adolescents with
those from historical adult controls.
Almotriptan (Axert)
The original written request for almotriptan was issued in October 2001. It speci-
fied three studies: (1) a single-dose pharmacokinetic study with 18 to 50 ado-
lescents, (2) a well-powered efficacy study (sample size not specified), and (3)
a long-term safety study with 300 adolescents. The amended written request in
February 2005 dropped the request for the pharmacokinetic study and decreased
the sample size of the third study to 200 participants. Inclusion criteria for the ef-
ficacy trial did not change between the original and the amended request.
Eletriptan (Relpax)
The written request for eletriptan was issued in July 2004. It specified two stud-
ies: (1) a well-powered efficacy study (specific sample size not specified) and
(2) a long-term safety study with 200 adolescents. There were no amendments.
(Note that in August 2005 this written request was declined by the sponsor and
was referred to the Foundation of the National Institutes of Health in accord with
provisions of BPCA of 2002 [71 FR 6081; see Chapter 3 for a discussion of these
provisions].)
SOURCE: Personal communication, P. Brian Smith, Department of Pediatrics, Duke Univer-
sity Medical Center, June 22, 2011.
OCR for page 185
185
OUTCOMES
tion for clinicians who care for children. These requests include those for
studies of drugs that
• had a new mechanism of action compared with existing medica-
tions approved for pediatric use (e.g., aripiprazole for schizophre-
nia) or were improvements over first-generation drugs in the class
(e.g., many second- or later-generation antibiotics);
• offered a possible treatment for a serious or life-threatening condi-
tion for which few or no treatments had been demonstrated to be
safe and effective (e.g., irinotecan hydrochloride for treatment of
solid tumors and bisphosphonates for treatment of osteogenesis
imperfecta);
• were commonly used off-label with no controlled studies of phar-
macokinetics, dosing, safety, or efficacy (e.g., antibiotics for various
infections and proton pump inhibitors for GERD);
• lacked important information on safety (e.g., desflurane for the
maintenance of anesthesia in nonintubated children);
• would potentially allow safe and effective use of the drug in a new
pediatric population if a new formulation was developed (e.g.,
terbinafine hydrochloride for tinea capitis); or
• had new safety concerns suggested by new data (e.g., remifentanyl
hydrochloride for anesthesia).
Some requests had elements that, in the committee’s judgment, could
limit the potential of the requested studies to yield strong information to
guide the care of children. Box 7-2 provides examples.
To focus on one therapeutic area, details of the written requests for
pediatric studies of a number of drugs used to treat hypertension in adults—
and the resulting trials—have been criticized for a number of reasons. An
analysis by Benjamin and colleagues reached the following conclusions:
Successful studies showed large differences in doses, with little or no over-
lap between low, medium, and high doses; failed trials used narrow dose
ranges with considerable overlap. Successful trials also provided pediatric
formulations and used reduction in diastolic, not systolic, blood pressure
as the primary endpoint. Careful attention to pediatric pharmacology and
selection of primary endpoints can improve trial performance. We found
poor dose selection, lack of acknowledgment of differences between adult
and pediatric populations, and lack of pediatric formulations to be associ-
ated with failures. (Benjamin et al., 2008, p. 834)
The authors have also suggested that for these trials feasibility was
more important to the sponsors than strong trial design, particularly since
exclusivity can be granted regardless of whether studies demonstrate ef-
OCR for page 186
186 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
BOX 7-2
Elements in Written Requests That Could Limit the
Potential of Studies to Yield Useful Information
Toxicity profile of drug. One request involved a drug (stavudine [Zerit]) with a
known toxicity profile that made its use for HIV-exposed or -infected neonates or
young infants unlikely given available alternative treatments.
Timing of pharmacokinetic study. For a drug (levalbuterol tartrate [Xopenex HFA])
to treat asthma, a request did not specify that pharmacokinetic studies be per-
formed early enough to guide the efficacy and safety trial.
