Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 111
5 Safety and Efficacy Assessments in Studies Conducted Under BPCA and PREA T he goal of the Best Pharmaceuticals for Children Act (BPCA) and the Pediatric Research Equity Act (PREA) is to improve pediatric therapeutics through preclinical and clinical studies of drugs and bio- logics that are prescribed for children or that have the potential to benefit children. Ideally, such studies lead to the addition of useful information to the labeling of these products and then to the effective dissemination and application of that information to improve clinical care and child health. BPCA and PREA are components of a complex system for ensuring the drugs for children and adults are safe and effective. The Food and Drug Administration (FDA) and its statutory and regulatory foundations are central elements of this system. As summarized in Chapters 3 and 4, FDA not only assesses and monitors the safety and effectiveness of drugs but also requires protections for adults and children who participate in the trials that are the basis for agency assessments. The agency’s effectiveness in its multiple roles depends on science-based decision making, credible leadership, committed and well-trained staff, adequate financial resources, and timely and trustworthy communication to professionals and the public (FDA Science Board, 2007; IOM, 2007). Beyond FDA, the system for ensuring safety and efficacy extends to the organizations and individuals responsible for conducting drug studies and protecting research participants and research integrity. It thus includes commercial and other sponsors of research, clinical investigators, and in- stitutional review boards (IRBs), as well as health services researchers and others who analyze medication use in clinical practice in an effort to im- prove the quality, effectiveness, and efficiency of health care. The system 111
OCR for page 112
112 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN also encompasses clinicians who consider available evidence about drug safety and effectiveness as they care for children. Parents have a role, too, including in drug research when they administer test drugs or placebos at home and keep diaries or other records necessary for the assessment of safety and efficacy outcomes. This chapter discusses selected aspects of FDA’s assessments of the safety and efficacy of drugs and biologics based on data from pediatric studies requested under BPCA or required under PREA. For safety, these as- pects include reviewer conclusions about overall safety signals, risk-benefit assessments, and extrapolation of safety and findings of the 1-year safety reviews first required in BPCA of 2002. For efficacy, the discussion focuses on the use of alternative endpoints and extrapolation. SOURCES OF INFORMATION ABOUT SAFETY AND EFFICACY RESULTS IN PEDIATRIC DRUG STUDIES The most comprehensive perspective on the pediatric study data sub- mitted by sponsors and evaluated by FDA is provided in the clinical re- views prepared by staff of the Center for Drug Evaluation and Research (CDER) or the Center for Biologics Evaluation and Research (CBER). For this report, these reviews were the primary source of information on the characteristics and findings of pediatric studies conducted under BPCA or PREA. The committee also consulted clinical pharmacology and statistical reviews (if any), product labeling, and letters describing FDA’s approval action and any further requirements (e.g., further pediatric studies). FDA managers may prepare memoranda that provide additional context for decisions or explain why a reviewer’s recommendations were not accepted. For some labeling changes, the committee consulted minutes from FDA advisory committee meetings. Following congressional directives described in Chapter 3, CDER and CBER now post the reviews for products approved on or after September 27, 2007.1 For products approved earlier, clinical and other reviews are posted for a few products, but the committee had to request that FDA make public the reviews for most products approved before September 2007. (Appendix A describes how the committee selected the sample of requests, studies, and labeling changes assessed in this report.) As described by CDER, the clinical review (sometimes called the medi- cal review) is a “comprehensive summary and analysis of the clinical data 1 For CDER and CBER respectively, the reviews posted after September 26, 2007, are at http://www.fda.gov/Drugs/DevelopmentApprovalProcess/DevelopmentResources/ucm049872. htm and http://www.fda.gov/AboutFDA/CentersOffices/OfficeofMedicalProductsandTobacco/ CBER/ucm122938.htm.
OCR for page 113
113 SAFETY AND EFFICACY ASSESSMENTS submitted in support of a marketing application . . . [that] also includes the clinical reviewer’s assessment of and conclusions about: (1) the evidence of effectiveness and safety under the proposed conditions of use; (2) the adequacy of the directions for use; and (3) recommendations on regulatory action based on the clinical data submitted by an applicant” (CDER, 2010, p. 3). Clinical reviews may summarize findings from other areas of scientific review (e.g., toxicology and microbiology), and reviewers may also cite their own literature searches. In the years since BPCA and PREA and their predecessor policies went into effect, FDA has improved the organization and completeness of the clinical reviews. In 2004, CDER added to its policy manual a standard- ized template for clinical reviews, although some reviewers had been using a similar format for some time. Box 5-1 shows the major headings of the CDER template as revised in 2010. (Details of the safety and efficacy sec- tions of the template are presented later in this chapter.) CDER has also created templates for clinical pharmacology and biopharmaceutics reviews and for statistical reviews. In addition, CDER has created a 65-page desk reference guide that provides staff with an accessible resource of principles and procedures (CDER, 2011a). The guide also describes the roles of review team members, including those with specialized expertise (e.g., pediatrics) who may be included as needed. As described in the desk reference guide, the primary audience for the clinical review includes the review team (i.e., those with responsibility for BOX 5-1 CDER Template for Clinical Reviews (2010) 1. Recommendations/Risk-Benefit Analysis 2. Introduction and Regulatory Background 3. Ethics and Good Clinical Practices 4. Significant Efficacy/Safety Issues Related to Other Review Disciplines 5. Sources of Clinical Data 6. Review of Efficacy 7. Review of Safety 8. Postmarketing Experience 9. Appendices 9.1 Literature Review/References 9.2 Labeling Recommendations 9.3 Advisory Committee Meeting SOURCE: CDER Manual of Policies and Procedures, 6010.3R (issued December 14, 2010).
