2

Children’s Growth and Development
and Pediatric Drug Studies

As context for later discussions of ethics, safety, and efficacy in pediatric studies, this chapter provides an overview of how children’s growth and development may affect their responses to medications. Medications that are generally safe and effective for adults may be unsafe or ineffective—or both—for some or all pediatric age groups or may require changes in dosing forms, calculations, or schedules to be safe and effective. Such disparities underscore the necessity for pediatric drug studies. This chapter also discusses how differences between children and adults may require alterations in the design, conduct, and analysis of such studies.

As a prelude to the rather technical discussion of developmental pharmacology, the chapter begins with an example of the sometimes fatal consequences of the lack of drug studies with children, especially the youngest children. The case involves an antibiotic that was used to treat neonates before its safety had been documented in that age group.

THE CASE OF CHLORAMPHENICOL

Chloramphenicol was discovered in the late 1940s and found to be effective against many different infections caused by a wide range of organisms such as salmonella and rickettsia (Meissner and Smith, 1979). The pharmacokinetics of chloramphenicol in children were reported in 1951 (Kelly et al., 1951).

During the 1950s, as pediatricians made increasing use of the drug to treat a variety of infections, the American Academy of Pediatrics (AAP) Committee on Infectious Diseases offered dosing recommendations for the



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



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 43
2 Children’s Growth and Development and Pediatric Drug Studies A s context for later discussions of ethics, safety, and efficacy in pe- diatric studies, this chapter provides an overview of how children’s growth and development may affect their responses to medications. Medications that are generally safe and effective for adults may be unsafe or ineffective—or both—for some or all pediatric age groups or may require changes in dosing forms, calculations, or schedules to be safe and effective. Such disparities underscore the necessity for pediatric drug studies. This chapter also discusses how differences between children and adults may require alterations in the design, conduct, and analysis of such studies. As a prelude to the rather technical discussion of developmental phar- macology, the chapter begins with an example of the sometimes fatal con- sequences of the lack of drug studies with children, especially the youngest children. The case involves an antibiotic that was used to treat neonates before its safety had been documented in that age group. THE CASE OF CHLORAMPHENICOL Chloramphenicol was discovered in the late 1940s and found to be effective against many different infections caused by a wide range of or- ganisms such as salmonella and rickettsia (Meissner and Smith, 1979). The pharmacokinetics of chloramphenicol in children were reported in 1951 (Kelly et al., 1951). During the 1950s, as pediatricians made increasing use of the drug to treat a variety of infections, the American Academy of Pediatrics (AAP) Committee on Infectious Diseases offered dosing recommendations for the 43

OCR for page 43
44 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN drug (Kempe, 1955). Most of the studies reviewed as a basis for the rec- ommendations included children and infants (some as young as 1 month) but no newborns. Then, in response to the increasing survival rates for premature newborns, AAP sponsored a seminar in 1956 on a broad range of problems specific to premature and newborn infants. To reduce mortality from infections, some discussants recommended that premature newborns born after premature rupture of membranes (24 to 48 hours prior to deliv- ery) be treated prophylactically with antibiotics, including chloramphenicol (Day and Silverman, 1957), even though no controlled studies had investi- gated the drug’s safety and efficacy for use with neonates. In 1959, a report of three newborns who died without explanation dur- ing treatment with chloramphenicol (Sutherland, 1959) was soon followed by the report of a randomized clinical trial to evaluate the effectiveness of prophylactic antibiotics in reducing mortality in premature newborns fol- lowing prolonged premature rupture of membranes (Burns et al., 1959). In the trial, mortality rates for the two groups treated with chloramphenicol were 68 and 60 percent. In contrast, mortality rates for the placebo group and the group treated with different antibiotics (penicillin or streptomycin) were 19 and 18 percent, respectively. Other studies determined that newborns, in particular, premature new- borns, could not eliminate the drug from their bodies as fast as older infants and children (Weiss et al., 1960). As a result, dosing at levels used for older children and adults increased chloramphenicol concentrations to dangerous levels. This led to the “gray syndrome” (or “gray baby syndrome”), which was characterized by abdominal distension beginning 2 to 3 days after the start of chloramphenicol treatment and then by grunting respirations, cardiovascular collapse with gray skin color, and death. Although most off-label use of drugs does not have such dire consequences, the experience with chloramphenicol underscores the potential hazards of using new drugs in children, especially newborns, and the importance of controlled studies to guide decisions about when, how, and whether to use them. DEVELOPMENTAL PHARMACOLOGY AND PHARMACOGENOMICS Basic Aspects of Developmental Pharmacology1 The visible changes that occur as a newborn infant grows into a toddler, child, adolescent, and then a young adult are well known. As knowledge of 1 Resources for this discussion include the work of Kearns et al. (2003), Ward and Lugo (2005), and Rakhmanina and van den Anker (2009). The Food and Drug Administration provided draft guidance on the conduct of pediatric pharmacokinetic studies in 1998 (CDER/ CBER, 1998a).