Failure to request pharmacokinetic study. FDA requested safety and efficacy
studies but not a pharmacokinetic study for a drug to treat asthma (albuterol
sulfate inhalation aerosol [Ventolin HFA]) in children ages birth up to 2 years
and 2 years up to 4 years. The clinical reviewer concluded that the studies did
not show efficacy and that the dose chosen for the studies might not have been
optimal (Wang, 2008).
Failure to request relevant studies. For a drug (inhaled nitric oxide [INOmax]) to
prevent bronchopulmonary dysplasia in newborns, FDA did not request pharma-
cokinetic data, despite the diverse gestational ages of infants to be studied, and
did not specify safety endpoints other than those associated with prematurity.
Need to tailor studies to age groups. As specified in a request for studies of
esomeprazole magnesium [Nexium]), a study design with a run-in treatment stage
followed by randomized, placebo-controlled withdrawal, although appropriate for
older age groups, may not be optimal in a study of the treatment of GERD in in-
fants. If the initial run-in phase effectively heals erosive esophagitis, withdrawal is
not likely to show a significant difference between the placebo treatment and the
proton pump inhibitor treatment. An amendment to the written request eliminated
the efficacy study, although the drug is widely used by infants.
Selection of endpoints inappropriate for age group. In a study of a drug (salme-
terol xinafoate/fluticasone propionate [Advair]) for treatment of asthma in children
ages 4 to 11 years, the requested endpoint of forced expiratory volume in 1 sec-
ond (FEV1) did not adequately recognize the inability of the youngest children to
perform the necessary breathing maneuvers.
Insufficient definition or scope of intervention. For a requested study of an anes-
thetic agent (desflurane [Suprane]), the request specified management by either
a face mask or a laryngeal mask airway device, leaving the choice of approach to
the anesthesiologist rather than defined as part of the trial design.
ficacy. As outlined in Appendix E, FDA has amended the template for
requests for these studies by specifying stronger trial design options, which
are more likely to provide useful information about efficacy.
Chapter 6 notes that some requests for studies with neonates yielded
little information, in part because of uncertainty about the nature and
OCR for page 187
187
OUTCOMES
means of assessment of the condition in children. At the same time, the
chapter noted the shortage of studies of drugs often prescribed off-label
for neonates. In addition, certain categories of requests, for example, re-
peated requests for studies of similar (“me too”) drugs, were generally not
compelling, although studies of such drugs might still provide some useful
pharmacokinetic and dosing information. For the variety of drugs used for
the treatment of AIDS, often in new combinations, pediatric studies (some-
times requested, sometimes required) provide reassurance about safety and
appropriate dosing in children.
Except for sponsors, who may propose pediatric study requests, no or-
ganized process currently exists to obtain broader public input. Moreover,
neither Congress nor FDA has spelled out the criteria to be considered in
assessments of the health benefits of a request, and written requests usually
contain little or nothing by way of rationale for the request. It is not clear
that the magnitude or importance of the expected benefit matters. Particu-
larly for requests that follow several other requests for studies of drugs in
the same class, especially when the initially requested studies do not find
efficacy, it would be in the public interest for FDA to discuss whether the
expected benefit of a drug proposed for a written request would exceed that
of existing therapies for all or some subgroups of children (e.g., because the
drug allowed less frequent dosing or had a safer formulation).
In addition, although it did not consider alternatives to the current
period of exclusivity, the committee was troubled by the disparity in effort
associated with more demanding requested studies that lead to the same
reward as less demanding studies. Six months of exclusivity is the reward
whether the requested studies primarily involve small pharmacokinetic,
pharmacodynamic, and safety studies or larger, well-controlled studies of
safety and efficacy. These concerns do not imply the need to change the
current policy that allows the granting of exclusivity for both studies with
positive results and studies with negative results, as long as they meet the
terms of the written request.
