In September 2017, Charles Schmidt published an article in Scientific American entitled “Many Pediatric Studies Are a Waste of Time.” His conclusion was based on a 2016 Harvard University study which found that between 2008-2011 more than 40 percent were never finished or that they were completed but never published even five years later.
But how valid is his assertion?
Legislation requiring disclosure of results on databases has been enacted in both EMEA and the U.S. and should do much to address the issue of non-publication, although given that few pediatric studies complete within a 5-year period, the full impact of this will not be apparent for a couple of years yet.
However, the “elephant in the room” remains the number of trials which never complete.
One of the advantages of working within a large Contract Research Organization (CRO) is the opportunity to experience a wider range of target conditions, using a broader range of mechanisms, often in more diverse geographic settings than most sponsors.
At any given time, a large CRO may be managing upwards of 200 pre-registration pediatric drug trials across nearly as many clinical indications, a breadth which few industry sponsors or academic groups can match. From our experience, the challenges relating to trial completion largely fall into two categories: study design and documentation.
As Schwartz states, “by definition, most childhood diseases are rare.” Disregarding other confounding factors, this of itself renders trials in children with most conditions challenging.
In Europe and in the U.S., the well-intended measures introduced some 20 years ago to encourage pharmaceutical companies to assess their new products in children are now creating situations not envisioned when they were implemented.
For example, the adult population of oncology patients has clearly been sufficient to support the development of multiple CAR-Ts and PD1/PDL1 inhibitors, but the requirement for the manufacturers of each of these drugs then to execute a pediatric program strains credulity beyond breaking point.
A similar situation for proton pump inhibitors (PPIs) used in esophagitis, and the 5HT-3 antagonists used to reduce chemotherapy-induced nausea are approaching this situation. Every month, the authorities consider requests from manufacturers to be excused of their PIP/PSP requirements, and in most cases the reason is the same: an insufficient number of children in accessible locations with research-experienced sites have the condition to be able to execute multiple, simultaneous studies of new drugs with similar mechanisms of action which are sufficiently robust to allow the regulators to make their assessments.
Last year, the EMA convened a meeting involving five CAR-T manufacturers, recognizing that the pediatric population was too small to support simultaneous development of five new drugs. All parties agreed that some form of prioritization was required, otherwise none of these drugs was likely to be properly evaluated in children. However, no agreement was reached regarding the methodology for prioritization. Until and unless processes are agreed to manage such situations, this challenge will not be solved.
The robustness of a clinical trial relies not only on subject numbers. All pivotal, pre-registration clinical trials involve a comparison of the new drug to some reference. For a condition with no specifically-approved treatment, a placebo-controlled study may be ethically-acceptable (to IRBs, ECs, parents/guardians and pediatric patients).
A closer look
Let’s look at a simple scenario: a condition for which one drug, an ABC-123 inhibitor, has recently been approved in children, having been evaluated against a placebo, and shown similar superiority to placebo in both adult and pediatric trials.
Is a placebo-controlled trial in children ethically-acceptable with a second ABC-123 inhibitor which shows similar superiority to placebo in adults? If the answer is “yes,” then how many ABC-123 inhibitors need to be approved based on placebo-controlled trials in children before placebo-controlled trials become unjustifiable?
A better solution may be a platform trial, in which children are randomized to one of multiple active treatment arms, or to placebo. Legal and statistical issues aside, the difficulty here is persuading different manufacturers to agree not only on the principle, but also on the details, including the time-frame, stopping rules, discontinuation criteria and so on, and participating in such a trial would require disclosure of much confidential information to potential competitors. Until an independent group is imbued with the responsibility for conducting such comparative studies, maintaining the requisite firewalls, they seem unlikely to be undertaken to a significant extent.
The alternative – evaluating a new drug against an existing treatment – opens up two issues. Whether the objective of the trial is to demonstrate similar or superior efficacy or safety or both, it will substantially increase the sample size required – and perhaps that is the right outcome.
Once we have four ABC-123 inhibitors to treat the same condition in children, do we need a fifth? But the conduct of pediatric trials is now a regulatory and legal requirement, beyond the control of manufacturers, upon which approval to market the drug for adults may be predicated, so unless we wish to restrict the range of new drugs available to the much larger adult population with a given condition, it would seem that we do need to evaluate the fifth, and sixth, and perhaps even seventh ABC-123 inhibitor in children.