OCR for page 43
45 CHILDREN’S GROWTH AND DEVELOPMENT the biology underlying this normal growth and development has increased, so has the recognition that these changes significantly affect the responses of growing children to medications. Such changes require evidence-based methods for selecting safe and effective doses of medications for children at different stages of development and for engineering appropriate delivery systems for these medications. Adjustments in dosing are often more com- plicated than simply scaling down the dose determined for adults on the basis of a child’s age or weight. The study of what happens to a drug in the body is a key focus of the field of clinical pharmacology. Developmental pharmacology studies the changes that take place in the clinical pharmacology of drugs as a child grows from birth to adolescence. Once administered, drugs undergo biochemical changes that allow their absorption, distribution, metabolism, and removal from the body (processes collectively referred to as the pharmacokinetics of a drug). These biochemical changes—which may occur in the intestinal tract, liver, or other organs through the action of drug-metabolizing enzymes—may facilitate absorption or elimination. Some of these enzymes are not fully active at the time of birth, especially premature birth. An important group of enzymes involved in drug metabolism includes cytochrome P450 (CYP), which is primarily present in the liver. One specific CYP can often metabolize several drugs that belong to the same drug class and carry out similar actions in the body. Conversely, a specific drug may also be metabolized by several different CYPs. After a drug is absorbed into the bloodstream, it can quickly move throughout the body. For drugs taken orally, absorption from the gas- trointestinal tract occurs more rapidly for drugs that are small molecules (those with a molecular mass of less than 500 daltons), not ionized, and fat soluble. Ionization—and therefore absorption—of drugs varies with the pH in the gastrointestinal tract, which ranges from very acidic in the stomach to more alkaline in the small intestine. Absorption differs between premature and term infants, and stage of development may also affect absorption for other modes of administration (e.g., through the skin). After a drug is moved or distributed throughout the body, its concentra- tion in the blood generally decreases. The extent to which a drug is distrib- uted throughout the body depends on a number of factors, including how readily it dissolves in water. For drugs that are water soluble, this lowering of the concentration by dilution in body water is particularly important in premature newborns, who have proportionately more body water than do adults and older children. Individual dosages of water-soluble drugs for pre- mature newborns must often be increased to adjust for this increased body water so that the drugs reach an effective concentration in the bloodstream. After enzymatic changes, many drugs are eliminated in the urine. Oth- ers continue to undergo further biochemical changes that allow the drug or