The committee concluded that, in general, changes in the basic tem-
plate for written requests and the amendments to specific written requests
have improved requests. Although the committee examined few written
requests that were issued after the PeRC initiated its reviews and many of
these requests are not yet public, it expects that the additional pediatric and
methodologic expertise provided by these reviews are improving the quality
of requests (and subsequent studies). As described in Chapter 6, the lack of
availability of expertise in neonatal research and clinical care, including for
the specification of appropriate written requests for studies of the youngest
pediatric patients, remains a concern.
OCR for page 196
196 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
In a few cases, however, problems appeared to arise as much or more
from the execution of requested or required studies as from the specifica-
tions for the studies. Box 7-3 provides examples of aspects of study plan-
ning or execution that may have limited the usefulness of the information
submitted. (Chapter 6 discussed problems with written requests and studies
that stem from uncertainties about the nature of GERD and bacterial con-
junctivitis in neonates.)
In one instance, the committee found unusual labeling language that
conveyed FDA’s dissatisfaction with the sponsor’s design and conduct of a
study of a drug to reduce mortality and morbidity in neonates and infants
with cyanotic congenital heart disease palliated with a systemic artery-to-
pulmonary artery shunt. The reviewer particularly cited deficiencies in the
sponsor’s approach to selecting the dose for study, which the reviewer and
others at FDA concluded was too low to have the desired antiplatelet effect
(Rose, 2010; Grant, 2011). Other problems included the concomitant use
of aspirin and the late initiation of therapy. After noting the study results
and these likely contributing factors, the label goes on to state that “[i]t
cannot be ruled out that a trial with a different design would demonstrate
a clinical benefit in this patient population” (BMS/SPP, 2011).
In addition to problems with various aspects of study design, studies
may not be completed to the standard desired—or at all—because sponsors
encounter difficulties with enrollment of sufficient numbers of children,
despite reasonable efforts. This challenge was highlighted in Chapter 1. In
the committee’s sample, one example of such enrollment problems involved
a study comparing leflunomide (Arava) to methotrexate for treatment of
juvenile rheumatoid arthritis. Enrollment shortfalls prompted the amend-
ment of the written request to specify a superiority trial with 94 participants
instead of the originally requested noninferiority or equivalence trial with
120 participants (Yancey, 2004). The eventual findings for the randomized,
double-blind trial favored the active comparator.
Another example of enrollment difficulties involved a combined phar-
macokinetic, safety, and efficacy trial testing pegfilgrastim (Neulasta) to
reduce episodes of febrile neutropenia in children with sarcoma. Of 50
eligible study sites, only 18 agreed to participate in the trial; of these, only
10 enrolled any children (Summers, 2008). A likely contributing factor was
that pefilgrastim was already marketed and available, so parents may have
been reluctant to have their child participate in a trial comparing this drug,
which involved a single injection, to neupogen, which required daily injec-
tions. Some pharmacokinetic information was added to the labeling. FDA
judged the sponsor to have made diligent effort to fulfill PREA requirements
and noted that the Children’s Oncology Group (COG; which is centrally
involved in the conduct of most pediatric cancer trials in the United States)
had indicated to the sponsor that the conduct of an additional efficacy study
OCR for page 197
197
OUTCOMES
of the drug was not a priority. For another pediatric cancer drug study,
which was ended early for lack of evidence of test drug activity, a differ-
ent reviewer noted that because relatively few children are diagnosed with
cancer compared with the number of adults, “COG prioritizes its trials to
study the most promising agents first” (Honig, 2002, unpaged).
PEDIATRIC STUDIES AND CHANGES IN LABELING
Types of Labeling Changes
All but one of the products in the committee’s sample had labeling
changes that resulted from the studies conducted under BPCA or PREA.
Three labeling changes involved one product. Of the 45 labeling changes
in the sample, 17 involved the extension of age limits for an indication
that had already been approved in adults or another pediatric age group.