OCR for page 114
114 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN various aspects of the overall review), division staff, and CDER managers. The guide notes that reviewers should anticipate “the availability of the document to a public audience” (CDER, 2010, p. A-1). The Center for Biologics Evaluation and Research (CBER) has not com- pleted work on a standard format for reviews (personal communication, Catherine Lee, Office of Pediatric Therapeutics, FDA, August 8, 2011). Some CBER reviewers have, however, used an outline format similar to that used in CDER reviews. In general, the committee found that CBER reviews were more variable than CDER reviews. Overall, the committee found that the FDA reviews from recent years tended to be more systematic and focused than earlier reviews. The recent reviews were more likely to highlight key conclusions about safety and ef- ficacy, although they did not invariably follow the template. (Reviews may not follow the template for submissions that involve only pharmacokinetic and limited safety data, as requested by FDA.) Recent reviews also tended to provide more regulatory and other context about the origins and ratio- nales for studies. Occasionally, the reviews summarize interactions between the FDA and sponsors and provide insights into how and why studies changed over time. ASSESSING AND MONITORING SAFETY IN PEDIATRIC DRUG STUDIES: SELECTED ISSUES A sponsor’s submission of a new drug application (NDA) or biologics license application (BLA) will generally report safety data from preclini- cal and clinical studies and offer the sponsor’s assessments of these data. Submissions may also include data from adult pharmacokinetic and other studies, a review of relevant literature, and postmarket safety reports for already marketed drugs. As noted in Chapter 7, almost 10 percent of label- ing changes attributed to studies requested under BPCA or required under PREA involved no information from new pediatric studies. During the course of a clinical trial, the sponsor is responsible for trial monitoring. Depending on the anticipated risks in a trial (usually a Phase III trial), the sponsor may appoint a data monitoring committee (DMC; some- times called a data safety monitoring board or data and safety monitoring committee) to evaluate safety data as it accumulates.2 If a DMC identifies serious safety concerns in interim assessments of trial data, it can recom- 2 As described in FDA guidance, a DMC is “a group of individuals with pertinent expertise that reviews on a regular basis accumulating data from one or more ongoing clinical trials. The DMC advises the sponsor regarding the continuing safety of trial subjects and those yet to be recruited to the trial, as well as the continuing validity and scientific merit of the trial” (CDER/CBER/CDRH, 2006, p. 1).
OCR for page 115
115 SAFETY AND EFFICACY ASSESSMENTS mend modification or early termination of a trial. It is the sponsor’s respon- sibility to report serious adverse events and DMC recommendations related to such events to FDA. Unlike the National Institutes of Health (NIH), FDA regulations do not require the appointment of a DMC except in rare circumstances (CDER/CBER/CDRH, 2006). However, CDER’s template for written requests includes the option for the agency to require a DMC under other circumstances (CDER, 2011c).3 The Pediatric Review Com- mittee (PeRC, described in Chapter 3) discusses whether a DMC should be required, and FDA may place a clinical hold on a protocol if it concludes that the absence of a DMC puts research participants at unreasonable and significant risk (personal communication, Robert Nelson, Office of Pedi- atric Therapeutics, FDA, January 16, 2012). An assessment of the use of DMCs in pediatric clinical trials was beyond the task for the committee but may warrant future examination. CDER Template for Review of Safety in Drug Studies The CDER template for clinical reviews outlines a comprehensive eval- uation and discussion of safety that covers key topics and data sources in a systematic order (Box 5-2). One subsection of the template provides for a discussion (if relevant) of pediatrics and assessment of effects on growth. In practice, reviewers may tailor the format of their assessments to take into account the specifics of a particular submission, for example, whether it presents only a pharmacokinetic and pharmacodynamic study, as requested by FDA. Similarly, although the agency prefers that analyses pool data across studies, some sponsor submissions may not support this strategy. New Rules to Improve Reporting of Adverse Events and Analysis of Safety Data from Clinical Trials Central to the assessment of drug safety are the identification and evaluation of adverse events both during clinical trials and after marketing approval. FDA regulations define an adverse event as “any untoward medi- cal occurrence associated with the use of a drug in humans, whether or not considered drug related” (21 CFR 312.32(a)). Such an event can involve, for example, a laboratory or other test result, a symptom, a hospitaliza- 3 For example, in 2005, FDA requested a study of griseofulvin (an off-patent drug ap- proved for treatment of tinea capitis in children 2 years of age or older) to provide more data on pharmacokinetics, safety, and efficacy related to different dosing recommendations. FDA stated that a “Data Monitoring Committee with pertinent expertise must be used to provide ongoing oversight of patient safety” (Beitz, 2005, p. 4).
OCR for page 116
116 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN BOX 5-2 Safety Review Section of CDER Clinical Review Template (2010) Safety Summary Methods Studies/Clinical Trials Used to Evaluate Safety Categorization of Adverse Events Pooling of Data Across Studies/Clinical Trials to Estimate and Compare Incidence Adequacy of Safety Assessments Overall Exposure at Appropriate Doses/Durations and Demographics of Target Populations Explorations for Dose Response Special Animal and/or In Vitro Testing Routine Clinical Testing Metabolic, Clearance, and Interaction Workup Evaluation for Potential Adverse Events for Similar Drugs in Drug Class Major Safety Results Deaths Nonfatal Serious Adverse Events Dropouts and/or Discontinuations Significant Adverse Events Submission-Specific Primary Safety Concerns Supportive Safety Results Common Adverse Events Laboratory Findings Vital Signs Electrocardiograms (ECGs) Special Safety Studies/Clinical Trials Immunogenicity Other Safety Explorations Dose Dependency for Adverse Events Time Dependency for Adverse Events Drug-Demographic Interactions Drug-Disease Interactions Drug-Drug Interactions Additional Safety Explorations Human Carcinogenicity Human Reproduction and Pregnancy Data Pediatrics and Assessment of Effects on Growth Overdose, Drug Abuse Potential, Withdrawal, and Rebound Additional Submissions/Safety Issues SOURCE: CDER Manual of Policies and Procedures (Section 7 of Clinical Review Template), 6010.3R (issued December 14, 2010).
