Commentary on “The Use of Placebo in Pediatric Inflammatory Bowel Diseases”

See “Use of Placebo in Pediatric Inflammatory Bowel Diseases: A Position Paper From ESPGHAN, ECCO, PIBDnet, and the Canadian Children IBD Network” by Turner et al on page 183. The current process of pediatric drug development under the auspices of regulatory agencies has not resulted in decreasing the amount of time for the approval of new drugs in children once they are approved for use in adults. During this interval, children are treated “off-label” with medications for which an appropriate dose is not determined, long-term efficacy is not established, and meaningful safety data are not collected. Recently, the US Food and Drug Administration and the European Medicines Agency (EMA) have proposed that a placebo arm be included in pediatric inflammatory bowel disease (IBD) maintenance trials of biologic medications that are approved in adults and seeking approval in children. This decision has resulted in the objection of many pediatric gastroenterologists, who believe that extrapolating data from adults is sufficient to support the use of a medication for children and that clinical trials involving placebo in this instance are not ethical. Others believe that although medications used to treat adults with IBD are effective in children, the durability of response and exposure-response may differ, and clinical trials in children should be performed to obtain evidence to support approval. There is enough doubt about the durability of treatment to warrant a placebo-controlled trial in children with ulcerative colitis (UC) who have had a clinical response during induction. The position paper published in this issue of the Journal of Pediatric Gastroenterology and Nutrition, supported by many gastroenterologists and the leadership from the European Society for Pediatric Gastroenterology, Hepatology, and Nutrition (ESPGHAN), the European Crohn's and Colitis Organisation, the Global Pediatric IBD Network, and the Canadian Children IBD Network, two Canadian bioethicists, and the ethics committee of ESPGHAN, addresses concerns about the use of placebo in children with IBD (1). They focus specifically on a trial in children with moderate to severe UC who achieved clinical response with a biologic agent. In contrast, the leadership of NASPGHAN as well as other pediatric gastroenterologists declined an invitation to support this position paper. Furthermore, this study has been approved by ethics committees in the Czech Republic, Australia, Israel, Japan, Spain, Slovakia, Belgium, New Zealand, Poland, the United Kingdom, Italy, Austria, Sweden, the United States, and Hungary. Mark Sheehan, an ethicist at Oxford University, has addressed the issue of including children in clinical trials, citing the results of focus groups in which children expressed being upset about not having the option of being included in clinical trials, and warned against the overuse of the term “vulnerable population.” He notes ethical concerns related both to including and excluding children in clinical trials. The ethical decision about including any subject in a clinical trial centers on the concept of clinical equipoise, “the uncertainty within a relevant expert community consistent with competent clinical care… .” (2) Sheehan stated that when experts disagree, equipoise usually exists. In the trial focused on by Turner et al, equipoise is supported by the existence of differing opinions among respected and knowledgeable individuals as well as professional societies. The position paper also addresses situations in which equipoise may be achieved in children with IBD. They state that the use of a placebo may be considered “when a new therapy with a novel mechanism of action without established data in adults is to be evaluated in children, especially when no other alternatives with proven efficacy exist outside the trial.” (1) To place a child for whom there are no reasonable alternatives into a placebo-controlled trial with a medication for which there are no established data even in adults is a position that many would not support even if it met the uncertainty part of the definition of equipoise. The other aspect of equipoise, however, concerns the issue of competent clinical care. Up to now, regulatory agencies have not considered including children in placebo-controlled trials for induction, especially for a medication without demonstrated efficacy in adults. If medications are given to children without evidence of efficacy in adults, clinical equipoise may not be met on the basis of competent clinical care. In the situation in which there are no reasonable alternatives and there are no efficacy data in adults to support treatment in a child, one may consider expanded access or an investigational new drug application. The central question dividing experts in the clinical trial model proposed by the US Food and Drug Administration and EMA is the issue of competent clinical care. If extrapolation of adult data to children is accepted, then equipoise does not exist and conducting a clinical trial in the pediatric population would not be necessary (2). What are the data to support extrapolation of adult data to children? The EMA's Paolo Tomasi emphasized that one of the best examples in which adult clinical data do not extrapolate to children and adolescents is with type 2 diabetes mellitus. Despite similar pathophysiology, the responses to medications are very different (2). In UC, infliximab is approved for treatment of adults and children at a dose of 5 mg/kg (3,4), yet in contrast to the evidence, many key opinion leaders start treating children at a dose of 10 mg/kg. The design of the clinical trial leading to approval of this dose did not include a higher dose for induction, but adult trials did test 10 mg/kg and the results did not differ significantly from the lower dose. Current clinical practice and existing data would suggest that for selecting a dose of a biologic agent used for induction of children with UC, extrapolation of the adult dose may not be reliable. The experience with certolizumab pegol provides additional evidence that the response to antitumor necrosis factor (anti-TNF) medications may not be the same in children and adults. Although approved in adults with Crohn disease (CD), the certolizumab pegol clinical trial in children was terminated because of a “higher than projected discontinuation rate during maintenance phase.” (5) Data in other pediatric IBD clinical trials suggest that maintenance therapy in children may not be the same as in adults. The position paper (1) points out that “an early trial by Markowitz et al on the effect of thiopurines in an inception cohort of children with CD … showed clear superiority of maintenance treatment over placebo, after an open-label induction period with steroids.” (6) This seminal placebo-controlled trial provided a high level of evidence based upon the inclusion of a placebo arm. All of the pediatric subjects were induced with corticosteroids, and 89% achieved clinical remission. Although 91% of subjects who were treated with 6-mercaptopurine maintained clinical remission after 548 days, 53% of the subjects who were randomized to maintenance on placebo remained in clinical remission. This study was conducted in children with CD, but the