Another 10 changes involved approval of a new product with pediatric
labeling or a new indication that had not previously been approved for
adults. Thus, 60 percent of labeling changes in the sample resulted from
analyses that found efficacy and safety. (A few changes occurred without
requirements for efficacy studies.) The addition of an indication to labeling
was generally accompanied by information on dosing, pharmacokinetics,
and safety. As described in Chapter 6, five products studied with substantial
numbers of neonates did not have a labeling change that incorporated any
information from these studies.
For the labeling changes that did not involve the addition or expansion
of a pediatric indication, changes generally included the addition of some
information about safety and pharmacokinetics. For changes that followed
from studies that did not show efficacy, the presentation of that information
varied. Some labels added statements to the effect that use of the product
was not indicated or recommended, whereas most stated that safety and
efficacy had not been established for all or some pediatric age groups.
The latter language is rather imprecise, in itself not making clear whether
studies have not been conducted and submitted to FDA or whether studies
have been submitted and did not show safety and efficacy. Additional text
in the label may clarify the situation, but the key summary sentence is still
ambiguous.
FDA has not evaluated information added to the label as a result of
studies required under PREA, but FDA staff have published two articles
that have reviewed labeling changes associated with BPCA (Roberts et al.,
2003; Rodriguez et al., 2008). The most recent article presents data from
an analysis of labeling changes from July 1998 through October 2005
(Rodriguez et al., 2008). For the 108 drugs with labeling changes resulting
from studies conducted under BPCA, 77 changes extended the age limits
OCR for page 198
198 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
BOX 7-3
Aspects of Studies as Planned or Executed That May
Have Limited the Usefulness of Information Submitted
Questions about participant characteristics and dosing issues. For the pivotal
study of omalizumab (Xolair) for the treatment of moderate to severe persistent
asthma in children ages 6 to 11 years (inclusive), the children enrolled in studies
had, on average, normal pulmonary function (determined from the forced expira-
tory volume in 1 second [FEV1]). As summarized by the clinical reviewer, an FDA
advisory committee was concerned that “the applicants had not studied patients
for whom the drug is intended, namely the most severe asthmatic patients who
are not responding to alternative therapies” and was “very concerned that the
applicants had not explored any dose ranging” for this age group (Starke, 2009,
p. 95). Taking the results for all efficacy endpoints and safety data into account,
the advisory committee concluded that the risk-benefit assessment did not favor
approval of the product. Almost all of the overview of the risk-benefit section of
the review was redacted.
Questions about adequacy of dosing. In requested studies of leflunomide (Arava),
children with juvenile rheumatoid arthritis receiving this drug showed less im-
provement than children in the active comparator (methotrexate) control arm of
the trial (68 versus 89 percent) (Yancey, 2004). On the basis of questions about
the adequacy of the dosing used for lower-weight children, the drug was labeled
in 2004 as having not been fully evaluated. The label included information about
pharmacokinetics and safety and a summary of the trial results.
Problems with data quality. In analyzing a submission of studies of zolmitriptan
(Zomig) for treatment of migraine in adolescents, the statistical reviewer described
“extreme difficulties in analyzing the data due to poor data quality, missing infor-
mation (information not entered in the data by the sponsor), poor organization of
the data, and various errors” (Yan, 2008, p. 4). The reviewer also noted problems
with poor patient compliance and with the deviations from the statistical analysis
plan in the sponsor’s imputation of efficacy values. The reviewer concluded that
no statistically significant difference existed between the test drug and the placebo
for either 1-hour headache response or 2-hour sustained headache response.
Inadequate enrollment of relevant age groups. One of two studies described in
the written request for propofol (Diprivan), which anesthesiologists use in all age
groups, was a randomized, open-label trial comparing 1 percent propofol versus
standard anesthetic technique for induction and maintenance of general anes-
thesia in children from birth to 3 years of age (Raczkowski, 1999). The request
specified “substantial representation” of three age groups, including children from
birth to 2 months of age. In reviewing the study as conducted, the clinical reviewer
concluded that “the only age group not adequately covered was the birth to <2
month age group” (Hartwell, 2000, p. 66); only one neonate was in the propofol
arm, whereas four were in the standard anesthetic arm. The labeling states that
the product is not recommended for maintenance of anesthesia in this age group
because safety and effectiveness have not been established.