OCR for page 117
117 SAFETY AND EFFICACY ASSESSMENTS tion, or a death. An adverse reaction is an adverse event that is attributed to use of the drug. In 2010, FDA issued new regulations and guidance on safety report- ing for clinical trials (CDER/CBER, 2010a). The goal was to “increase the interpretability of and usefulness of safety data available to the clinical investigators, IRBs, and the FDA” (Sherman et al., 2011, p. 5). The rules require clinical investigators to report all serious adverse events to trial sponsors. They shift the responsibility for assessing whether an isolated adverse event is likely to be drug related from individual investigators to sponsors. As a result, sponsors should have a larger and more complete pool of data to support assessments of causality. These assessments should improve the relevance of their reports to FDA. The 2010 rules also offered several examples of the kinds of events on which sponsors should focus. They include the following: • A single occurrence of an event that is uncommon and known to be strongly associated with drug exposure (e.g., angioedema, hepatic injury, and Stevens-Johnson syndrome); • One or more occurrences of an event that is not commonly associ- ated with drug exposure but that is otherwise uncommon in the population exposed to the drug (e.g., tendon rupture); and • An aggregate analysis of specific events observed in a clinical trial (such as known consequences of the underlying disease or condi- tion under investigation or other events that commonly occur in the study population independent of drug therapy) that indicates that those events occur more frequently in the drug treatment group than in a concurrent or historical control group. (CDER/CBER, 2010b, p. 4) IOM Review of Safety Assessments in Pediatric Drug Studies As explained in Chapter 1, the Institute of Medicine (IOM) was asked to assess “the number and type of pediatric adverse events” in a sample of studies conducted under PREA or precursor regulations. The committee also included a sample of studies stemming from requests under BPCA. This broader scope provided additional context for understanding FDA’s evaluation of safety findings in pediatric drug studies. Unfortunately, the FDA clinical reviews examined by the committee were completed before FDA’s shift to the new, more targeted strategy for reporting adverse events. The typical clinical review included numerous, sometimes lengthy tables and reports of various categories of adverse events that correspond to topics in the review template. The reviews focused on
OCR for page 118
118 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN serious and unexpected adverse events reported in clinical trials, but they also discussed less serious events. The sponsor and FDA reviewers judged many adverse events described in the clinical reviews not to be related to the test product. Given the thoroughness of most reviews and the usual judgment that a substantial proportion of reported adverse events were not related to the test drug, the IOM committee decided that it would not be productive to re- view and assess the numbers and types of these events. Instead of counting and categorizing individual adverse events, the committee focused on the clinical reviewer’s more general and relevant conclusions about a product’s safety profile. For example, for products that had been studied in adults, did the FDA reviewer conclude that pediatric studies of a drug or biologic showed a safety profile that was similar to that reported for adults? Alter- natively, did the profile for children differ from that for adults in ways that, at a minimum, warranted discussion in the product’s labeling? If the FDA reviewer did not compare pediatric safety findings to adult safety findings, did he or she make other appropriate comparisons (e.g., with findings for a control group or with safety findings in other pediatric studies of similar drugs for the same condition)? Because safety is relative, FDA must weigh findings about the risks of a product against expected benefits and judge whether the expected benefits sufficiently outweigh expected harms to justify approval for marketing. (FDA may disapprove the labeling of a product for pediatric use but pro- vide for the addition of safety or other information from pediatric studies to the product labeling for already marketed products.) In assessing clinical reviews, the committee looked for a risk-benefit assessment (to use FDA’s language), that is, an explicit overall judgment about risks in relation to expected benefits. In some cases, the committee found that a reviewer’s discussion of the risk-benefit assessment was redacted without explana- tion. A memo from a division director or review team leader sometimes indicated that agency management reached different conclusions from the primary reviewer. The committee initially intended to assess the extent to which label- ing changes were consistent with the reviewer’s conclusions about safety signals or significant adverse events. However, after discovering that FDA generally redacted all or much of the discussion of labeling in clinical re- views, the committee decided that it could not be confident in making such assessments. As discussed in Chapter 3, the sponsor owns the label, and new labeling or changes in labeling usually result from a process involving negotiation between the sponsor and FDA about the sponsor’s proposed wording.