OCR for page 43
46 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN metabolite to be excreted into the bile. The steps to change a drug molecule into a form that is more readily eliminated by the body often require the ac- tion of a number of enzymes. In developing children, the individual enzymes for drug metabolism and conjugation usually do not mature at the same rate, nor does the maturation of an individual enzyme occur at a constant rate. For newborns and young children, the dose of a drug is often adjusted to the child’s body weight or body surface area to take into account not only size but also the maturation of enzymes that occurs with growth. Studies have demonstrated, however, that neither body weight nor body surface area fits the maturation process exactly. At some stages dur- ing growth, especially from a few months to several years of age, the rate of increase in liver activity for some CYPs exceeds the rate of growth, so the dose of a drug per unit of body weight must be as high as twice that in an adult to keep the concentration in a therapeutic range. In contrast, for premature newborns, many CYP enzymes are underdeveloped, and the drug doses must be given at intervals much longer than those used for older children or adults. Without knowledge of the rates of drug removal from the body, dosing in the wrong amount and at the wrong interval can cause drugs to accumu- late in newborns and infants, sometimes to toxic or even lethal concentra- tions. The only way to determine the correct dose of medications is to test them in children at different stages of development. Otherwise, children can be harmed. Although a dose that is too high may be toxic, a dose that is too low may be ineffective. Premature newborns are a special challenge in determination of the ap- propriate dosages of medications because of their unique physiology as well as the difficulty of studying drugs in this fragile population. In the neonate, the liver’s capacity for drug metabolism is immature for many but not all drugs, and the kidney is similarly immature in filtering drugs selectively into the urine. Given that neonates born as early as 24 weeks (or 6 months) prematurely now commonly survive, the challenge for developmental phar- macology has increased. For some drugs (e.g., aminoglycoside antibiotics), changes in the rate of clearance or elimination of drugs from the body may correlate with both gestational age (the number of weeks since the mother’s last menstrual period) and chronologic age (age after birth).2 For other drugs (e.g., pantoprazole), clearance may correlate more closely with chrono- 2 Postmenstrual age may also be used to describe the age of a preterm infant. It is the in- fant’s gestational age at birth plus his or her chronological age (AAP Committee on the Fetus and Newborn, 2004). For preterm infants, chronological age differs from corrected age. The latter, which is used for preterm infants below the age of 3, is determined by subtracting the number of weeks that an infant was born before 40 weeks of gestation from his or her chronological age.

OCR for page 43
47 CHILDREN’S GROWTH AND DEVELOPMENT logic than gestational age (Ward et al., 2010). As a general rule, how a drug is removed from the body needs to be studied both in preterm newborns, that is, infants born at less than 34 weeks of gestation, and in newborns born from 34 weeks of gestation to term. Separate studies may be needed for the most immature newborns (those born at 24 to 28 weeks of gestation). Different diseases may also influence renal and liver clearance of drugs in children in ways that require dosage adjustments. For example, infants with intestinal problems who are unable to eat and must be fed intrave- nously often develop cholestasis (impaired bile flow). This condition re- duces bile acids in the small intestine, which in turn reduces the absorption of fat-soluble drugs and the excretion of conjugated drugs into the bile and requires adjustments to some drug doses. In contrast, drugs such as phenobarbital and rifampin increase the activity of many drug-metabolizing enzymes in the liver. Again, the only way to determine the appropriate ad- justments is by study with relevant pediatric populations. As children move from infancy through childhood and adolescence, their developmental maturity—as it affects responses to drugs—more closely approaches that of adults (Carr and Ensom, 2003). Adolescent development is, however, highly variable. The onset of puberty in children who are living in similar environments and have no medical conditions that could accelerate or delay puberty may vary by as much as 4 to 5 years (Parent et al., 2003). For that reason, some studies of drugs in older children and adolescents use a measure of pubertal development (Tanner staging) rather than age to specify the upper or lower developmental bound- ary for enrollment in a trial. Behavior can also be an issue, for example, when uncertainties about adolescent compliance with self-administered dosing regimens complicate interpretation of clinical response or study measurements. A National Institutes of Health (NIH) working group on adolescent therapeutics has recommended more research on a number of topics, in- cluding how pubertal development and body weight affect drug distribution and metabolism (NICHD, 2010). The group noted, for example, the need for studies to understand risk factors and other aspects of weight gain in adolescents using antipsychotic and certain other medications. As cited in Chapter 5, the Food and Drug Administration’s (FDA’s) Pediatric Advisory Committee has recommended that information about the possible risk of pediatric weight gain be added to the labeling of these drugs. In addition, some have argued that dosing strategies for studies of drugs for major de- pression in children, particularly adolescents, have not consistently taken into account the results of pharmacokinetic studies (Findling et al., 2006). The concern about weight may apply to medications prescribed for younger children as well as adolescents.