The second issue is the acquisition of sufficient supplies of the approved drug to conduct the gold-standard trial – a double blind study. Competing parties won’t easily reach an agreement to supply active drug and matching placebo to allow a double-blind, double-dummy trial to be conducted, without wishing to exercise considerable control over the design of the study, having access to all of the data it generates and probably gaining a full understanding of the properties of and experience with the new drug. Similarly, the manufacturer of the new drug will wish to protect their intellectual property, but will also wish to understand more about the approved drug which, should the new drug also be approved, will become a competitor.
At present, within a free market, manufacturers cannot be compelled to provide materials for such studies; they will do so only if it brings some commercial advantage. Perhaps one route to a solution is to work out what might be done to provide such an advantage. An alternative would be for such studies to be managed by a neutral party or parties. In the oncology area, the Children’s Oncology Group would be an obvious candidate, but similar groups do not exist for all target conditions, and of those which do, few have the infrastructure to manage large studies to the regulatory standards required, e.g., data management, or regulatory applications, legal requirements regarding assent and consent and site contracts across a large number of countries.
The decisive factor for trial participation is the human element. Children can be approached to participate in a clinical trial only after permission from the parent/guardian, other than in very particular circumstances. As part of deciding whether to give permission, the parents/guardians must be given sufficient information for their needs.
The regulations which govern this are clear, and include a requirement to explain treatment alternatives. Going back to the ABC-123 inhibitors, if an anxious mother were told that four such drugs have been approved for use in children, but the investigator would like to enrol her sick child into a clinical trial involving a new ABC-123 inhibitor, not yet evaluated in children, and placebo, requiring additional blood samples and hospital visits, how likely is she to give permission for her child to be enrolled?
Staying with parent/guardian consent, information sheets and consent forms have long attracted criticism from commentators for being overly long, and littered with technical, medical, and legal terminology which render them difficult to understand.
Working in a CRO, we have reviewed, literally, dozens of such documents every year. Some are excellent; their wording and structure correspond to the target age-range and focus on aspects likely to be important to children.
Others have been disappointing, with readability scores suitable for someone at post-doctoral level, and addressing topics unlikely to be meaningful to children. Do 10-year old children need to understand compensation for iatrogenic injury, or data privacy requirements, or their rights regarding samples retained for bio-banking or other future genetic tests?
That information, and more, is contained in the documents given to the parents/guardians. The concept of assent, unlike consent, is not widely recognized legally, so little defense is afforded to the trial sponsor by including that information in the documents provided to the child.
Yet corporate attorneys often insist upon the inclusion of such wording, not permitting changes. This, and the seeming desire to provide a detailed explanation of what will happen at each visit, mean that information sheets commonly exceed 20 pages. How many children have the patience to read such documents, far less assimilate the information they contain?
Parents and children are often altruistic; they want to help generate information which will help others, but they are also seeking treatment for their own conditions. We rarely treat children with nothing; even if no treatment is approved, physicians use their experience to come up with the best option they can. Trials adding the new drug or placebo on top of the standard of care are becoming more common, but platform trials, enabling the assessment of multiple new drugs simultaneously, would reduce the number of children allocated to placebo, and enable comparative assessments to be made among these new drugs.
The challenge is finding a method acceptable to manufacturers. One option might be to require active-comparator studies, and one way to achieve that may be to require the manufacturer of the first drug approved for pediatric use in an indication to supply active drug and matching placebo for two such studies as a condition of approval, with the regulatory agencies deciding when to implement this option.
But changes such as these will not make a difference to enrollment in pediatric trials until the documentation provided is more readily understood by parents and children. IRBs/IECs must shoulder their obligations, and reject documents which are overly-complex. In addition, legal clarity is required in all the major jurisdictions regarding the requirement to include information in documents given to children, to relieve corporate attorneys of their perceived obligation to give children as much information as their parents.
Finally, the issue of publishing negative or neutral results is not entirely an issue for manufacturers. Respectable journals are not inclined to publish such studies, unless they demonstrate that an entire class of drug is unlikely to be effective (or not), and with single agent-versus-placebo designs, that is unlikely to happen. Active-comparator and platform trials seem more likely to increase the prospect of publication.
So, are pediatric studies are a waste of time?
No, because they all generate some data, they allow physicians to gain experience with a new drug, and because even incomplete studies enable regulators to bring together data regarding the safety of drugs within a class.
Could they be better?
Yes, very much better, enabling between-drug comparisons, and enrolling a wider section of the population than is currently the case by using clearer documentation. But the will to make, and if necessary to enforce, changes to our current practices must be present, otherwise 10 years from now we will be in exactly the same position as we are at present.