OCR for page 119
119 SAFETY AND EFFICACY ASSESSMENTS Analysis of Safety Profile For products that had also been studied in adults, most clinical reviews that the committee examined offered relatively straightforward and easily understood conclusions about whether the safety findings from pediatric studies showed results similar to those found from adult studies. The ma- jority of the reviews that included comparisons of the results for children with the results for adults (or with the known safety profile of the product) concluded that the safety profile was similar for children. For example, in the assessment of leflunomide (Arava) for the treatment of juvenile rheu- matoid arthritis, the reviewer’s summary conclusion was that the “overall profile of adverse events was consistent with the underlying disease and known serious adverse events of leflunomide” (Yancey, 2004, p. 68). The summary also notes hepatotoxicity to be a known risk of the drug. To cite another example, the clinical reviewer for tenofovir disoproxil fumarate (Viread) noted that “[o]verall, the safety issues identified in the adolescent study are similar to those previously identified in the adult clinical trials and are included in the current product label” (Levorson, 2010, p. 40). The reviewer then described several specific safety issues, including reductions in bone mass density, renal toxicity, and gastrointestinal events. For one prod- uct (eletriptan hydrobromide [Relpax]), the labeling—but not the redacted clinical review—stated that the profile of adverse events in a pediatric study was similar to that reported in studies with adults. Because one objective of FDA’s evaluation of adverse events in pediatric studies is to determine whether a product’s labeling needs to be revised, reviewers sometimes explicitly noted whether the findings about treatment- related adverse events in children were reflected in the existing labeling (for previously approved products) or whether some revisions were needed. As already noted, reviewers’ specific discussions of the text of proposed label- ing were mostly or entirely redacted. For some products, reviewers found different safety signals, usually in the form of events that, although expected, were more common in children than in adults. In a few instances, the findings were unexpected on the basis of the data for adults. Box 5-3 provides examples of these kinds of reports. Some drugs were studied in populations and for indications that did not lend themselves to comparisons with the findings of studies with adults. An example is nitric oxide (INOmax) for the treatment of neonates with bronchopulmonary dysplasia, a condition not diagnosed in adults. Even when comparisons with adult safety findings were possible, some review- ers chose to make other informative comparisons. To cite an example, in the clinical review of a combination salmeterol xinafoate and fluticasone propionate product (Advair Diskus), the comparison was with the safety
OCR for page 120
120 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN BOX 5-3 Examples of Products with Different Safety Profiles for Children and Adults Identified in FDA Clinical Reviews Adalimumab (Humira) for treatment of juvenile idiopathic arthritis. “Safety was similar to that seen in adults but there were several safety signals not observed in the adults, including elevations of creatine phosphokinase (CPK). In addition, a higher rate of immunogenicity was observed in children as compared to adults as well as a higher rate of non-serious hypersensitivity reactions. There was disagreement between the primary clinical reviewer and the secondary reviewer on the specific details of the post-marketing registry that should be conducted” (Siegel, 2008b, p. 3). Aripiprazole (Abilify) for treatment of schizophrenia. “Based on a comparison of the results of five short-term adult studies in schizophrenia with the results of this pediatric schizophrenia study, the safety profile of aripiprazole in adolescents with the diagnosis of schizophrenia is comparable to the adult schizophrenia population, with the exception of dose-related occurrence of higher frequency of somnolence and extrapyramidal symptoms observed in the pediatric population” (Zhang, 2008, p. 35). Desflurane (Suprane) for induction or maintenance of anesthesia. “The clinical data submitted in this supplement demonstrated a marked increase in the inci- dence of both major (associated with significant oxygen desaturation) and minor respiratory events including laryngospasm, airway obstruction, secretions, and breath holding in non-intubated pediatric patients who underwent maintenance anesthesia with desflurane compared to a cohort of children treated similarly with isoflurane. The incidence of these respiratory events appeared to be related to the inspired concentration of desflurane. These data do not support the use of desflurane for induction (which was a prior finding) or maintenance of anesthesia in non-intubated children” (Shibuya, 2006, p. 4). Olmesartan (Benicar) for treatment of hypertension. “In this whole study program, transient minor to moderate headache was the major adverse event with this prod- uct in pediatric population. Other than that, there does not appear to be any other unexpected adverse events in children compared to adults” (Xiao, 2009, p. 10). Omalizumab (Xolair) for treatment of asthma. “[I]n patients 6–11 years of age with IgE [immunoglobulin E] levels above 500 IU/mL, circulating trough levels of omali- zumab and omalizumab-IgE complexes are higher than those achieved in patients 12 years of age and older with IgE levels up to 700 IU/mL. These complexes take months to clear after termination of Xolair treatment. Although no urinary abnormalities or evidence of serum sickness was noted in the safety database, the clinical meaning of higher circulating immune complex exposure, particularly over many years of chronic exposure, is unknown. Thus, lack of evidence support- ing the long-term safety of a dosing regimen associated with circulating immune complex levels that are higher in children higher [sic] than those studied and approved in adults is a safety concern with this application” (Starke, 2009, p. 12).
OCR for page 121
121 SAFETY AND EFFICACY ASSESSMENTS profiles for the individual components of the product, which were similar to those for the combination product (Johnson, 2000). For some products, the comparison was with previously studied products or formulations. In the review of mometasone furoate (Asmanex) for treatment of asthma, for example, the reviewer noted that the adverse events identified were com- mon and consistent with those found in other trials of similar drugs in pediatric patients “and do not suggest a new safety signal” (Karimi-Shah, 2007, p. 11). In reaching overall conclusions about safety, some reviewers did not make comparisons with other populations or products. For example, the reviewer for alendronate (Fosamax) for osteogenesis imperfecta stated that “the safety and tolerability profile of alendronate in this population were acceptable, with few serious adverse events (only three of which were pos- sibly related to alendronate) and no deaths” (Schneider, 2003, p. 3). The reviewer also noted one case of leukopenia—a condition not identified to be a risk for adults—and suggested that that this type of event be monitored as a safety issue, regardless of whether FDA approved the drug for treatment of the studied indication. Some reviews stated only that no unexpected