*US Food and Drug Administration, Silver Spring, MD
†European Medicines Agency, London, UK
‡Health Canada, Ottawa, Ontario, Canada
§Pharmaceuticals and Medical Devices Agency, Tokyo, Japan.
Address correspondence and reprint requests to Andrew E. Mulberg, MD, FAAP, Division of Gastroenterology and Inborn Errors Products, OND/CDER, US Food and Drug Administration, 10903 New Hampshire Ave, Silver Spring, MD 20993 (e-mail: Andrew.Mulberg@fda.hhs.gov).
Received 5 September, 2013
Accepted 22 January, 2014
Members of the i-IBD Working Group include primary authors and Jessica J. Lee, MD; Anil Rajpal, MD; Robert Fiorentino, MD; Klaus Gottlieb, MD; Aisha Peterson Johnson, MD; Zana Handy Marks, MD; Wes Ishihara, BS, and Kevin Bugin, MS, RAC, of the Division of Gastroenterology and Inborn Error Product in the FDA; Kader Kourad; Catherine Njue; Cora Chen; Talia De Laurenti of Health Canada; Mutsuhiro Ikuma, MD; Yosuke Kobayashi; Keiko Ueda, MD; Hana Sugai; Akiko Nitta of PMDA; other members from the FDA—E. Papadopoulos, MD, MPH (Study Endpoints and Labeling Development), and Jean Temeck, MD; William Rodriguez, MD, PhD; Suzanne Malli of the Office of Pediatric Therapeutics.
The views expressed in this article are those of the authors and do not necessarily reflect official positions or policies of the US Food and Drug Administration, European Medicines Agency, Health Canada, and the Pharmaceuticals and Medical Devices Agency of Japan.
The authors report no conflicts of interest.
See “Pediatric UC Drug Development: A GREAT Idea Now Needs a GRAND Conversation” by Rosh and Hyams on page 677 and “Steps Toward Harmonization for Clinical Development of Medicines in Pediatric UC—A Global Scientific Discussion, Part 1: Efficacy Endpoints and Disease Outcome Assessments” by Mulberg et al on page 679.
Ulcerative colitis (UC) remains a chronic debilitating disease associated with abdominal pain, diarrhea, blood in stool, and other clinical signs and symptoms, in spite of currently available Food and Drug Administration (FDA)–approved treatments (1). Thus, there is a need for further development of drug and biological products to treat UC. As discussed in Part 1, the lack of global harmonization regarding study endpoints and disease activity indices has presented a hurdle for drug development in pediatric IBD (2). The lack of scientific consensus on other basic questions, such as the need for and extent of extrapolation from adult trial data, use of placebo-controlled trial designs, and optimal use of pharmacokinetic (PK) data, has also impeded drug development for pediatric UC.
In our accompanying paper, clinical trials designed to evaluate the efficacy of medical therapies require clearly defined study endpoints and well-recognized definitions of disease remission (3). Literature review suggests that there is little consistency in the use of primary efficacy endpoints and the instruments to assess clinical outcome in adult and pediatric IBD trials (3,4). Valid and clearly defined efficacy endpoints are needed to extrapolate a drug's efficacy in UC from adults to children. In short, there should be a similar response to intervention as assessed by an instrument, clinical, or pharmacodynamic (PD) endpoint that is considered validated, along with a similar disease progression in adults and children.
According to Dunne et al, complete extrapolation of efficacy from an adult to a pediatric population relies on robust data supporting the assumptions of a similar disease progression, response to intervention, and exposure-response (ER) relation. The effective dose is identified by matching systemic exposures between adult and pediatric populations. Complete extrapolation is supported by pediatric PK and safety data or, in certain patients, pediatric safety data only. Data supporting complete extrapolation do not currently exist for pediatric IBD because there is a lack of adequate evidence for currently available therapies that establish a sufficiently similar ER in children, when compared with adults.
Alternatively, partial extrapolation of efficacy is possible when available data support some but not all of the above assumptions underlying complete extrapolation. For example, there may be data to support partial extrapolation from a single adequate, well-controlled trial confirming efficacy, or a PK/PD study confirming ER in the pediatric population. The ability to use a PK/PD study for partial extrapolation of efficacy assumes that there is a PD measurement that is associated with efficacy in the pediatric population (5).
