Discussion
Thoughtful drug development and inclusion of pediatric patients in
trials is critical to public health. Over the years, because of rarity
of pediatric cancer, pediatric cancer drug development has relied on
BPCA rather than PREA to promote the development and inclusion of
pediatric information into the drug label. Under FDA Reauthorization Act
(FDARA) of 2017 and RACE for Children Act [25], pediatric studies
for oncology products will no longer be exempt based on orphan
designation, and the pediatric studies will be required for drugs
directed at molecular targets observed in pediatric cancers [26].
Original marketing applications in the US for certain adult oncology
drugs that are submitted on or after August 18, 2020 would be required
to be preceded by plans for pediatric cancer investigations. With the
change, pediatric oncology drug development should generally be
coordinated with oncology drug development for adults, as part of an
overall drug development plan.
Before the changes of FDARA and RACE Act take place, we conducted
research on the written requests requested through BPCA over 19-year
period. This would serve the purposes to study what has been requested
over the years and what have been completed in response to the
requirements. We consider this to be the anchor of pediatric cancer
development for current stage and can potentially provide insight on how
pediatric cancer drug development would change for the future years.
Since 2001, FDA has issued 42 written requests in the area of oncology.
Among those, 18 drugs have been granted additional pediatric
exclusivity, and 4 received approval for use in specific age groups with
indication in drug label. Most oncology pediatric trials have been
designed to evaluate both safety and efficacy. The majority of pediatric
trials in response to WRs included monotherapy activity evaluation.
Request over the years has seen that more recent pediatric trials
changed to request whether phase 2 studies should be conducted are based
on results from the phase 1 study. Most pediatric trials are single arm
design, especially for hematology indications. Although controlled
trials are preferred, they are difficult to conduct in a pediatric
population for ethical reasons. Most efficacy endpoints involved overall
response rate and assessed by investigators. Only a few assessed the
response rates by an independent review committee (IRC).
More pediatric studies are still in urgent need. Accrual challenge is
the main reason for study termination. Many studies terminated because
of difficulties with enrollment. On the other hand, many completed
trials had the problem of over-accrual. Most of the completed pediatric
trials over the past 19 years have led to conclusions that cancer drugs
used in adult populations do not work as well in children. Since
children are a vulnerable population, it is challenging to enroll them
in clinical trials where the effects of these developmental drugs are
unclear. Given the uncertainty in drug efficacy and for the protection
of pediatric patients, the implementation of futility criteria into the
trial design would be recommended in future pediatric trials. Studies
should stop early if evidence of no efficacy is observed to minimize the
harm to pediatric patients.
Because of many challenges in pediatric cancer drug development,
innovative analysis should be explored to improve the efficiency of the
pediatric trials. A Bayesian approach is a flexible tool that could be
used in pediatric trials to sequentially monitor efficacy and futility
as data accumulate [27]. This approach provides an option to stop
trials for efficacy or futility if enough evidence is observed, and thus
with the advantages of possibly requiring fewer patients and shorter
trials. Bayesian approach in pediatric trials can formally incorporate
prior information from adults, older age groups, and other external
sources if appropriate and quantify the uncertainty. To overcome the
limitation of a small population and limited opportunities for
extrapolation, an innovative approach using Bayesian strategy to allow
more flexibility in statistical design for future pediatric trials
should be considered.
Identifying an optimal starting dose remains important for successful
pediatric oncology drug development. Modeling and simulation is a
powerful tool to inform selection of safe and efficacious doses as early
as possible in the drug development process and to avoid overexposing
children to subtherapeutic doses. When the disease pathobiology and the
mechanism of action for drug would not differ by age, a starting dose
may be selected by targeting similar exposure to the adult therapeutic
dose without the need to evaluate multiple dose levels. When similarity
in disease between pediatric and adults cannot be assumed, a more
extensive PK and dose finding study may be required. While traditional
designs like the 3+3 and rolling 6 are commonly used for dose
escalation, Bayesian designs like the continual reassessment method or
adaptive logistic regression design could be preferable, especially for
targeted therapies with adult’s safety profile well characterized.
We believe this research provides the sponsors, patient advocate groups,
academic groups, and treating physicians on the current landscapes of
pediatric studies in support of labeling changes and approved pediatric
use of oncology and hematology drugs. Overall, this study will benefit
the planning of future drug development in oncology and hematology in
the pediatric population.