Article Text

Download PDFPDF

Prerequisites to support high-quality clinical trials in children and young people
  1. Steven Hirschfeld1,
  2. Florian B Lagler2,3,
  3. Jenny M Kindblom3,4
  1. 1Uniformed Services University of the Health Sciences, 4201 Jones Bridge Road, Bethesda, Maryland, 20814 USA
  2. 2Institute for Inherited Metabolic Diseases and Department of Pediatrics, Paracelsus Medical University, Clinical Research Center Salzburg GmbH, Strubergasse 21, 5020 Salzburg, Austria
  3. 3European Society of Developmental, Perinatal and Pediatric Pharmacology (ESDPPP) Council, Leuven, Belgium
  4. 4Pediatric Clinical Research Center, Queen Silvia Children’s Hospital, Sahlgrenska University Hospital, Gothenburg, Sweden
  1. Correspondence to Dr Jenny M Kindblom, Sahlgrenska University Hospital, S-413 45 Goteborg, Sweden; jenny.kindblom{at}vgregion.se

Abstract

Children have the right to treatment based on the same quality of information that guides treatment in adults. Without the proper evaluation of medicinal products and devices in paediatric clinical trials that are designed to meet the rigorous standards of the competent authorities, children are discriminated from advances in medicine. There are regulatory, scientific and ethical incentives to address the knowledge gap regarding efficacy and safety of medicines in the paediatric population. High-quality clinical trials involving children of all ages can generate data that will ultimately close the knowledge gaps and support decision making.

For clinical trials that enrol children, the needs are specialised and often resource intensive. Prerequisites for successful paediatric clinical trials are personnel with training in both paediatrics and neonatology and expertise in clinical trials in these populations. Moreover, national and international networks for efficient collaboration, dissemination of information, and sharing of resources and expertise are also needed, together with competent, efficient and high-quality local infrastructure with effective processes. Monitoring and oversight bodies with the relevant competence, including expertise in paediatrics, is also an important prerequisite for paediatric clinical trials. Compromise in any of these components will compromise the downstream results.

This paper discusses the structures and competences needed in order to perform effective, high-quality paediatric clinical trials with the ultimate goal of better medicines and treatments for children. We propose a model of examining the process as a series of components that each has to be optimised, then all the components are actively optimised to function together as an ecosystem, and the resulting ecosystem functions well with the general research system and the healthcare delivery system.

  • evidence based medicine
  • general paediatrics
  • outcomes research
View Full Text

Statistics from Altmetric.com

Introduction

Data from clinical trials are needed to drive evidence-based child health and to enable safe and effective use of medicines in children. The knowledge gap regarding efficacy and safety of medicines in children is well known, has been acknowledged by authorities and is reflected in legislation,1 but conducting and completing paediatric clinical trials remain as one of the biggest challenges.2 The rationale for organising clinical trials is to generate data that will support decision making and, ultimately, forecasting. If the data were available elsewhere, there would be no need to invest in the infrastructure, operations and quality assurance practices or to expose participants to the risks related to participating in trials, including potential physical harm, delay in receiving alternatives and potential breach of trust through misunderstanding and false expectations and misplaced hope.

Following the paradigm that data lead to information, which in turn leads to inference and then knowledge, clinical trials provide the data in the form of observations and measurements. The data are transformed into information through analysis, and from the analysis can come inference. The inferences are subsequently used, generally in combination with other information and inferences, to support decision making. Given the range of decisions that clinical trial data could be applied to, an essential challenge is to generate data with the highest quality and stringency and the lowest bias and uncertainty. The ability to achieve these goals is context dependent in that the intersection of clinical trial participant, infrastructure, personnel and process must align in the most favourable manner possible.

The aim of this paper was to discuss the structures and competences needed in order to perform effective, high-quality paediatric clinical trials with the ultimate goal of better medicines and treatments for children.