An important question that needs to be addressed in pediatric UC trials is the use of a placebo-controlled design. According to Temple and Ellenberg, when known effective therapies are available, placebo-controlled trials conducted in adults may be ethical provided that the withholding of known effective treatment will not result in serious morbidity or mortality. In addition, active-control trials, although valuable, informative, and appropriate under many circumstances, often cannot provide reliable evidence of the effectiveness of a new therapy given the lack of assay sensitivity. In other words, it cannot be concluded that the active comparator is effective under the conditions of the clinical trial absent the inclusion of a placebo-controlled arm (6). The use of placebo treatment in pediatric trials is more restricted than for adults because children cannot consent. Placebo-controlled trials are acceptable in situations in which no approved or adequately studied therapies exist for children with the condition under study. In addition, according to the International Conferences of Harmonization guidance on Choice of Control Group (E10), under certain circumstances, a placebo-controlled study of an investigational drug or biologic may involve the withholding of known effective treatment. In such situations, the risks of withholding known effective treatment in a placebo control group should present no more than a minimal risk or a minor increase over minimal risk (7). The application of these general principles may depend on the extent and severity of UC being studied and the individual therapy being studied.
Consequently, the goal of global harmonization of endpoints, reliance on extrapolation, use of PK studies, and trial design is to facilitate future pediatric UC drug development and address unmet medical needs.
The International Pediatric Inflammatory Bowel Disease Working Group (i-IBD Working Group) is an international collaboration of regulators seeking to advance scientific knowledge on study endpoints, extrapolation, trial design, and PK considerations to harmonize global drug development in pediatric IBD, including UC and Crohn disease. Scientific discussions concerning extrapolation, trial design, and the role of PK studies are conducted. The i-IBD Working Group consists of representatives from the US FDA, European Medicines Agency (EMA), Health Canada, and the Pharmaceuticals and Medical Devices Agency (PMDA) of Japan. The views of the i-IBD Working Group expressed in the present article are individual personal opinions or suggestions for drug development in pediatric UC and do not represent regulatory consensus or official guidance from the represented agencies.
Representatives from FDA, EMA, Health Canada, and PMDA discussed existing approaches when evaluating applications submitted in support of new medicines intended to treat pediatric UC at monthly teleconferences from January to December 2012. The discussions were focused on study endpoints, extrapolation, trial designs, and the role of PK studies in pediatric UC. A literature review assessing extrapolation, trial design, and the role of PK studies that have been used in pediatric IBD trials was conducted. The electronic database PubMed/Medline and phrase “inflammatory bowel disease/ AND (extrapolation or trial design or PK)” was used in the literature search. No language and time restrictions were applied during the search. The last search was conducted on January 31, 2013. The Working Group's perspectives on pediatric UC drug development based on the available medical literature are also provided.
Is the Course of UC Sufficiently Similar Between Adults and Children?
Although some differences in disease severity exist, the pathogenesis of UC in adults and children is the same, and the disease courses in these 2 populations are similar enough to allow extrapolation of efficacy outcomes from adults to children (8–11). It has been reported that, when compared with UC in adults, pediatric UC has a more severe phenotype, often involving the entire colon, a shorter time to first UC-related surgery, and a higher corticosteroid failure rate (12–15). Differential diagnoses, such as genetic diseases and irritable bowel syndrome, are more difficult to ascertain and confirm in children with pediatric UC.
Is the Response to Treatment Sufficiently Similar?
The response to treatment is considered similar in both populations (8–11). There are currently no validated tools to measure treatment response including remission in children in a way that corresponds with tools that measure treatment response including remission in adults. Efforts to understand the role of the Mayo Disease Activity Score in assessing treatment response and remission in children and adults are ongoing in the United States, and will be supplemented by trial data from other countries, when available. On completion of the validation or development of appropriate instruments for treatment response and remission, extrapolation of adult efficacy data to children with UC will be possible. The Pediatric Ulcerative Colitis Activity Index (PUCAI) has also been developed as a pure clinical score by Turner et al (16). It has been used in clinical trials as a measure of efficacy supplemented by 1 sigmoidoscopy in children with UC. Currently, the PUCAI does not meet all criteria for a validated patient-reported outcome (PRO) measure as recommended by current FDA guidance because physician-based approach instead of a patient-based approach was used in the PUCAI during the instrument development, implementation, and assessment process (17). In general, 1 type of outcome assessment is not considered to be more important than another type of outcome assessment. Each type plays a critical role in drug development, depending on the context of its use and the goal of measurement. For example, if symptom intensity is the goal of measurement and the patient population can respond itself, a PRO is most appropriate. If clinical judgment is required to interpret an observation, a clinician-reported outcome is appropriate. If the concept of interest can be adequately captured only by observation in daily life, and the patient cannot report for himself or herself, then an observer-reported outcome is chosen.