OCR for page 199
199
OUTCOMES
Safety concern not addressed. In a study of sotalol (Betapace) for treatment of
arrythmias, the reviewer noted higher peak concentrations of the drug in neonates
and infants than older children and attributed some of the difference to differences
in renal function (Karkowsky, 2000). The studies enrolled fewer neonates than
planned (6 rather than 20). In general, the reviewer notes that the studies provided
no information about dosing of children who have diminished renal function.
Questions about pediatric subgroup. Guanfacine hydrochloride (Intuniv) was stud-
ied for treatment of attention deficit hyperactivity disorder in children ages 6 to
17 years of age. For the 13- to 17-year-old age group, the studies did not find a
statistically significant different result for the study drug than for the placebo. The
clinical reviewer noted that the sponsor used fixed rather than flexible, weight-
based doses in the trials and concluded that “it is highly likely that one contributing
factor [to the study results] was the lower serum guanfacine exposures observed
in the Intuniv clinical program” (Levin, 2007, p. 43) The product was approved for
the entire age group with a weight-based dosing regimen, labeling that described
the study results, and a postmarket commitment for an additional study with ado-
lescents to confirm efficacy.
Weak trial design. Etodolac (Lodine XL) was studied for treatment of juvenile
rheumatoid arthritis in children 6 to 12 and 12 to 16 years of age in an open-label
uncontrolled trial to assess pharmacokinetics, safety, and efficacy. The clinical
reviewer concluded that “especially without some arm for comparison, it is difficult
to understand how any of this information can be placed into a proper context
short of historical controls either in an adult or pediatric population” (Witter, 1999,
p. 17). The pharmacometrics reviewer concluded “that no statistical comparison
can be made on pediatric and adult PK [pharmacokinetics] based on the studies
submitted” (Wang, 2000, p. 14). The pediatric use section of the label approved
in 2000 read, “If a decision is made to use Lodine XL for patients six years of
age or older, as with other NSAIDs [nonsteroidal anti-inflammatory drugs], such
patients should be monitored periodically” (http://www.accessdata.fda.gov/drug
satfda_docs/nda/2000/20-584S005_Lodine_prntlbl.pdf). By 2005, however, that
section of the label had been amended to state that safety and effectiveness in
patients 6 to 16 years of age were supported by extrapolation from adult trials and
by safety, pharmacokinetic, and efficacy data from an open-label trial with children
in that age group (http://www.accessdata.fda.gov/drugsatfda_docs/label/2005/02
0584s004,006,007lbl.pdf). It is not clear what prompted that change, which is not
recorded in FDA’s overview table of BPCA- and PREA-related labeling changes.
Problems with administration of test and control drugs. In a trial of fluticasone
inhalation aerosol (Flovent) involving children 6 to 23 and 24 to 47 months of age,
the report for the pharmacokinetic study revealed detectable levels of the study
drug in some participants in the placebo control arm. Further investigation also
showed that some participants in the active drug arm had no detectable levels of
the test drug. The reviewer concluded that “the studies could not be meaningfully
interpreted, and no conclusions may be drawn regarding either efficacy or safety
from the clinical studies” (Starke, 2003, p. 5).
OCR for page 200
200 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
for an approved indication; 19 changes added information about lack of
efficacy. Of the other changes, 23 involved dosing or pharmacokinetic
information, 34 involved safety, and 12 described a new pediatric formula-
tion. The discussion and examples focused on the changes related to new
information on pharmacokinetics or dosing.
The analysis by Rodriguez and colleagues (2008) stressed the impor-
tance of studies requested under BPCA to generate knowledge important
for safe and effective dosing. It noted that the results of studies were not
necessarily predictable on the basis of weight differences and data from
adults.
