Sandra Blumenrath and Ronald Portman provide an overview of the challenges of developing medicines for children, and explore recent innovations that will spur advancements in the field
Paediatric drug development has seen a significant paradigm shift in recent years. Globally converging policy reforms, pragmatic solutions to the complexities of paediatric clinical studies, and an evolving conviction that children have the right to benefit from scientific progress rather than to be protected against it have all contributed to improvements made in the paediatric drug development process. Yet it remains challenging to develop and provide quick access to innovative, high-quality medicines that are safe and effective for children, especially in rare paediatric diseases.
Addressing the Need
Ignoring paediatric considerations in drug development is no longer an option. More than 50% of drugs used in children and more than 90% of those used in newborns lack proper paediatric labelling. To ensure that medicines are assessed for children, paediatric provisions exist in several regions around the globe. Specifically dedicated, complete paediatric legislation, however, currently only exists in the US and the European Union (EU). Here, legislation requires specific paediatric development plans, such as the initial Paediatric Study Plan (or iPSP) and the Paediatric Investigation Plan (or PIP), respectively. In the US, two companion legislation acts – the Paediatric Research Equity Act (PREA) and the Best Pharmaceuticals for Children Act (BPCA) – and the FDA Safety and Innovation Act (FDASIA) have significantly increased the number of drugs with paediatric labeling. More funding and scientific advances in areas of special benefit for children, such as platforms for cell and gene therapy, have also spurred the increased focus on paediatric clinical trials and products. But with this new research environment also comes a unique set of challenges for today’s drug developers.
The challenges
The challenges of paediatric drug development are manifold and include ethical, methodological, operational and financial limitations. Developing medicines for children is essentially comparable to developing orphan drugs. With at least five paediatric subpopulations – based either on age or developmental stage or, more recently, on rapid changes in pharmacokinetics and pharmacodynamics (PK/PD) – there are very few patients with a given disease. A single disease and the corresponding treatment response patterns may manifest themselves slightly differently in each subpopulation, and inclusion/exclusion criteria further narrow the pool of eligible patients for study enrollment.
Because of the small patient pool, paediatric studies have high infrastructure needs, requiring trials to take place across multiple sites and countries, often with an average of only one to three patients per site per year. Many trials fail or stall, and those that do succeed take too long to complete, resulting in nearly a decade between indication in adults and paediatric labelling. And, despite new regulation, systematic guidelines and methodologies are still lacking. What’s more, although new regulation has increased the inclusion of children and adolescents in paediatric drug development, trials are still not widely accepted by society for fear of exposing children to uncertain treatment effects.
Luckily, product development for children has matured over the past years. Compared to only a decade ago, clinicians, developers, regulators, and patients/parents have gained substantially more experience with innovative paediatric clinical trial design (including paediatric extrapolation) and paediatric formulation. The true partnership of this paediatric drug development community is therefore much better equipped to address both remaining and new challenges.
Addressing the challenges
Regulatory directives to conduct paediatric clinical trials have spurred innovation by the paediatric community at various stages of drug development. Several methodological and operational innovations leverage existing information and infrastructure to build effective paediatric studies, including: paediatric extrapolation; pharmacometrics (simulation and modelling), often with the use of Bayesian statistics; innovative clinical trial designs; real-world evidence; inclusions of adolescents in adult clinical trials; global clinical trial networks; and tissue agnostic approaches to drugs with multiple potential indications for the mechanism of action.
Early planning, however, is essential for the success of these innovative approaches.
With paediatric extrapolation, efficacy data from adult studies is extrapolated to paediatric patients if the disease course and response to treatment are sufficiently similar between adults and children. Because children are considered vulnerable, they should not be enrolled in clinical trials if data from adult studies can sufficiently answer the scientific question at hand. Extrapolation is becoming the default approach whenever adult and paediatric indications are similar. However, early planning during adult development is key if the trials are meant to support paediatric drug development.
When there is uncertainty as to whether a drug will indeed work similarly in children, Bayesian statistics can be a powerful innovative analytical tool allowing for smaller, more efficient paediatric studies with valid statistical conclusions. Bayesian methodology is essentially a variant of extrapolation; it borrows data from adults and adolescents and combines them with new paediatric response data to arrive at an estimate for overall paediatric treatment response. Some argue that, if used appropriately, it provides scientifically robust evidence that greatly reduces the sample sizes needed in paediatric studies and other operational challenges, such as trial duration and slow or reluctant enrollment.