Is the Exposure-Response Relation Sufficiently Similar?
The ER relation has not been addressed sufficiently in pediatric studies of currently available therapeutic agents for UC conducted to date. Attention to addressing the ER relation for investigational drugs is also needed.
Available data show interindividual differences in PK and PD that could possibly cause underdosing (or overdosing) in individual pediatric or adult patients. Thus, therapeutic drug monitoring and individual dose adjustment may be warranted.
In summary, there are insufficient data regarding both the response to treatment and the ER relation in pediatric UC to support the complete extrapolation of efficacy data from adults to pediatric patients. Absent an accepted PD marker of clinical response and remission, partial extrapolation of efficacy may be accomplished through the use of a clinical endpoint. There is currently no international agreement about the optimal choice of endpoint for clinical investigation for the components of clinical response or remission.
Whether one can extrapolate efficacy, either completely or partially, affects the choice of an appropriate study design. Currently most pediatric UC trials are open-label, randomized trials comparing different dosing regimens of the investigational product. Ideally, a dose-finding study should be done early in drug development to identify at least 2 doses or dosing regimens that can be studied in later-phase trials.
Examples of open-label studies in pediatric UC include that submitted in support of Remicade (infliximab) to the FDA, EMA, and other regulatory agencies (eg, Health Canada). This study was a phase III, open-label, randomized, parallel-group, multicenter trial to assess the safety and efficacy of Remicade induction and maintenance treatment in 60 pediatric patients. The trial was divided into an open-label, single-arm induction phase, in which all patients received Remicade 5 mg/kg at weeks 0, 2, and 6, and an open-label, randomized, parallel-group maintenance phase, in which only patients in clinical response at week 8 (2 weeks after the last induction dose) were randomized in a 1:1 ratio to 1 of 2 groups (q8w and q12w groups). The pediatric Colazal (balsalazide) trial submitted to the FDA was a multicenter, randomized, double-blind, parallel-group study of 2 different dosage regimens in children with a diagnosis of mild-to-moderately active UC.
The i-IBD Working Group also discussed whether an active-controlled 3-arm study design (with 1 arm a placebo) could be considered a potential option. In a putative example of evaluating a new molecular entity against placebo in a clinical trial with 2 active arms, the hypothesis comparing difference between these 2 active arms and placebo arm will yield larger treatment differences than that between the 2 active arms. Hence, the number in each treatment arm will be smaller in the 3-arm trial than in the 2-arm trial, but the total number of subjects needed for the study may be larger to accommodate the additional arms of a trial. The members suggested that enrollment in the placebo arm should not present more than a minor increase over minimal risk to children with mild UC (and thus be acceptable in this population). For children with moderate to severe UC, the placebo assignment may present more than a minor increase over minimal risk during the induction phase, but not in the maintenance phase. For these reasons, use of placebo controls in studies in children with mild UC and during the maintenance phase in children with moderate to severe pediatric UC may be scientifically and ethically appropriate (6,18,19). Although the FDA, EMA, and Health Canada favor the appropriate use of placebo, PMDA states that a randomized controlled trial may be difficult to conduct in Japan because of a low prevalence of pediatric IBD. For this reason, if the similarity of the drug exposure and clinical response between children and adults are confirmed, extrapolation from adult trials and pediatric clinical data obtained from the trials conducted outside of Japan could be used to support the approval of a pediatric UC indication in Japan.
The concerns associated with randomizing children with UC to a placebo group can be mitigated by introduction of an early escape procedure and/or a randomized withdrawal trial design. Escape refers to the prompt removal of subjects whose clinical status worsens or fails to improve to a defined level in a trial.
Dose Selection and PK Studies
Generally, when neither complete nor partial extrapolation of data from adults to children is possible, PK studies are necessary to establish dosing before conducting safety and efficacy trials in children (5). In pediatric UC, in which complete extrapolation is currently not possible but partial extrapolation is feasible, PK/PD studies using a clinical endpoint (until an appropriate biomarker has been identified) are needed to establish an appropriate ER relation that can link to the adult studies; once a proper pediatric dose is determined, safety trials may be conducted. If complete extrapolation becomes possible at some point in the future, PK studies in children would suffice to inform dosing with an aim of achieving drug levels similar to those for adults.