adverse events had been noted. In context, such statements probably can be interpreted as suggesting that the safety profile was similar to that for adults if the product had been previously studied in adults. For example, in an assessment of irinotecan hydrochloride (Camptosar) for refractory solid tumors, a clinical reviewer concluded that the pediatric studies provided no meaningful new safety information (Ibrahim, 2003). Although reviewers differed in how they summarized and presented the information, the reviews typically supplemented the overall assessment of safety with a summary of serious adverse events that are considered to be related to the drug and a summary of common treatment-related adverse events. One example of a clear, relatively brief summary of such adverse events is provided in the clinical review of a sponsor submission involving almotriptan (Axert) for the treatment of migraine in adolescents (Harris, 2009). In three short paragraphs, the reviewer notes that 67 percent of the study participants had some kind of adverse event (all causality), that 8 percent had an adverse event that was judged to be related to the product, that these events were most often nausea and somnolence (each reported by 1.4 percent of participants), and that 2 percent of participants experienced a serious adverse event, none of which was judged to be treatment-related. Risk-Benefit Assessment Explicit statements of risks in relation to benefits usefully underscore the reality that the use of drugs involves the potential for harm as well as
OCR for page 130
130 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN endpoint such as bone mass density that is used in place of an endpoint or outcome that is more directly meaningful to patients, such as a bone frac- ture.7 In studies of drugs to treat osteoporosis in adults, the rate of fractures is the primary efficacy measure. In pediatric studies of the same drugs to treat low bone mass in osteogenesis imperfecta, FDA has specified change in bone density (a surrogate measure) to be the primary endpoint; fracture rate is one of several secondary endpoints (see, e.g., Schneider, 2003). (In studies focusing on prevention rather than the treatment of osteoporosis in adults, a bone density measure has been a primary endpoint.) Consultants from CDER’s Study Endpoint and Labeling Development Group may be involved in consultations about pediatric endpoints without being cited in clinical reviews. The group is also involved in the process that FDA created to evaluate and qualify biomarkers, patient-reported outcome tools, and other measures that sponsors may use in specific drug develop- ment efforts so that the appropriateness of each such use does not have to be individually evaluated (CDER, 2010). In addition, FDA may support analyses of alternative or other end- points in various contexts. For example, in the context of an advisory committee discussion of modifications to a 2001 written request for the study of sildenafil (Revatio) for the treatment of pediatric hypertension, a staff member from CDER’s Office of Biopharmacometrics discussed data on the use of a hemodynamic measure (the pulmonary vascular resistance index) as an alternative to the 6-minute walk test used for adults (Brar, 2010; CRDAC, 2010). As noted in Chapter 2, alternative endpoints may be used in pediatric studies in several circumstances. These include when • the use of the adult endpoint is impossible, for example, when that endpoint depends on a pulmonary function test that cannot be reli- ably performed by young children or when it requires self-reporting of symptoms and the children to be studied are preverbal; • the use of the adult endpoint is too risky given the circumstances, for example, when a measurement process used only for research purposes (such as an evaluation by magnetic resonance imaging that has no prospect of benefit) requires a research participant to remain still and would require sedation for children in the age group to be studied; 7 As defined elsewhere by an NIH working group, a surrogate measure is a “biomarker that is intended to substitute for a clinical endpoint. A surrogate endpoint is expected to predict clinical benefit (or harm or lack of benefit or harm) based on epidemiologic, therapeutic, pathophysiologic, or other scientific evidence” (Biomarkers Definitions Working Group, 2001, p. 91).
OCR for page 131
131 SAFETY AND EFFICACY ASSESSMENTS • the condition being studied has somewhat different manifestations in children (e.g., juvenile rheumatoid arthritis versus adult rheu- matoid arthritis); • the adult endpoint involves measures of common social interac- tions or functioning (e.g., at work) that do not reflect children’s situations; and • a before-and-after treatment measure could be affected by children’s development as well as treatment-related change (e.g., change in bone mass density). Results of Committee Assessments The clinical and other reviews and the written requests that the com- mittee examined usually did not note whether the endpoints used for pediatric studies were different from the endpoints used for adult studies. They likewise typically did not discuss the rationale for the endpoints. In some cases, the committee consulted descriptions of studies with adults to determine whether different endpoints were used in the pediatric studies. For the sample of requested or required pediatric studies and labeling changes that the committee examined, almost half (23 of 49) used pri- mary efficacy endpoints that were the same as those used in adult studies. Roughly one-fifth (11 of 49) involved alternative endpoints. For one prod- uct for which two primary endpoints were specified, one of the endpoints was also used in studies with adults and the other was an alternative end- point. In most of the remaining cases, the studied indications were primar- ily or entirely found in the pediatric population (seven cases) or primary efficacy endpoints were not required or requested (six cases). Three of the 49 product assessments involved efficacy studies that had different primary efficacy endpoints for different age groups. For one efficacy study (for moxi- floxacin ophthalmic [Vigamox] for the treatment of bacterial conjunctivitis in neonates), the section of the clinical review that presumably described the endpoint and results was redacted. Box 5-5 presents examples of the different categories of endpoints re- ported in the clinical reviews that the committee examined. More than one indication could be evaluated for a single product, and different efficacy endpoints could be used for different age groups. For the most part, FDA reviewers did not raise concerns about the use of alternative endpoints as such. Some reviewers noted that the endpoints were based on measures validated for the indication and age group studied (see, e.g., Siegel, 2008b). In general, it would be desirable for specification of alternative endpoints to be accompanied by some discussion of evidence supporting their reliability and validity. For several studies of asthma drugs, the committee had concerns about
OCR for page 132