durability of clinical remission in children with IBD is difficult to predict without reliable evidence. Reliable evidence in children with UC does not exist to guide clinicians as to the durability of response. The possibility exists that a medication that effectively induces clinical remission in children with UC may have a prolonged benefit during maintenance. For these reasons, uncertainty about the durability of clinical remission supports equipoise. Investigators need compelling reasons to include children in a placebo-controlled trial, especially with a medication that is perceived to be effective and is being used “off-label.” In the placebo-controlled anti-TNF study in children with moderate to severe UC, only children who achieve a clinical response during the 8-week open-label induction will continue in the maintenance phase (1). The rerandomization of subjects who continue into the 3-arm maintenance phase is weighted to allocate more patients to 1 of 2 active drug arms (ie, the randomization schedule is not 1:1:1) to limit the proportion of subjects who will receive placebo. The weighting of the randomization to place more study subjects into the active drug arms allows for the capture of additional safety data with the active drug in the pediatric UC population while retaining sufficient statistical power to test the maintenance of remission with study drug compared with placebo. With the exception of corticosteroids, subjects entering the trial are maintained on the medications with which they were treated before entering the trial. For example, children receiving mesalamine or an immunomodulator would continue on that medication after achieving clinical remission. Turner et al state, “It is questionable whether continuing existing treatment, such as thiopurines, can be considered effective treatment since the drug has failed previously, as evident from the inclusion of the child with active disease to a trial.” (1) There is uncertainty that continuing a medication that was not previously effective in controlling UC would be effective once a child has a clinical response or is in clinical remission. Children with UC who flare while taking mesalamine and are brought into clinical remission with corticosteroids often are successfully restarted on mesalamine as a maintenance medication. Few clinicians would advocate for the prolonged use of corticosteroids to maintain clinical remission, suggesting that medications used to re-induce clinical remission may not need to be maintained throughout the maintenance phase. The statement “there has not been hitherto any single precedent where an IBD drug proved to be effective in adults but not in children” should be closely scrutinized (1). Hon and Wang presented data at the 2015 EMA meeting to support that in some trials with anti-TNF agents the exposure response is not consistent between children and adults (2). These data were presented with the caveats that there are differences in endpoints and study design. Nevertheless, the possibility exists that responses to some biologic agents may be different in children and adults and that these differences could affect the selection of a dose in children based on extrapolation from data in adults. Randomization favors treatment with drug over placebo, but concern is raised about loss of efficacy in those subjects who are randomized to placebo. The STORI (infliximab diSconTinuation in CrOhn's disease patients in stable Remission on combined therapy with Immunosuppressors) trial in adults with CD found that when infliximab was discontinued, approximately half of the patients relapsed within 1 to 2 years, but that almost every patient who relapsed was in remission 1 month after a single dose of infliximab. Furthermore, no patient experienced acute or delayed infusion reactions to retreatment despite a hiatus of more than 6 months in >50% of the subjects (7). The extrapolation of results from a trial in adults with CD to a pediatric UC trial may not provide reliable data on which to determine the risk of loss of response; however, in the proposed placebo-controlled anti-TNF trial, rescue therapy is available at 12 weeks, which is 4 weeks after the last dose of active drug. The possibility of developing loss of efficacy with a humanized anti-TNF agent in the ongoing trial in this short time is unlikely and certainly not supported by the data in the STORI trial, which was done with another medication. Using data from a trial in a different disease, population, and medication does not support the validity of extrapolation to evaluate loss of function and durability of clinical remission in children with UC. There is a distinction in children between placebo-controlled trials that assess efficacy during induction (for which there is most likely no equipoise) and placebo-controlled trials that assess durability during maintenance. If equipoise exists, the number of subjects needed to power a placebo-controlled maintenance study is much less than in comparator designed trials, and the data will provide valuable evidence about the durability of response in children with UC treated with biologic agents. The existence of equipoise in the clinical trial in children with UC of an anti-TNF agent is supported by lack of consensus among experts, societies, and ethics committees in many countries in Europe and North America that have approved the trial; examples such as type 2 diabetes mellitus, in which treatment in adults cannot be extrapolated to children; failure of data from clinical trials in adults to predict response in children with IBD; and a lack of evidence for the durability of response to maintain clinical remission in children with UC. The lack of evidence to support extrapolation of adult UC data to children supports equipoise. Allowing children in this specific study to be maintained on medications that are used to sustain clinical remission supports the justification to include children into the placebo-controlled anti-TNF trial. Future medications may have less efficacy at maintaining remission than anti-TNF agents, and placebo arms may be even more necessary. Accepting equipoise does not require that investigators enroll patients, but provides only the ethical evidence that allows an investigator to enroll patients. Clinical trials in children with IBD are essential to provide credible evidence that beliefs about treatment are justified and not anecdotal. Some children and families will want to participate in an effort to determine optimal treatment, but all should be given the option to decide for themselves about the risks and benefits. All stakeholders share in the responsibility to expedite approval of medications used to treat children with IBD, but additional input from bioethicists, parents, and children on what constitutes both an ethical and feasible pediatric clinical trial is needed. As patient-related outcomes, mucosal healing, and surrogate biomarkers begin to be incorporated into study design, the place for a placebo arm will need to be considered in a new light. It is hoped that pediatric gastroenterologists, industry, and regulators will be able to work together to find evidence to support the approval of medications used to treat IBD in children, and “off-label” use will cease to exist.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.