For the sample that the committee examined, Box 7-4 presents exam-
ples of informative changes to labeling resulting from requested or required
pediatric studies. Most changes supported the use of the drug with children
but some did not. Some changes reflected safety findings for children that
differed from findings for adults.
As discussed in Chapter 5, FDA sometimes requests only pharmacoki-
netic and safety information and expects to extrapolate efficacy on the basis
BOX 7-4
Examples of Informative Labeling Changes
Vinorelbine tartrate injection (Navelbine) (2002). Requested studies did not show
activity of the drug against recurrent malignant solid tumors, which is important
information for clinicians. Labeling noted that toxicities were similar to those in
adults. Recent studies suggest that the drug may have value against other can-
cers; the clinical review is not publicly posted by FDA but includes pharmacoki-
netic data that could be useful to investigators.
Remifentanil (Ultiva) (2004). Requested studies with infants from birth to 2 months
of age showed high variability in the drug’s phamacokinetics in neonates, which
led FDA to recommend careful titration of individual doses. Given concerns about
possible negative neurodevelopmental effects of anesthetics in young children,
the information about an ultra-short-acting opioid without suspected neurotoxic
effects is valuable.
Desflurane (Suprane) (2006). Requested studies clarified the risks from use of
this anesthetic, which is approved for maintenance of anesthesia in pediatric
patients with intubation and after induction with another agent. The studies led to
stronger safety information in the labeling stating that the product is not approved
for maintenance of anesthesia in nonintubated children. The warning now appears
at the front of the labeling, a change made possible by the switch in 2010 to the
structured labeling format that FDA has been phasing in since 2006.
Aripiprazole (Abilify) (2007, 2008). This drug has a different mechanism of action
than other antipsychotic medications available at the time that written requests
OCR for page 201
201
OUTCOMES
of efficacy studies with adults, absent unexpected safety findings. In the case
of sotalol (Betapace), FDA requested pharmacokinetic, pharmacodynamic,
and safety information to guide pediatric use but did not request efficacy
studies and did not extrapolate safety and efficacy from adults. The labeling
for the product notes that safety and efficacy have not been established, but
it includes pediatric dosing and pharmacokinetic information (for children
more than 2 years of age and children younger than that) based on two
requested studies (FDA, 2001b).
Box 7-5 presents examples of committee concerns about the labeling
changes that followed pediatric studies. Most involve how labels presented
information about pediatric studies that did not demonstrate efficacy.
In some cases, labeling seemed to convey contradictory information,
as illustrated in the first example in Box 7-5. Such labeling may stem from
the dilemma faced by FDA in labeling of products that it expects may have
continued off-label use, despite studies that do not demonstrate efficacy.
It may also stem from FDA concerns about the shortcomings of efficacy
studies (e.g., enrollment problems) that might have limited the possibility
were issued. Studies led to labeling for pediatric use for the treatment of schizo-
phrenia and mania associated with bipolar disorder. Under PREA, the drug has
also been approved for treatment of irritability associated with autism.
Adalimumab (Humira) (2008). Required studies of children with juvenile idiopathic
arthritis demonstrated efficacy. They also found several safety signals that had not
been identified in adults, including elevations of creatine phosphokinase, a higher
rate of immunogenicity, and a higher rate of nonserious hypersensitivity reactions.
Tenofovir disoproxil fumarate (Viread) (2010). Several factors complicated the
required study of this drug’s efficacy for treatment of HIV infection in adolescents,
but the pharmacokinetic and safety data combined with adult data allowed the
extrapolation of efficacy to this pediatric age group. Although the drug has been
used in adolescents on the basis of a favorable toxicity profile in adults and
pharmacokinetics that allow once-a-day dosing, the studies provided reassur-
ance for such use on the basis of the safety and pharmacokinetic results. (Based
on these studies and studies with adults suggesting adverse bone effects, FDA
required a postmarket clinical trial to further investigate the drug’s effects on bone
in pediatric patients.)