The goal of extrapolation is to facilitate pediatric drug development by reducing primarily the number of required clinical trials and secondarily the number of patients required for study. Although it is reasonable to assume that pediatric UC, when compared with UC in adults, has a similar disease progression and response to intervention, the similarity of the ER relations has not been adequately established. Pediatric-specific endpoints reflecting clinical benefit and disease pathobiology remain to be established. Currently, only a PD measurement, such as clinical remission defined by the Mayo score, can be used to explore efficacy in children. In the absence of defined clinical endpoints, partial extrapolation with conduct of appropriate safety trials needs to be considered.
The known differences in disease course, if any, between adults and children and evaluation of disease course activity must be taken into account when one considers extrapolation and trial design for an individual drug treatment. The overlap of clinical signs and symptoms of other diseases with UC should encourage the careful differential diagnosis to exclude irritable bowel syndrome and genetic diseases, that is, immunodeficiencies and inflammasome mutations that would not necessarily be treated with the study drug.
At the present time, there is no globally acceptable disease activity outcome measure for pediatric subjects and adults. Validated activity indices and outcome measures for evaluation of treatment response in children are needed. Comparison of adult and pediatric UC activity indices is needed to allow extrapolation of adult efficacy data to the pediatric population. Representatives from 4 regulatory agencies agreed that ideally the same outcome assessment(s) would be validated for use in both pediatrics and adults. Adequate and well-controlled studies in children are needed to determine the correlation of PRO measures with other measures of efficacy such as endoscopic and histological evidence of reduction in inflammation.
Although PK studies are useful to establish doses that achieve target concentrations similar to those for adults, a PK study alone may be inadequate if there is a lack of understanding of the ER relation in children. A dose-finding study that explores a range of doses and consideration of weight-based dosing may be necessary. Multiple-dose regimens may be operationally difficult to evaluate given difficulties in recruiting adequate numbers of pediatric patients with UC.
The need for a pediatric efficacy trial and the choice of an appropriate study design are guided by whether efficacy can be extrapolated from informative adult trials. Use of placebo controls in studies involving children with mild UC and during the maintenance phase in children with moderate to severe UC may be scientifically and ethically appropriate to establish the ER relation, thus supporting efficacy. When a placebo control is used, early escape withdrawal criteria must be allowed and 3-arm studies or randomized withdrawal studies can be considered. Using withdrawal-retreatment trial design for a biological agent can be complicated by the concerns of whether the potential immunogenicity of the agent would lead to an ultimate loss of response for the trial subject and any associated complications. Although such design is commonly used in the psoriasis trials of biological agents, generalization across disease conditions may not necessarily be applicable. Paul et al reported in a recent prospective IBD trial that a high proportion of subjects who developed antibody to infliximab in a particular concentration window achieved clinical remission on dose intensification (20). Although results from this single trial are yet to be replicated, it suggests that anti-drug antibody development can be overcome by dose intensification. Immunogenicity of the agent may or may not necessarily be a key determinant of whether efficacy can be regained on retreatment. It would be necessary to evaluate the immunogenicity risk and its impacts on the controlled trials in patients with UC to better implement a withdrawal-retreatment trial design for biological agents in the future.
In summary, based on the available literature and scientific discussion, the i-IBD Working Group provides the perspectives as shown in Table 1. Briefly, partial extrapolation from informative adult studies is a necessary element to construct a pediatric drug development program. Studies that are essential for a development program in pediatric UC include initial dose-finding studies that incorporate PK and preliminary efficacy assessments that support ER modeling followed by an efficacy and safety study to explore further the ER relation. In the setting of partial extrapolation, clinical studies do not need to be fully powered for efficacy, but a comparison with placebo and/or active comparator could facilitate interpretability of the ER data obtained in the trial. When a placebo control is used, early escape withdrawal criteria must be allowed. Three-arm studies or randomized withdrawal studies can be considered in mild pediatric UC and during the maintenance phase in children with moderate to severe UC because the disease severity in these settings is amenable to consideration of use of a placebo. Initiating pediatric studies at the completion of adult efficacy studies that have demonstrated efficacy, and before the authorization process for adults is completed, will allow earlier labeling for children and facilitate the recruitment of pediatric patients absent extensive off-label use. It would be helpful to include preliminary evaluation of PD and individual treatment response in PK studies. If significant interindividual treatment response differences are found, therapeutic drug monitoring and individual dose adjustment may be warranted.