A Conceptual model

The rationale and justification for clinical trials in general, and especially clinical trials enrolling children, is that this is the most valid way to obtain the data needed for decision making.3 4 The types of decisions that require data go far beyond individual clinical patient care and include, but are not limited to, product development, research funding, marketing authorisation, policy development, public health promotion and healthcare provider organisation purchases. However, decision making in these settings and others can have additional dimensions than selecting the best option among alternatives. Decision making is a form of and closely tied with forecasting. An implied expectation of making a decision is that future results will align with values and goals that define a given programme or context, that betterment of the lives of others will be achieved and that trust in the decision maker, the decision process and the consequences of the decision are upheld by a promise to maintain or extend a beneficial outcome. A direct implication of the model is that clinical data that are both of high quality (internal validity) and interoperable with other data (external validity), be it clinical data or other data types, have greater utility and therefore greater value because they can be used to support many decisions.

Regulatory and stakeholder initiatives leads to building of networks

In the last decade of the 20th century and early part of the 21st century, the regulatory authorities in the USA and Europe built legal frameworks for the development of paediatric medicines.1 A detailed description of the drug approval process and regulation for paediatric medicines in different regions is beyond the scope of this paper but is available in the literature.5 6 The submission of paediatric data on already marketed medicines is incentivised by the prolongation of market exclusivity.1 Consequently, these regulatory initiatives increased the demand for clinical data and therefore a need for clinical trials enrolling children. In the late 20th century, the resources for performing high-quality, high-stringency studies that could be submitted to regulatory authorities to support product labelling were not well established. Paediatric infrastructures were lacking on a national and subnational level, and the availability of networks with coordinated actions to facilitate multicentre/multinational paediatric trials was poor. Therefore, a combination of stakeholders started initiatives to facilitate the generation of high-quality paediatric clinical research data in a cost-effective manner.

In 2010, the Eunice Kennedy Shriver National Institute of Child Health and Human Development initiated a new model through the establishment of an umbrella organisation that would coordinate the participation on a trial-by-trial basis of a roster of prequalified sites. The initial purpose of the Paediatric Trials Network, as the entity was named, was to perform mainly pharmacokinetic and pharmacodynamic studies on off patent products for providing paediatric dosing information to be incorporated into the approved package product label. It has since developed into an alliance of clinical research sites cooperating in the design and conduct of general paediatric clinical trials (table 1). The authority for the programme was through the Best Pharmaceuticals for Children Act of 2002, which directed the National Institutes of Health to issue a contract to perform such studies.1 A similar model of using a central coordinating entity to prequalify and assume responsibility for a flexible roster of primarily academic performance sites was used to establish conect4children (C4C) (www.conect4children.org), a European-based initiative, and two other US-based initiatives, the Institute for Advanced Clinical Trials for Children and the Institutional Development Awards States Paediatric Clinical Trials Network. Each of these initiatives are attempting to address some of the historical challenges to paediatric research that include lack of alignment of stakeholder interests, lack of an integrated infrastructure, operational inefficiencies and overcoming feasibility barriers.6

Table 1

Examples of paediatric networks

In Europe, a key initiative is the European Network of Paediatric Research at the European Medicines Agency (Enpr-EMA). It is a ‘network of research networks’, investigators and centres with recognised expertise in performing clinical trials in the paediatric population. Enpr-EMA aims to (1) increase high-quality research into medicines for children, (2) increase the availability of authorised medicines for children and (3) to increase the available information on medicines. Unnecessary studies in children should be avoided as well as delaying the authorisation of drugs for adults. The European Medicines Agency (EMA) acts as a facilitator by providing secretarial support to the activities of the network, ensuring exchange of information between network partners and providing information to external partners and stakeholders. Enpr-EMA members perform research in children, from newborns to adolescents, in multiple therapeutic areas and ranging from pharmacokinetics to pharmacovigilance. Networks, centres or investigators can apply as members. The involvement of patients and parents in the planning, design and conduct of clinical research is increasingly important. To meet this need, patients and their representatives need to have access to specific information, normally only available to professionals. The C4C consortium is a recently initiated public–private network with the ultimate goal to contribute to better medicines for babies, children and young people that promotes innovation in the design of paediatric clinical trials. C4C strives to involve actively patients and families through their participation in reviewing protocol sections, assessing patient documentation and obtaining their view on unmet needs. C4C offers training and educations for young people and patients specifically within the area of paediatric clinical trials (conect4children.org/patient-and-public-involvement). EPTRI, TEDDY, PedCRIN are other European initiatives with the goal to facilitate different parts of the paediatric research process (table 1). The European Patients’ Academy (EUPATI) is not primarily focused on paediatric trials, but it is a trusted source of credible information on drug development and clinical research in lay language. EUPATI efforts to develop patient experts based on three pillars: the Certificate Training Programme, the EUPATI Toolbox and a very comprehensive online glossary.