132 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN BOX 5-5 Examples of Efficacy Endpoints in Pediatric Studies Alternative Endpoint Adalimumab (Humira) Indication: juvenile rheumatoid arthritis Primary efficacy endpoint: disease flare measured by a 30 percent worsen- ing in at least three of six juvenile rheumatoid arthritis core set criteria and a minimum of two active joints AND 30 percent improvement in not more than six juvenile rheumatoid arthritis core set criteria specified by American College of Rheumatology (Siegel, 2008b) Alendronate (Fosamax) Indication: osteogenesis imperfecta Primary efficacy endpoint: change in lumbar spine bone mass density (BMD) Z-score (standard deviations from the mean for age-matched healthy controls) from baseline (Schneider, 2003) Buspirone hydrochloride (Buspar) Indication: generalized anxiety disorder (ages 6 up to 17 years) Primary efficacy endpoint: change from baseline in the sum of four scores from C KSADS GAD (Columbia Kiddie Schedule for Affective Disorders and Schizophrenia–General Anxiety Disorder scale) that are specific to anxiety (Laughren, 2000) Endpoint Also Used in Adult Studies Aripiprazole (Abilify) Indication: schizophrenia (ages 13 up to 17 years) Primary efficacy endpoint: Positive and Negative Syndrome Scale (Zhang, 2007) Hydrocortisone butyrate (Locoid) Indication studied: atopic dermatitis (ages 3 months or older) Primary efficacy endpoint: Physician’s Global Assessment score (Katz, 2007) Other (Primarily or Entirely a Pediatric Condition) Methylphenidate (Concerta) Indication studied: attention deficit hyperactivity disorder (ages 6 up to 12 years) Primary efficacy endpoint: IOWA Conners Teacher Rating Scale (Inattention/ Overactivity Subscale) (Mosholder, 2000) the endpoint specified for studies in children ages 4 to 11 years. The end- point, forced expiratory volume in 1 second (FEV1), is widely accepted for use with adults and older children, but it requires physical maneuvers that children under age 6 years cannot reliably perform (see Chapter 2). As a result, for levalbuterol hydrochloride [Xopenex inhalation] for the treat- ment of asthma, FDA approved labeling for use only in the age group 6 to 11 years old, even though the requested study was supposed to assess drug
OCR for page 133
133 SAFETY AND EFFICACY ASSESSMENTS safety and efficacy in the age group 4 to 11 years old. Three other products (albuterol sulfate [Ventolin HFA], levalbuterol tartrate [Xopenex HFA], and salmeterol xinafoate [Advair Diskus]) were approved for children in the age group 4 to 11 years old, but on the basis of data that were less than adequate for the youngest children in this group. At least one requested study of asthma in a younger age group (birth up to 4 years of age for albuterol sulfate [Ventolin HFA]) reported the use of alternative endpoints. These involved asthma symptom scales that used parents’ assessments of symptoms (cough, wheeze, and shortness of breath) in one trial and a clinician assessment using the Modified Tal Asthma Symptoms score, which “included components of respiratory rate, wheez- ing, cyanosis, and accessory respiratory muscle utilization” (Wang, 2008, p. 14). (In an Internet search, the committee did not find an assessment of the latter instrument.) For the assessment of studies of pantoprazole sodium (Protonix) for the treatment of gastroesophageal reflux disease (GERD), the clinical review ex- plicitly noted that different symptoms in different age groups required dif- ferent efficacy endpoints (Chen, 2009). The reviewer also noted concerns, expressed by a consultant from the agency’s study endpoints and labeling development team, about the appropriate description for labeling purposes of measures for infants (vomiting/regurgitation, irritability/fussiness, refusal to feed, choking/gagging, arching back) that were observer (parent) rather than patient based. The consultant also observed that the sponsor did not discuss translation or cultural adaptation of the measures for infants, even though the trial had sites in six countries other than the United States. On occasion, FDA and a sponsor may not identify a measure suitable for a specific age group, and FDA may waive studies required under PREA for that group. For example, when FDA approved dextromethorphan hy- drobromide and quinidine sulfate (Nuedexta) for the treatment of pseu- dobulbar affect, it waived required studies with children less than 2 years of age. The approval letter explained that the condition “involves exag- gerated or contradictory episodes of laughing or crying given the patient’s actual emotional state” and “verbal and non-verbal communication is not adequately developed [in this age group] to allow for accurate appraisal of the patient’s actual emotional state” (Katz, 2010, p. 3). Use of Extrapolation Chapter 1 described the FDA initiative in the early 1990s to increase pediatric studies. Among other steps, FDA allowed, under certain circum- stances, the extrapolation of efficacy findings from studies with adults to children. Specifically,
OCR for page 134
134 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN a pediatric use statement may also be based on adequate and well con- trolled studies in adults, provided that the agency concludes that the course of the disease and the drug’s effects are sufficiently similar in the pediatric and adult populations to permit extrapolation from the adult efficacy data to pediatric patients. Where needed, pharmacokinetic data to allow determination of an appropriate pediatric dosage, and additional pediatric safety information must also be submitted. (59 FR 64240 at 64241) In 2007, FDAAA added that “a study may not be needed in each pediatric age group if data from one age group can be extrapolated to another age group” (21 USC 355C(a)(2)(B()ii)).8 Allowance for the use of extrapolation is intended to make pediatric drug studies less onerous and thereby increase the number of such studies undertaken. Although the allowance in 1994 for extrapolation of efficacy to pediatric age groups had little effect on its own as a stimulus to pediatric studies, it became more significant after Congress created the incentives and requirements for pediatric studies under BPCA and PREA and their predecessor policies. Decision Tree for Extrapolation Decisions Working from the regulatory framework described above, FDA has developed a decision tree to guide determinations about when extrapolation can be permitted (Figure 5-1). The determinations can differ by age groups (e.g., with extrapolation accepted for adolescents but not for younger children). As interpreted by FDA, the extrapolation decision is not a simple “allow” or “do not allow” decision. FDA must also specify the extent to which extrapolation can be relied upon for determinations about efficacy. In guidance issued in 1998, FDA stated that evidence relevant to the de- terminations about similarity of disease course and disease effect included “evidence of common pathophysiology and natural history of the disease in the adult and pediatric populations, evidence of common drug metabo- lism and similar concentration-response relationships in each population, and experience with the drug, or other drugs in its therapeutic class, in the disease or condition or related diseases or conditions” (CDER/CBER, 1998b, p. 8). Occasionally, the written requests or FDA clinical reviews that the 8 In addition to the FDA provisions for extrapolation that were explicitly directed at pediatric studies, FDA also has more general authority to determine effectiveness based on “data from one adequate and well-controlled clinical investigation and confirmatory evidence (obtained prior to or after such investigation)” (21 USC 355(d)). That is, legislation provides for one form of what FDA terms partial extrapolation to be used by sponsors to support the labeling of products for adult uses.