The authors recognize and thank Dr Julie Beitz of the US Food and Drug Administration for valuable editing and insightful input during the manuscript preparation. They also acknowledge Dr Yow-Ming C. Wang, PhD, of the US Food and Drug Administration for helpful discussion during manuscript preparation.
1. McQuaid KR, Alimentary tract. Tierney LMJ, McPhee M, Papadakis A. Current Medical Diagnosis, Treatment. 43rd ed.New York:Lange Medical Books/McGraw-Hill Medical Publishing; 2004. 603–607.
2. Sun H, Mulberg A. Steps towards harmonization for clinical development of medicines in pediatric ulcerative colitis—a global scientific discussion, part 1: efficacy endpoints and disease outcome assessments. J Pediatr Gastroenterol Nutr
3. Cooney RM, Warren BF, Altman DG, et al. Outcome measurement in clinical trials for ulcerative colitis: towards standardisation. Trials
4. D’Haens G, Sandborn WJ, Feagan BG, et al. A review of activity indices and efficacy end points for clinical trials of medical therapy in adults with ulcerative colitis. Gastroenterology
5. Dunne J, Rodriguez WJ, Murphy MD, et al. Extrapolation of adult data and other data in pediatric drug-development programs. Pediatrics
6. Temple R, Ellenberg SS. Placebo-controlled trials and active-control trials in the evaluation of new treatments. Part 1: ethical and scientific issues. Ann Intern Med
7. Department of Health and Human Services/Food and Drug Administration. Children's Health Act/Additional Safeguards for Children in Clinical Investigations of Food and Drug Administration-Regulated Products
. 21 CFR parts 50 and 56/docket no. FDA-2000-N-0009 (formerly 2000N-0074; RIN 0910-AG71), 12937-12951. 2-26-2013. 21 CFR parts 50 and 56 (docket no. FDA-2000-N-0009) formerly 2000N-0074; RIN 0910-AG71.
8. Ardizzone S, Bianchi PG. Inflammatory bowel disease: new insights into pathogenesis and treatment. J Intern Med
9. Carter MJ, Lobo AJ, Travis SP. Guidelines for the management of inflammatory bowel disease in adults. Gut
2004; 53 (suppl 5):V1–V16.
10. Sandhu BK, Fell JM, Beattie RM, et al. Guidelines for the management of inflammatory bowel disease in children in the United Kingdom. J Pediatr Gastroenterol Nutr
2010; 50 (suppl 1):S1–S13.
11. Sawczenko A, Sandhu BK. Presenting features of inflammatory bowel disease in Great Britain and Ireland. Arch Dis Child
12. Heyman MB, Kirschner BS, Gold BD, et al. Children with early-onset inflammatory bowel disease (IBD): analysis of a pediatric IBD consortium registry. J Pediatr
13. Sauer CG, Kugathasan S. Pediatric inflammatory bowel disease: highlighting pediatric differences in IBD. Med Clin North Am
14. Turner D, Walsh CM, Benchimol EI, et al. Severe paediatric ulcerative colitis: incidence, outcomes and optimal timing for second-line therapy. Gut
15. Van LJ, Russell RK, Drummond HE, et al. Definition of phenotypic characteristics of childhood-onset inflammatory bowel disease. Gastroenterology
16. Turner D, Otley AR, Mack D, et al. Development, validation, and evaluation of a pediatric ulcerative colitis activity index: a prospective multicenter study. Gastroenterology
17. Food and Drug AdministrationGuidance for Industry, Patient-Reported Outcome Measures: Use in Medical Product Development to Support Labeling Claims. Silver Spring, MD:US Department of Health and Human Services, Food and Drug Administration; 2011.
18. Ellenberg SS, Temple R. Placebo-controlled trials and active-control trials in the evaluation of new treatments. Part 2: practical issues and specific cases. Ann Intern Med
19. Roth-Cline M, Gerson J, Bright P. Seyberth HW, Rane A, Schwab M, et al. Ethical considerations in conducting pediatric research. Pediatric Clinical Pharmacology
1st ed.Berlin/Heidelberg:Springer-Verlag; 2011. 219–244.
20. Paul S, Del TE, Marotte H, et al. Therapeutic drug monitoring of infliximab and mucosal healing in inflammatory bowel disease: a prospective study. Inflamm Bowel Dis
© 2014 by European Society for Pediatric Gastroenterology, Hepatology, and Nutrition and North American Society for Pediatric Gastroenterology,