Recruitment, power and endpoints relevant in paediatrics

Children are generally healthy. Compared with adults, children have lower prevalence of diseases and medical conditions, and eligible populations are often small or very small in number and, in addition, geographically scattered. The heterogeneity caused by developmental aspects of the child physiology and pharmacology and more diverse background conditions in children than adults, make recruitment of a sufficient number of patients an even greater challenge. Consequently, a very high number of study sites often is needed in order to enrol a sufficient number of patients for the desired statistical power. To some extent this hurdle is overcome by a plethora of networks around paediatric diagnoses and clinical research,7–11 but other issues remain. Even with a high number of sites, recruiting children in time and to target remains one of the biggest barriers, if not the biggest, with substantial impact on study success and data quality.2 12 An accepted estimate is that approximately one-third of paediatric studies are delayed due to inadequate recruitment, and one-fifth are discontinued.12 An international consensus is that children who are considered normal and healthy based on prevailing community standards should not be enrolled in therapeutic interventional studies. This means that children are not involved in the classic type of phase I trial where a new compound is given to ‘healthy subjects’. Consequently, screening and population enrichment methods must be applied in order to efficiently locate the target population. With 70% of children in the USA being treated in general hospitals, focusing recruitment only on paediatric specialty hospitals, treating approximately only 30% of paediatric patients, may miss the opportunity to enrol eligible patients.13 A future scenario could be that research can be done at facilities that are local to where the patients are. This could be in the form of providing the training and support to local facilities for high-quality research data capture using existing infrastructure or have mobile research units that can travel to local healthcare facilities or other locations, including homes. The cost for either training and equipping local facilities or offering mobile facilities could be justified if the time and efficiency of recruitment and data capture were increased sufficiently to lower the overall study costs.

The use of hard clinical endpoints (for example death and disease progression) can be particularly challenging or simply not feasible, in small populations, for example, when the prevalence of the events of concern is low. Other endpoints may be unsuitable or irrelevant to use in children; for example, the 6 min walking test is not possible in the very young, and strongly impacted by the test circumstances (eg, motivation from bystanders etc) even in older children. Lung function tests using spirometry in the very young is only feasible in very specialised centres. Clinical trial endpoints need to be age and development appropriate, and the catalogue of such endpoints that span multiple ages and are properly calibrated is small. As an example, lymphocyte counts and lymphocyte subset ratios change with age, so endpoints based on haematological assessments have to be carefully calibrated with proper age and size categories. The Report ‘Children’s Health, a Nation’s Wealth’ noted that a lack of proper age-adjusted endpoints hampered proper analysis for longitudinal studies.14 Endpoints that reflect the many dimensions of child growth and development are an ongoing area of research.15 To provide sufficient precision to minimise both bias and the number of study participants, study endpoints need to be well characterised by properties such as sensitivity (proportion of false negative), specificity (proportion of false positive), and receiver operator characteristics (changes in readout over a range of inputs). Outside clinical laboratory assessments, few clinical research endpoints have such quantitative properties defined across multiple age and development groups. There are no central libraries or repositories for endpoints as there are for terminologies (https://evs.nci.nih.gov/), and the performance characteristics of endpoints used in studies are generally not reported as part of the outcome. This makes comparison across studies challenging and requires the use of concurrent controls, particularly for time-to-event endpoints, to interpret properly results. Assessment of multiple endpoints can be helpful in these situations, but careful consideration of multiplicity is required.16 Particularly in rare diseases, composite endpoints, combining several outcomes into one measure, are increasingly used. However, all components should be relevant to, assessable and validated in the whole study population. Including many study sites may help to recruit a sufficient number of patients, yet it adds complexity and in most cases heterogeneity. Particularly in multinational studies, divergent reference values for laboratory analyses can be a challenge. A value considered just normal in one site may need to be considered an adverse event in another site, if not specifically addressed within the study protocol. The gold standard treatment, which might be the comparator in a controlled study, may also differ between countries. Studies with several treatments in many arms for one or similar indications, often called basket studies or platform studies, can be applied in such scenarios.17