Candesartan (Atacand) (2009). Requested studies of children ages 1 to 17 years
showed safety and efficacy of the drug for the treatment of hypertension. The
pharmacokinetic data provided the basis for dosing recommendations for children
ages 1 up to 6 years and children ages 6 up to 17 years. Other data for children
less than 1 year of age led FDA to drop the requested study with children in this
age group and to specifically warn in the label that the product must not be used
by this age group.
OCR for page 202
202 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
BOX 7-5
Concerns About Clarity of Labeling Changes
Information appears to be contradictory. The labeling for zoledronic acid (Zometa)
states that it is not indicated for use in children but also states (as in the previ-
ous label) that “[b]ecause of long-term retention in bone, Zometa should only be
used in children if the potential benefit outweighs the potential risk” (NPC, 2012,
unpaged). That advice applies to the use of any medication by children or adults.
Lack of efficacy is downplayed. The pediatric use section of the labeling for bus-
pirone (Buspar) does not state that efficacy has not been demonstrated. Rather, it
describes safety and pharmacokinetic data from two placebo-controlled trials and
that the trials found “no significant differences between buspirone and placebo
with regard to the symptoms of GAD [generalized anxiety disorder] following
doses recommended for the treatment of GAD in adults” (BMS, 2010, p. 11).
Lack of efficacy in an age group not explicitly stated. The pediatric use section of
the labeling for olmesartan (Benicar) notes that it was studied in children ages 1 to
16 years and that it was generally well tolerated and had an adverse experience
profile similar to that for adults. It does not explicitly state that studies did not show
efficacy in the younger age cohort studied (ages 1 to 5 years).
Lack of advantage of higher dose could have been clearer. In the highlights sec-
tion of prescribing information for aripiprazole (Abilify) for treatment of schizophre-
nia in adolescents, the dosing chart lists a maximum dose without noting that it
was not shown to be more effective than the recommended dose. The discussion
of dosing later in the labeling notes this. The discussion of adverse events does
not discuss the effects of the higher dose on adverse events (e.g., somnolence
and extrapyramidal effects).
Relevant data about dosing were not highlighted. Studies of mometasone furoate
(Asmanex) yielded convincing data that twice-a-day administration of the 110-mg
dose to children ages 4 to 11 years was more efficacious than once-a-day dosing
for severe asthma. These data do not have a prominent place on the label.
Placement of information is unexpected. Data on the pharmacokinetics of irino-
tecan hydrochloride (Camptosar) are included in the precautions section of the
labeling rather than in the section on clinical pharmacology. The latter section
does not provide a cross-reference to the precautions section, which begins by
explaining that studies had not demonstrated effectiveness for the treatment of
solid tumors in pediatric patients.
of finding statistically significant positive findings for a drug that is, in fact,
efficacious.
Aside from specific language in labeling, another concern stems from
the incomplete transition from the old labeling format to a new format,
which was introduced in 2006, as described in Chapter 3. Of the 45
OCR for page 203
203
OUTCOMES
labeling changes in the committee’s sample, the labeling for 15 products
remained in the old format at the time that it was consulted. That is, FDA
has not required the sponsor to revise the label to meet current standards
for new labeling that, in particular, requires an initial highlights section that
summarizes key information about approved uses and age groups (ideally),
warnings, and use by special populations.
Reformatting can significantly clarify information. For example, when
the labeling for desflurane (Suprane) was reformatted, the highlights seg-
ment on pediatric use stated that for safety reasons the product was not rec-
ommended for induction of anesthesia in or for maintenance of anesthesia
in nonintubated children. In the old format, the indications and usage sec-
tion did not explicitly state that it was not recommended for the latter use.