OCR for page 43
48 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN Pharmacogenomics and Developmental Pharmacology One area of challenge and opportunity for pediatric drug studies requested under the Best Pharmaceuticals for Children Act (BPCA) or required under the Pediatric Research Equity Act (PREA) involves phar- macogenomics (see, e.g., Cohen and Ness, 2009; Hudson, 2011; Neville et al., 2011). Pharmacogenomics is the study of how individual genetic variability affects the body’s response to medications (SACGHS, 2008). As of October 2011, FDA had identified almost 100 drugs with labeling that included pharmacogenomic information (FDA, 2011c). The inclusion of pharmacogenomic information in labeling is most common for oncology and psychiatry drugs. To cite an example in psychiatry, the labeling for aripiprazole (Abilify) advises dosing adjustments for patients identified by cytochrome CYP2D6 genotype as poor metabolizers (BMS, 2011). In some cases, the inclusion of pharmacogenomic information in label- ing takes the form of a boxed (“black box”) warning. For example, a boxed warning on the label of the drug abacavir sulfate (Ziagen) states that hyper- sensitivity reactions to the drug can be fatal and that “patients who carry the HLA-B*5701 allele are at high risk for experiencing a hypersensitivity reaction” (GSK, 2010, p. 1). This drug is approved for treatment of HIV infection in patients 3 months of age or older, and testing for this allele is now an accepted element of the standard of care for HIV-infected children (Panel on Antiretroviral Therapy and Medical Management of HIV-Infected Children, 2011). Advances in pharmacogenomics may affect other drug therapies for children. To cite examples, two common childhood conditions—attention deficit hyperactivity disorder (ADHD) and asthma—have known genetic components that affect responses to certain drugs. In children with ADHD, the response to methylphenidate (which is found in drugs such as Ritalin and Concerta) is affected by polymorphisms in the dopamine transporter gene (DAT1) (see, e.g., Gruber et al., 2009). In the treatment of asthma, bronchodilation or the worsening of asthma in patients on continuous short-acting and long-acting beta-agonists is associated with polymor- phisms in the β2-adrenergic receptor gene (ADRB2) (see, e.g., Lima et al., 2009). In patients using inhaled corticosteroids, other genetic variations contribute to variability in airway responsiveness, lung function response, and clinical exacerbations. As in other areas, the developmental variability described in the first part of this chapter adds complexity and may limit the generalization to children of findings from pharmacogenomic studies with adults. For example, researchers recently reported that a pharmacogenetics- based dosing algorithm for warfarin that was derived from adult data consistently overestimated the pediatric dose of the drug (Biss et al., 2012). In addition to affecting treatment decisions, pharmacogenomics can aid

OCR for page 43
49 CHILDREN’S GROWTH AND DEVELOPMENT the design of pediatric drug trials and other studies. Genotypic data can be included as a covariate in population-based analyses of pharmacokinetic or pharmacodynamic data, in which the contribution of the genotype to outcome can be examined (Neville et al., 2011). In addition, genotypic information can be useful in identifying the reason for outlier pharmacoki- netic or pharmacodynamic data in a given cohort of research participants, which may in turn allow a fuller understanding of variability in drug ac- tion. Incorporation of pharmacogenomics in clinical trial designs to better identify patient characteristics associated with differences in drug response could reduce the number of pediatric trials that fail to show efficacy because of a lack of sufficient information on such characteristics. Incorporation of pharmacogenomics could likewise allow reductions in sample sizes, which is a particular issue in pediatric studies. These and other applications of pharmacogenomics have ethical im- plications that are beyond the scope of this brief discussion (see, e.g., Issa, 2002; Freund and Clayton, 2003; Moran et al., 2011). Nevertheless, con- sideration of these implications is relevant for both pediatric research and pediatric medicine. TAILORING PEDIATRIC RESEARCH TO DEVELOPMENTAL VARIABILITY Paraphrasing a common theme in pediatrics, children are not just small research participants. At different ages from birth through ado- lescence, children who participate in research differ from adult research participants—and from each other. An understanding of developmental pharmacology and the appropriate conduct of pharmacokinetic and pharmacodynamic trials is an essential ele- ment for most pediatric drug research plans. Those designing, conducting, and assessing the data from pediatric drug studies must also deal with other challenges related to developmental variability. This section outlines some of these challenges and responses to them. Later chapters provide further discussion of selected issues, including ethical considerations and the use of extrapolation. Appropriate Drug Formulations and Drug Delivery Systems In planning clinical evaluations of the safety and efficacy of medications in children, one early question is whether the formulation of a medicine de- veloped for adults will be suitable for children in the age groups to be stud- ied. If not, one element of the research program will be the development of an age-appropriate formulation or formulations. A few examples illustrate the ways in which adult formulations may be unsuitable for children.