Infrastructures

Several critical components must coincide in time and location to generate high-quality, high-stringency data for paediatric clinical trials. These are infrastructure, personnel, process, and monitoring and oversight around the study participant. The general history of a dedicated research infrastructure for paediatric clinical research began in the 1950s in the field of paediatric oncology. The concept evolved over the subsequent half century with the development of networks of investigators at different institutions linked by interest and training in a particular disease or subspecialty. The result is that the same institution could participate in more than a dozen networks, not including ad hoc networks for a single study, and there would be no operational overlap or resource sharing other than use of general activities such as human subject protection oversight or clinical laboratories. Participation in several networks is still often the case today. The inconsistencies and variations in practice introduced delays and even contradictions that needed to be resolved prior to studies proceeding. During the last decades, great efforts have been put into building up national and international networks in order to facilitate high-quality paediatric trials through efficient collaboration, dissemination of information and sharing of resources and expertise (see also the section Regulatory and stakeholder initiatives leads to building of networks). Today, paediatric oncology and paediatric rheumatology represent areas where the number of clinical trials have increased over the last 10 years and with strategies for advanced trials in early phase of drug development.2 8

The concept of local infrastructure is emerging as a complement to network development and means that an entity establishes a framework, culture, workflow processes, quality assurance and control mechanisms that are consistent with all applicable legal, ethical, scientific and regulatory principles, and demonstrates operational compliance through a certification process with the purpose of local facilitation (table 2).18 19 A recent study where researchers, regulators and sponsors gave their views on paediatric clinical trials identified infrastructural barriers as one of the major challenges with paediatric clinical trials.18 Important benefits of an infrastructure are the use of standard operating procedures and methodologies grounded in standards and reference resources. These include but are not limited to paediatric-specific terminology,20 site performance standards (reference https://sasi-accreditation.org/), data standards (reference https://www.cdisc.org/) and mechanisms to function well with collaborating organisations. Especially in studies that occur in different geographical and demographic regions, where assessment techniques can vary based on local practices and resource availability, the reliability and interpretation of events and data can be enhanced by capturing the method used as well as the actual data values.21 Such an approach is currently in development for global neonatal research by the International Neonatal Consortium (https://c-path.org/programs/inc/).

Table 2

Components of a paediatric clinical trial infrastructure

Personnel and paediatric clinical trial expertise

Staffing and staff training for research are not standardised. A general paradigm is that a physician or scientist will lead a study; a nurse will coordinate patient encounter logistics and data collection; and a data manager will have local responsibility for data transmission, archiving, initial processing and security. Even so, the contribution by residents, PhD students, research fellows and other groups of staff in training should not be underestimated. For paediatric clinical trials, relevant paediatric competence and experience is crucial and necessary, but there are no universally accepted formal degrees for most aspects of research, although individual certifications and master’s level training programmes for different types of support staff continue to proliferate. Several initiatives have tried to address this lack. For example, the Global Research in Paediatrics initiative (www.cordis.europa.eu) and recently the C4C consortium (www.conect4children.org) have addressed this and will contribute to greater insight and expertise into all relevant aspects of the organisation, implementation and delivery of paediatric clinical trials through an educational platform (https://conect4children.org/academy/). A recent study showed a dearth of paediatric trial expertise.18 The general paradigm is serving some time in an apprenticeship relationship during clinical specialty training and then taking online or in-person classes for certification in human subject protection principles and regulations. Role-associated training is available through organisations such as the Collaborative Institutional Training Initiative (https://about.citiprogram.org). In general, the training emphasises compliance and oversight.

Aside from compliance and general introductions to frequentist statistical methods, the operationalisation of research, the application of tools for interoperability, the principles and practices of data integrity and security, use of terminologies and common data elements, and quality assurance and control are not universally taught nor readily available in a format designed for paediatric clinical research. Consequently, the experience and expertise that the research community brings to the endeavour is variable.

Monitoring and oversight bodies

Clinical trials in children are subject to additional protections and oversight compared with clinical trials in adults. While the justification and rationale for the protections are consistent around the world and reflect the value and status of children in society, the interpretation and implementation of the protections are inconsistent with many perceptions of what is expected. The consequence is that different review panels can come to different conclusions and research projects are delayed.