Similarly, pharmacometrics (simulation and modelling) and Physiologically Based Pharmacokinetics (PBPK) can optimise already available data to inform future paediatric clinical trials. Dose-response modelling based on available PD/PK data, for example, allows for a much quicker and more scientifically accurate dose selection for paediatric trials. Finally, the use of real world evidence can be utilised to assess the feasibility of paediatric studies and provide needed data for proper study design selection.
Although innovative analytics do not replace the need for clinical trials from a regulatory standpoint, they certainly increase efficiency by providing a formal approach for incorporating prior information into the planning and analysis of the paediatric study.
Another step in the right direction is that increasingly more clinical studies enrol adolescents in adult clinical trials. US legislation, for example, now requires that adolescents be included in the early phase studies of most cancer treatments. For many conditions, efficacy and safety data are very similar for adolescents and adults, so the inclusion of adolescents raises less of an ethical concern. However, some unresolved issues with adolescent studies remain.
For one, there are no clear and scientifically validated guidelines as to the proper number of adolescents to include in adult clinical studies. In European clinical studies it has been proposed that adolescents comprise roughly 10% of the study population, but this is by no means an official policy. Also of particular concern are timing issues: adding adolescents to an adult trial can shorten the time for regulatory reporting requirements which can be a strong disincentive to adolescent inclusion.
Despite the issues, there are clear benefits to including adolescents in adult clinical trials. Information from adolescent trials can be used to extrapolate to younger patients. As a result, instead of having to wait an average of nine years between adult programme approval and paediatric labelling, drugs can be approved much faster and prescribers can gain quicker access to information on dose and side effects.
A major root cause for why paediatric study programmes are challenging or even fail is due the lack of sustainability of the ad hoc ‘networks’ formed for each trial. Many fewer children than adults typically need treatment for a particular condition that the adult drug was initially developed for. Therefore, eligible paediatric patients can be few and far between, requiring developers to essentially find as many study sites as patients – a process that is time-consuming, inefficient and expensive. Stakeholders have long recognised that clinical research networks can break down operational and other barriers and make patient recruitment and clinical studies more efficient.
In Europe and the US, two paediatric clinical trial networks – The Collaborative Network for European Clinical Trials for Children or conect4children (c4c) and the Institute For Advanced Clinical Trials For Children (I-ACT) – have joined forces to build the capacity for multinational paediatric studies for all disease areas and all phases of paediatric drug development. These independent, non-profit networks are committed to facilitating high-quality clinical trials that generate enough data to ensure the safety and efficacy of innovative therapies for children of all ages.
Conclusions
Paediatric drug development is complex and difficult. The challenge of obtaining adequate paediatric efficacy and safety data is the major cause for the unacceptable lag between adult approval and incorporating paediatric information in labelling. Children should only be part of clinical trials if adult clinical studies cannot provide the answer, and they certainly should not be enrolled in trials that have little chance of completion. We are therefore faced with a strong ethical obligation to find appropriate alternative solutions to providing children with safe and effective treatments in a timely manner.
Most companies are beginning to pay attention to paediatric considerations early in development, focusing on both possible paediatric formulations and potential toxicity issues as well as on gathering adult data that can assist with paediatric programmes. With the emergence of innovative analytical tools and increased access to global clinical research networks, drug developers are now better equipped than ever to conduct paediatric studies. However, opportunities like these are easily missed if developers don’t consider the operational aspects of paediatric clinical trials very early in drug development.
The next chapter for paediatric drug development must focus on stimulating the development of drugs specifically for conditions that present themselves in particular paediatric ages, especially in the neonatal period. Although paediatric regulations have been pivotal in moving paediatric drug development forward, not all paediatric subpopulations have benefited equally. With their unique, transitional physiology, premature and newborn infants constitute a distinct developmental population that differs considerably from older infants. For this population, even fewer therapies exists and nearly all prescribing occurs off-label and without appropriate clinical evidence. So far, attempts at neonatal policy reforms have received only limited traction. Successful policy reforms will have to address financial and operational concerns to stimulate research investments that specifically target neonatal needs.
Sandra Blumenrath is Science Writer at the Drug Information Association (DIA), while Ronald Portman is executive director of Paediatric Development, Science and Innovation at Novartis’ Paediatric Center of Excellence, Clinical Development & Analytics