OCR for page 43
50 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN • Children may be more resistant than adults to taking unpleasant- tasting medicines. • Younger children may be unable to swallow adult capsule or tablet forms. They may require a liquid formulation that is practical, safe, effective, stable, and also palatable. Other options include a chew- able tablet, a dissolvable powder, or a product that can achieve reliable doses when sprinkled on applesauce or a similar food. • The appropriate amount of medication in a tablet may vary for children of different ages. A tablet with a single strength may be sufficient for adults, but tablets with different strengths may be needed for children. • Intravenous drugs may be too concentrated for small infants (i.e., the appropriate volume for these patients is too small to measure reliably). In addition, preservatives, binders, and other additives that are safe for adults may not be safe in all pediatric age groups, particularly neonates and infants. The past problems with benzyl alcohol cited in Chapter 1 are a case in point. Today, unresolved issues include the safety of commonly used additives such as propylene glycol and ethanol (see, e.g., Nahata, 2009). In the absence of appropriate pediatric formulations and pediatric labeling of medications, pharmacists may create an extemporaneous formu- lation that differs from the formulation provided and studied by the drug makers. Such formulations present their own problems related to stability, sterility, palatability, additive safety, and limited evidence-based guidance (see, e.g., Nahata and Allen, 2008). An example of the research use of an extemporaneous formulation is described in the clinical review for sotalol (Betapace), which was studied in response to a request under BPCA with exclusivity granted in 2000. The FDA clinical reviewer described the compounding as follows: Five intact Betapace tablets (120 mg = 600 mg) were added to 120 ml of commercially obtained simple syrup (contained [sic] 0.1% sodium benzo- ate) in a six ounce amber bottle. The bottle was shaken and the tablets allowed to hydrate for >2 hours (or overnight). The tablets are shaken intermittently until the tablets disintegrated. The formulating was com- pleted when . . . the syrup contained a fine dispersion of particles. The final concentration of the formulation was 5 mg/ml. (Karkowsky, 2000, p. 6) Because FDA did not approve this product for pediatric use, the develop- ment of a commercial formulation did not arise. Nonetheless, the current labeling includes guidance for dosing in children, and it presents instruc- tions for compounding an extemporaneous oral formulation that are more informative than those just described (Bayer Healthcare, 2010).