The practice of serial approval through multiple oversight bodies that could take up to 2 years prompted experimentation with a single institutional review board (IRB). The US National Cancer Institute was the first to adopt a central IRB in paediatrics on a national basis, but participation was voluntary, and approximately 40% of the US-based Children’s Oncology Group institutions did not participate.22 A subsequent effort at a national single IRB enrolled 65 US-based institutions that were under contract to the National Institutes of Health (NIH) with 100% compliance.23 The success of this effort preceded the implementation of the NIH policy on single IRBs and the updated Common Rule policy on single IRB.24 A more efficient regulatory process is the goal also when the Clinical Trial Regulation is updated in Europe. The new Clinical Trial Regulation developed by the EMA has the key benefits of a harmonised electronic submission process for all member states that will cover all regulatory and ethics assessments, so that competent authorities in one country (chosen by the sponsor) will approve the protocol. Furthermore, it also aims at improved collaboration and information sharing, increased transparency regarding clinical trials and the highest standards of safety for all participants in clinical trials within EU. The new Clinical Trial Regulation is not applicable yet (https://www.ema.europa.eu/en/human-regulatory/research-development/clinical-trials/clinical-trial-regulation).

The use of independent monitors, whether individuals or committees, for clinical research in paediatrics continues to remain inconsistent due to the absence of national and international guidelines on the topic. Thus, ethical review for paediatric clinical trials needs clear regulations pertaining to paediatrics, standardised, simple and effective ethics processes, along with templates, education and a harmonised view on paediatric protocols.18

Conclusions

Children and youth have the right to therapeutic agents that have been evaluated through scientifically and ethically sane approaches. There is a well-known knowledge gap regarding the efficacy and safety of medicines in children of all ages, as evidenced by the lack of medicines labelled for paediatric age groups and conditions. The knowledge gap regarding efficacy and safety of medicines in children has been acknowledged by authorities and is reflected in legislation. When drugs are not studied in children in clinical trials that are designed to meet the rigorous standards of the competent authorities, children are deprived of treatments based on the same quality of information that guides treatments in adults. The results might be that treatments are delayed, withheld or administered at ineffective dose to children. The rationale for organising clinical trials is to generate data that will ultimately support decision making. For clinical trials that enrol children, the needs are specialised and often resource intensive. One of the biggest barriers, if not the biggest, to achieve the goal of capturing high-quality data from children is the ability to enrol children in a study. For successful paediatric clinical trials, competent, efficient and high-quality infrastructure, both on the national and international level and for local facilitation, is crucial, together with personnel, monitoring and oversight bodies with the relevant competence (including expertise in paediatrics), and an effective process. An effective process includes the use of standards, common, well characterised and age and development appropriate terminologies that capture assessment method, as well as outcome values, properly defined and calibrated endpoints, and relevant training, certification and process oversight, in addition to human subject protection and financial oversight (figure 1).

Figure 1

Using conventional approaches, each study becomes a parochial activity without alignment or interoperability with other studies. Consequently, the inferences for making decisions and forecasts are limited. Using externally validated resources and integrating them into a study enhances the quality and interoperability of the study to produce high-quality data for robust decision support and forecasting.

All these components can be considered part of an ecosystem, and the realisation of the goals of paediatric therapeutics will require a child-centric research ecosystem that can interact with both the more general research ecosystem and the paediatric healthcare delivery systems, wherever they may be and whatever their nature is in different settings. Ecosystems can evolve and achieve functional inter-relationships among the components and with other systems; however, there is an alternative. A process-based ecosystem can be designed and optimised by design. A vision for a robust future for paediatric research would actively engage relevant stakeholders in planning and continuously evaluating options for the various components to interact optimally for the benefit of all. Thus, not only improvements in individual components would be both goals and drivers for activity, but also the interfaces and functioning among the system components would become targets for system optimisation.

Acknowledgments

JK has support from the Swedish state under the agreement between the Swedish government and the county councils, the ALF-agreement (ALFGBG-723791).

References

View Abstract

Footnotes

  • Contributors All authors performed literature searches and read relevant literature. All authors drafted the manuscript and reviewed it critically. All authors approved the re-submission.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Commissioned; externally peer reviewed.

  • Data availability statement There are no data that can be made available with this review article.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.