OCR for page 135
135 SAFETY AND EFFICACY ASSESSMENTS Is it reasonable to assume that children, when compared to adults, have a similar: (a) disease progression? (b) response to intervention? No Yes to both Is it reasonable to assume a similar exposure- response (ER) in children when compared to adults? No Yes Is there a pharmacodynamic Conduct pharmacokinetic (PK) (PD) measurement that can studies to achieve drug levels predict e cacy in children? similar to adults, then safety trials at the correct dose No Yes Conduct PK/PD studies to Conduct PK studies to establish dose; establish an ER in children then conduct pediatric safety and for the PD measurement; e cacy trials conduct PK studies to achieve target concentrations based on ER; then conduct safety trials at the correct dose FIGURE 5-1 Use of extrapolation to support pediatric efficacy claims. SOURCE: Dunne (2010). Figure 5-1.eps committee assessed used the language presented in the decision tree to acknowledge the use of extrapolation. Only rarely did a written request or FDA clinical review provide a more substantive explanation with references to the scientific literature to justify decisions to allow extrapolation. One example of a justification with explicit citation to the literature appears in the written request for a study of aripiprazole (Abilify) for the treatment of schizophrenia in adolescents:
OCR for page 136
136 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN Under FDAMA, 1997, adequate assessment of adolescents (data sufficient to support a labeling claim) might be based on a single study in pediatric patients, together with confirmatory evidence from another source, per- haps adult data for that disorder. . . . This approach too requires that the adult data be considered reasonably relevant to the course of the disease and the effects of the drug in the pediatric populations. Although we are aware of only two published placebo controlled studies supporting the ef- ficacy of neuroleptics (haloperidol & loxitane) in the treatment of pediatric schizophrenia . . . we believe that a sufficiently strong case has been made for continuity between adult and adolescent schizophrenia to permit a pediatric claim for a drug already approved in adults to be supported by a single, independent, adequate and well-controlled clinical trial in adoles- cent schizophrenia. In addition, a pediatric schizophrenia program would need to include pharmacokinetic information and safety information. . . . Finally, although we are requiring only certain specific studies, you will be expected to maximize the potential of the studies to demonstrate an effect of the drug in adolescents, if there is one. Toward this end, then, we urge you to perform additional studies (see below) in order to ensure that the required studies meet this goal. (Temple, 2003, pp. 6–7) FDA requests and reviews have become somewhat more consistent in providing justification for extrapolation. Such justifications are often lim- ited in their descriptions and citations of relevant literature. FDAAA speci- fies only that “a brief documentation of the scientific data” supporting a conclusion about the use of extrapolation be included in agency reviews (21 USC 355c(a)(2)(B)(iii)). Nonetheless, given the significance of the reliance on extrapolation, it would be desirable for requests and reviews to provide the public with a justification somewhat fuller than that now provided in each case in which the agency accepts full or partial extrapolation. Extent of Use of Extrapolation Recently, an FDA working group on extrapolation has developed a categorization scheme to label and describe the basic options for the use of extrapolation (Dunne et al., 2011b). The options include • No extrapolation of efficacy: FDA requires pharmacokinetic data and demonstration of safety and efficacy from two adequate, well- controlled pediatric trials (or from a sequential response and safety trial strategy for oncology products). • Partial extrapolation of efficacy from studies with adults (or other pediatric age group) with a controlled efficacy trial: FDA requires pharmacokinetic data and confirmation of efficacy and assessment of safety from one adequate and well-controlled pediatric trial.
OCR for page 137
137 SAFETY AND EFFICACY ASSESSMENTS • Partial extrapolation of efficacy from studies with adults (or other pediatric age group) without a controlled efficacy trial: FDA speci- fies other acceptable sources of pharmacokinetic, safety, and ef- ficacy or response data. • Complete extrapolation of efficacy from studies with adults with assessment of safety: FDA requires only safety data or requires safety and pharmacokinetic data to assess age-appropriate dosing. The working group on extrapolation also analyzed the use of extrapo- lation studies requested under BPCA based on NDA submissions received between February 1998 and February 2009. As shown in Table 5-2, 29 (17 percent) of 166 submissions of requested studies involved no extrapolation of efficacy, and 24 (14 percent) involved the complete extrapolation of efficacy. The modal submission (67 [40 percent]) included one controlled safety and efficacy trial with additional pharmacokinetic data (which could be obtained during the safety and efficacy trial). For the most part, the fewer the data on efficacy requested by FDA, the more likely it was that a later application for a new or expanded pediatric indication would be approved. The FDA analysis did not examine the use of extrapolation in studies required under PREA. For its sample, the committee examined FDA’s acceptance of extrapo- lation to support labeling changes resulting both from studies requested under BPCA and studies required under PREA. Because the use of extrapo- lation was often not mentioned explicitly in clinical reviews or other docu- ments, the committee had to infer FDA’s reliance on it. For this analysis, as for the one described above, the more extensive that FDA’s acceptance of extrapolation was, the more likely the agency was to approve labeling for a pediatric age group (Table 5-3). TABLE 5-2 FDA Analysis of Use of Extrapolation of Efficacy from Adult to Pediatric Population, Studies Conducted Under BPCA, 1998 to 2009 No. of Studies with Characteristic/ Total No. of Studies (%) Extrapolation of Efficacy from Use for Products with New/Expanded Pediatric Adults or Other Sources Written Request Indication Achieved No extrapolation (two WCTa) 29/166 (17) 10/29 (34) Partial extrapolation (one WCT) 67/166 (40) 35/67 (52) Partial extrapolation (other) 46/166 (28) 34/46 (74) Complete extrapolation 24/166 (14) 15/24 (62) a WCT indicates data required from an adequate well-controlled safety and efficacy clinical trial or, for oncology products, from a two-stage trial process to assess response and safety. SOURCE: Dunne et al. (2011b).