OCR for page 43
51 CHILDREN’S GROWTH AND DEVELOPMENT In addition to developing different formulations of a drug, sponsors may need to modify products that combine a drug and a device because combination products or delivery instruments developed for adults may not be suitable for delivering medications to children. To cite one example, measuring devices such as calibrated spoons or droppers that are suitable for use with liquid formulations for adults may not provide sufficient precision for the small doses required for children. (A different concern is that some parents may not understand that household tableware is not standardized by volume and that medications must be measured with spe- cific devices to provide an accurate dose.) Measuring devices may also be marked in ways that do not assist with accurate dosing for either adults or children. FDA issued guidance on dosage delivery devices for liquid over- the-counter mediations in 2011 (CDER, 2011b). To cite another example of drug delivery issues, children may not be able to manipulate safely and effectively the inhalation devices used to deliver certain asthma or other respiratory tract medications to adults. For younger children who cannot reliably match inhalations to medication release from a handheld metered dose inhaler, companies have developed spacers or chambers that can hold the released medication so that coordi- nated breathing is not required. Each new drug delivery modality requires extensive documentation from clinical trials to show that the drug is delivered as anticipated or reaches effective concentrations in children. In 2011, NIH announced fund- ing opportunities for investigators to explore new strategies for the cre- ation and testing of drug formulations suitable for children (NIH, 2011a). It noted a number of questions specific to the task of creating palatable formulations for children, as well as questions related to advances in drug delivery alternatives (e.g., skin patches and dissolvable oral films similar to over-the-counter breath freshener strips) and different approaches to oral delivery of medications (e.g., nanotechnologies). In developing a written request or requirement for pediatric studies under BPCA or PREA, FDA may consider the need for a new pediatric for- mulation. For example, the final version of the written request for a study of terbinafine hydrochloride (Lamisil) for the treatment of tinea capitis (ringworm) specified that the sponsor use an appropriate formulation (e.g., suspension or rapid-dissolution tablets). Further, it specified the following conditions: If the studies you conduct in response to this Written Request demonstrate this drug will benefit children, then an age-appropriate dosage form must be made available for children. This requirement can be fulfilled by de- veloping and testing a new dosage form for which you will seek approval for commercial marketing. If you demonstrate that reasonable attempts to develop a commercially marketable formulation have failed, you must

OCR for page 43
52 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN develop and test an age-appropriate formulation that can be compounded by a licensed pharmacist, in a licensed pharmacy, from commercially avail- able ingredients. (Beitz, 2006b) As discussed in Chapter 3, the FDA Amendments Act of 2007, which reauthorized both PREA and BPCA, explicitly provides for a waiver of required pediatric studies if the sponsor can demonstrate why a pediatric formulation is not possible; the grounds for the waiver must be made public, however. Furthermore, FDA must report annually on the number of pediatric formulations developed, the number of such formulations not developed, and the reasons for a failure to develop a formulation. As of December 31, 2011, FDA reported the development of five pediatric for- mulations under BPCA and PREA (most related to studies required under PREA); the agency reported no formulations that were not developed.3 The legislation also requires FDA to publish a notice that identifies any drug formulation that was developed, tested, and found to be safe and effective for pediatric use but that was not marketed within a year following a deter- mination about pediatric exclusivity. Since the enactment of this provision, FDA has posted two such notices: one for a formulation of pantoprazole sodium oral suspension for delayed release and the other for valganciclovir (formulation not specified). Appropriate Research Endpoints and Procedures Developmental differences may entail not only the creation of different formulations of medications for use with children but also the creation of developmentally appropriate research measures and procedures that differ from those used in studies with adults. As discussed further in Chapter 5, efficacy endpoints in pediatric clinical trials may differ from the endpoints in studies with adults and may also vary across pediatric age groups. Alternative and Surrogate Endpoints Efficacy measures used for adults or older pediatric age groups are sometimes not suitable for use with younger age groups. For example, to study medications that are intended for the relief of symptoms such as pain or nausea, symptom scales designed and validated for use with pediatric age groups may be necessary, including different scales for early verbal children, somewhat older children, and children with intellectual or developmental disabilities (Tomlinson et al., 2010). For preverbal children, symptom mea- sures may be based on parent or investigator assessment of facial expres- 3 This information is posted and updated at http://www.fda.gov/downloads/ScienceResearch /SpecialTopics/PediatricTherapeuticsResearch/UCM194987.pdf.