Although the committee was not asked to evaluate the efforts by FDA
or others to disseminate information from pediatric studies and labeling
changes, it recognized that these efforts are important. The committee was
aware that clinicians often do not consult a product’s labeling. They instead
rely on intermediary sources, as described in Appendix B. Nonetheless, to
the extent that labels still in the old format are consulted by clinicians or
others, including parents searching the Internet for additional information
on a child’s treatment, the format hinders the identification of key informa-
tion about efficacy and safety. To acknowledge the importance of getting
information to clinicians, the committee commissioned the background
paper that appears in Appendix B. It underscores the challenges of getting
up-to-date information to clinicians who care for children.
CONCLUSIONS
Pediatric studies conducted under BPCA and PREA are yielding im-
portant information to guide clinical care for children. The information
generated varies by medical condition and age group. As discussed in
Chapter 6, studies with neonates are a particular challenge. Findings from
pediatric studies sometimes support and sometimes run counter to expecta-
tions about the efficacy, safety, and pharmacokinetics of a drug in children
of different ages.
Some studies requested under BPCA or required under PREA do not
achieve their full potential. Reasons vary. Some problems stem from the use
of weak study designs and underpowered samples, the lack of dose-ranging
studies to guide efficacy trials, and the omission of relevant information
from labels. Other problems stem from sponsor difficulties enrolling suf-
ficient numbers of children in clinical trials. One persistent need is for strict
and consistent attention by FDA, sponsors, and investigators to dose selec-
tion for evaluation in pediatric drug studies.
OCR for page 204
204 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN
The committee concluded that the steps that Congress and FDA have
initiated appear to be improving the quality of requests and requirements
for pediatric studies. These steps include increased review by pediatric ex-
perts, increased specificity in the template for written requests and amend-
ments to specific written requests, and earlier specificity about deferred
studies required under PREA. In addition, as suggested in Chapter 4, FDA
could more clearly articulate the health benefits expected of requested
studies so that children do not participate in research of minimal value.
Chapter 5 suggested similar articulation of the rationales for the acceptance
of extrapolation and the use of alternative endpoints.
Although FDA now monitors, analyzes, and reports more information
about the status of studies (e.g., required studies that are pending or delayed
and clinical areas represented by written requests), some information is not
readily available. If FDA creates a formal system for tracking pediatric drug
applications through the submission and review process as recommended
by GAO, it would be helpful for the system to track pediatric studies by
age group, including neonates specifically.
The organization and highlighting of key information in the current
structured labeling format are substantial improvements over the previous
version. Transitions to the new format provide FDA with the opportunity
to clarify inadequately described, ambiguous, or contradictory information
in older labeling.
The committee recognizes that FDA faces some dilemmas when submit-
ted studies do not show efficacy but the agency expects that physicians will
continue to use the drug off-label. If the agency includes pharmacokinetic
and safety data in labeling, it is important that the label be clear that the
provision of information about pediatric dosing does not mean that the
product is recommended for pediatric use.
FDA likewise faces a dilemma when off-label use of a medication is
common but controlled studies of efficacy are not or may not be feasible.
The agency may have to weigh competing risks. If it requests or requires
sponsors to conduct only pharmacokinetic and other studies to guide dos-
ing decisions, it risks encouraging increased use of a product that has not
been demonstrated to be effective. If it does not seek this information in the
absence of more comprehensive investigations, it leaves physicians without
data that could potentially reduce the harm or increase the benefit from
off-label use.
In the future, FDA’s efforts to strengthen regulatory science (e.g., meth-
ods for evaluating drugs and biologics) should support further improve-
ments as should a number of activities the agency has undertaken to analyze
specific challenges in pediatric trial design and analysis and propose innova-
tive strategies to meet these challenges. Examples include the analyses of
pediatric hypertension trials described in this chapter and the assessment
OCR for page 205
205
OUTCOMES
of pediatric studies of analgesic medication and other pain prevention and
alleviation strategies described in Chapter 6. To improve pediatric studies
of drugs and biologics and their evaluation, it is important for FDA to
continue and expand initiatives to strengthen the science base for its work,
analyze shortcomings in pediatric studies, and develop innovative strategies
to meet the specific challenges of pediatric trials.
OCR for page 206