OCR for page 138
138 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN For written requests, FDA may reject a sponsor’s proposal for the use of partial extrapolation from adult studies. For example, in the case of the drug buspirone hydrochloride (Buspar), FDA wrote the sponsor, “While we acknowledge your . . . commitment to conduct two clinical trials for this indication, we do not believe that your new proposal to submit one com- pleted clinical study and one completed pediatric pharmacokinetic study, as a substitute for submitting two completed clinical studies, would be suffi- cient to support the safety and effectiveness [of the drug] . . . in the pediatric population and to qualify for pediatric exclusivity” (Temple, 1999, p. 1). As noted earlier, FDA may allow the use of extrapolation for one age group but not another. In a request for studies of pantoprazole (Protonix) for treatment of erosive esophagitis and nonerosive GERD, FDA con- cluded that efficacy could be extrapolated from adult data to children 1 to 17 years of age because pathophysiology was similar in the two groups. However, for children younger than age 1 year, the agency concluded that extrapolation was not acceptable because, as described in the clinical re- view, “the pathophysiology of GERD in infants is believed to be unique” and “symptomatology and prognosis differ between infants and individuals greater than age 1 year” (Griebel, 2009). Nonetheless, the agency did not request two well-controlled safety and efficacy studies for infants. Rather, it requested one such study and another pharmacokinetic, pharmacodynamic, and safety study (Raczkowski, 2001). According to FDA’s current scheme for categorizing determinations, the request allowed for the use of partial TABLE 5-3 Use of Extrapolation in IOM Sample of BPCA and PREA Labeling Changes No. (%) of Studies Indication Granted, Use of Extrapolation Extent of Use by Extent of Use Extrapolation not accepted (two WCT) 17/55 (31) 8/17 (47) Partial extrapolation accepted (one WCT) 26/55 (47) 15/26 (58) Partial extrapolation accepted (other data) 6/55 (11) 5/6 (83) Complete extrapolation accepted 1/55 (2) 1/1 (100) Other 5/55 (9) 1/5 (20) NOTE: Data are for 55 actions, including different decisions for different age groups. WCT indicates data required from an adequate well-controlled safety and efficacy clinical trial or, for oncology products, from a two-stage trial process to assess response and safety; other indicates that the study could involve various combinations of sources of pharamcokinetics, safety and efficacy, response, or activity data. The category “other” includes some submissions for which efficacy was not requested; one for which FDA stated that two WCTs were required but the sponsor only submitted one (which did not show efficacy); and one that included no new pediatric studies.
OCR for page 139
139 SAFETY AND EFFICACY ASSESSMENTS extrapolation on the basis of one safety and efficacy trial in the age group 1 month up to 1 year old. As it turned out, the studies did not support ef- ficacy in children in this age group (see Chapter 6). CONCLUSIONS In general, FDA reviewers were careful and thorough in identifying drug-related adverse events, assessing their significance, and reaching con- clusions about the safety profile of drugs evaluated in studies with children and the need for any changes in the safety elements of a product’s label- ing (if it was already labeled). Summary assessments of a product’s safety profile were generally accompanied by an identification of serious adverse events. The committee noted variations in the thoroughness of reviews, al- though recent reviews are generally more thorough and complete. To fur- ther improve the quality of reviews, the committee believes that it is time for CBER to adopt formally a systematic, standardized template for clinical and other reviews similar to that used by CDER. The committee also en- courages FDA divisions to continue to guide reviewers to follow the safety assessment template, to provide explicit statements about their risk-benefit assessments, and to state clearly their overall conclusions about a product’s safety profile and significant or common adverse events. If successfully implemented, the agency’s new guidance on safety re- porting for clinical trials should improve identification and assessment of treatment-related adverse events and thereby provide a better foundation for conclusions about a drug’s safety profile with pediatric use. Likewise, the structured benefit-risk assessment framework promised by the agency could make an important contribution to FDA’s assessments of pediatric drug studies. Pediatric studies of drug safety and effectiveness over the long term are important but not commonly requested or required. The 1-year safety reviews mandated by Congress appear to provide a useful opportunity for FDA to examine safety experience and to consider overall safety informa- tion after products have had labeling changes based on pediatric studies. In several instances, the reviews have led to revisions of safety information in product labeling or pending recommendations for such changes. Still, the lack of information about the long-term safety of drugs is a particular worry for developing children—both for drugs that may be used for decades for chronic conditions and for drugs for which short-term use may have adverse consequences months or years later. Given such concerns, FDA might more frequently use its expanded authority to require spon- sors to undertake postmarket, follow-up studies of drug safety in pediatric populations.
OCR for page 140
140 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN Although agency staff generally state that the agency does not accept the extrapolation of safety from studies with adult or other pediatric popu- lations, the committee found examples of such extrapolation. This may be appropriate in unusual circumstances, but a public explanation and justifi- cation of these circumstances is desirable. For the most part, FDA’s specification of efficacy endpoints appears to be reasonable, including the use of alternative endpoints when measures used for adults are not appropriate. Written requests and clinical reviews rarely discuss the rationale for endpoints, whether they are alternative or not. For alternative endpoints in particular, FDA should consider provid- ing an explicit discussion of their use, including whether they have been validated in studies with children in the age groups to be studied. FDA and sponsors rely extensively on extrapolation of efficacy, usually based on requirements for the submission of some efficacy, response, or activity information as well as pharmacokinetic and safety data. The com- mittee found that the justifications were often limited in their descriptions and citations of relevant literature, and Congress requires only brief docu- mentation for the use of extrapolation. Nonetheless, it would be desirable for requests and reviews to provide the public with a justification somewhat fuller than that now provided in each case in which the agency accepts full or partial extrapolation. The committee recognizes that providing the public with the additional justifications and explanations suggested here adds to the demands on agency staff. In some cases, internal documents (e.g., memoranda for PeRC meetings) or sponsor submissions may already provide much of the basis for such explanations. Overall, the committee believes that the significance of the judgments for which more explicit public rationales or justifications are suggested warrants the additional attention.