OCR for page 43
53 CHILDREN’S GROWTH AND DEVELOPMENT sions and physical movements (see, e.g., Taddio et al., 2009). Both kinds of measures of symptoms are alternatives to those used for adults. An alternative endpoint may also be a surrogate endpoint. A surrogate endpoint in a clinical trial is a laboratory measurement or a physical sign used as a substitute for an endpoint that measures directly how a patient functions, feels, or survives. For adults as well as children, surrogate end- points may be used in a variety of clinical research situations in lieu of endpoints such as mortality or organ failure that may occur rarely or that may develop over a period of years. Examples that have been validated for some research uses include blood pressure, exercise capacity, and cho- lesterol levels. FDA has recognized in various contexts the value of sur- rogate measures in pediatric trials. For example, in 2000 draft guidance on pediatric oncology studies, the agency emphasized that approval of a drug for pediatric use could be based on a drug’s effect on tumor size or other surrogate measure that was likely to predict clinical benefit (CDER/ CBER, 2000). A particular surrogate measure may not be appropriate for children of all ages. For example, forced expiratory volume in 1 second (FEV1) is an accepted surrogate measure to assess the advance of lung dysfunction in patients with diseases such as cystic fibrosis. Although widely used in older children, it requires physical maneuvers (i.e., strongly inhaling and force- fully and completely exhaling) that can be difficult for young children and impossible for infants to perform (Castile, 2004). Training and experience may make measurement of FEV1 feasible with children as young as 5 years of age, but alternative measures and techniques are usually required for use with children less than 6 years of age. To cite another example, exercise capacity is often used as a surrogate measure in children with pulmonary hypertension or congestive heart failure, but its reliable measurement in children less than 7 years of age, who are often developmentally unable to perform the test, is difficult. This difficulty is further compounded in chil- dren with developmental delay, such as those with Down syndrome, who are predisposed to pulmonary hypertension and congestive heart failure (Walker, 2010a). As in studies involving adults, investigators may devise composite end- points for pediatric trials. Each single endpoint that is included in a com- posite endpoint should have clinical significance and interpretability in its own right. The composite endpoint then becomes a summary measure of effect from the different variables. The rationale for using a composite endpoint in a clinical trial is that it can reduce the size of the trial if the components of the composite increase the number of events. This can be a major advantage in pediatric trials. In addition, a composite endpoint can address broader aspects of a multifaceted disease and can combine compo- nents (e.g., rehospitalization) that occur more frequently than other com-

OCR for page 43
54 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN ponents (e.g., mortality). In general, these components should add to the total treatment effects, move in the same direction, be of generally similar significance, and be easily ascertained.4 Some studies of drugs to treat HIV infection offer an example of the use of a composite endpoint that reflects developmental considerations. Because infection with HIV can negatively affect children’s growth, growth has been incorporated into composite endpoint measures for some pediatric studies of antiretroviral drugs. Although changes in weight were the initial focus, studies have suggested that changes in height are more closely related to survival (Benjamin et al., 2004). Use of Alternative Biospecimen Sampling Procedures Alternative research procedures may also be necessary for studies that require frequent sampling and testing of blood and other biological speci- mens. This sampling can be stressful for adults, who typically understand the rationale and the procedure; it can be even more stressful for children, particularly young children. For these children, their small veins also com- plicate the drawing of blood, and they have a smaller volume of blood, which limits the amount of blood that can be safely drawn.5 Fortunately, technological advances allow accurate assays with smaller sample sizes than in the past. In addition to assay innovations, the greater use of population-based pharmacokinetics permits less frequent or dense individual sampling than in traditional pharmacokinetic studies (CDER/CBER, 1999a; see also Zuppa et al., 2011). This can, for example, reduce the burden of frequent blood draws on individual children. Population pharmacokinetics can be de- scribed as “the study of variability in drug concentrations between individu- als . . . [including] the assessment of variability within the population and . . . [the assessment of possible sources of] variability in terms of patient characteristics such as age, renal function or disease state” (EMA, 2009a, p. 3). The approach also allows the use of data from a variety of sources not normally used in pharmacokinetic analyses, for example, data from studies assessing the relationships between dose and efficacy or safety. 4 In guidance on the discussion of clinical studies in a drug’s labeling, FDA has advised that “[i]n general, the results for all components of a composite endpoint should be presented. Presentation of all components reveals which components are driving the result and which components may be unaffected, or even adversely affected, by treatment with the drug” (CDER/CBER, 2006, p. 5). 5 The institutional review boards that review research proposals for compliance with stan- dards for human research protections (see Chapter 4) may have guidelines on acceptable blood draw volumes by weight (see, e.g., http://www.ucdmc.ucdavis.edu/clinicaltrials/documents/ Blood_Draws_Maximum_Allowable.doc).

OCR for page 43
55 CHILDREN’S GROWTH AND DEVELOPMENT Aside from these kinds of procedural or methodological innovations, investigators studying hospitalized children may be able to obtain extra serum and plasma during clinically indicated blood sampling to allow repeat validation of an analysis without additional blood draws. Such “scavenged” samples can be used to enhance pharmacokinetic studies, especially in small premature newborns (Wade et al., 2008). In addition, pharmacokinetic studies of some drugs may be amenable to the use of samples of other bodily fluids (such as tears or urine) that can be obtained noninvasively (McCracken et al., 1980). Children’s Development and Adaptations in Research Strategies Development-related differences such as those described above may require a variety of adaptations or additions to pediatric research plans or strategies. As discussed in Chapter 4, ethical considerations may also dictate adaptations. Studies with Juvenile Animals Concerns about possible toxicities not seen in adults may prompt FDA to require short-term or long-term studies involving juvenile animals. Such studies generally supplement the studies with older animals that typically precede clinical trials with adults. For example, when FDA approved abatacept (Orencia) for treatment of juvenile idiopathic arthritis in patients ages 6 to 16 years, it deferred clinical studies for the 2- to 5-year-old age group until data from three safety stud- ies with juvenile rats had been submitted and evaluated (Rappaport, 2008). FDA’s online database for tracking postmarket study requirements shows that the data from rat studies have been submitted. (It also shows—without explanation—that FDA released the sponsor from the requirement for the deferred clinical studies with children in the 2- to 5-year-old age group.) Studies with Different Pediatric Populations As explained above, developmental differences within the pediatric population often require that separate clinical studies be undertaken with individuals in different age groups. For a number of the products discussed in this report, FDA required studies with neonates; infants up to 1 or 2 years of age; one or two groups of older, preadolescent children; and ado- lescents. Separate studies with each age group, however, may necessitate adjustments in the research plan, for example, if suitable efficacy measures are not available for the youngest age groups. Aside from the additional complexity and cost of separate studies, one

OCR for page 43
56 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN disadvantage of separate studies for different age groups is that the separate studies may fragment what is already a small population. Although such fragmentation presents problems, one alternative—inclusion of patients covering a broader age range in a single study that is not powered for sub- group analysis by age—presents the risk that the study will fail to enroll sufficient numbers of patients in relevant age groups to identify important developmental differences in a drug’s safety and efficacy. Use of Extrapolation Chapter 5 discusses one strategy that FDA commonly allows in an effort to encourage pediatric drug studies while reducing the costs to spon- sors. Instead of specifying the two adequate, well-controlled safety and ef- ficacy trials that are often required for studies of drugs in adults, FDA may indicate in advance that it will accept the use of extrapolation of efficacy from studies with adults to children (or from one pediatric age group to another), usually with requirements for the submission of some supportive pharmacokinetic, safety, and efficacy data. For a particular drug and indication, the appropriate use of extrapola- tion depends on a careful assessment of similarities and differences between adults and children in the course of the disease and the effects of the drug. FDA may thus accept extrapolation for some age groups (e.g., adolescents) but not others (e.g., neonates). Different Approaches to Pharmacokinetic Studies For adults, Phase I studies often start with a small number of healthy volunteers. The studies seek to investigate a drug’s pharmacokinetics in individuals not affected by a disease under study; they, therefore, carry no prospect of medical benefit to these volunteers. For pediatric drug studies, either the drug or the research procedures (e.g., extensive blood draws), or both, are often deemed to involve more than minimal risk without the prospect of direct benefit to the child. Such studies are restricted under the framework of the research protections described in Chapter 4. As a result, with FDA and institutional review board agreement, spon- sors of pediatric drug studies typically develop needed pharmacokinetic evidence by using a combination of data from previous studies with adults and new data from studies involving children who have the condition be- ing studied. For example, the clinical pharmacology review for the drug sotalol (Betapace) included a literature review of data from studies of healthy adults, ill adults, and ill children. It also evaluated the findings from two Phase I trials (Gobburu and Canal, 2000). One of these trials was a

OCR for page 43
57 CHILDREN’S GROWTH AND DEVELOPMENT single-dose study involving 34 children (ranging from neonates to children 12 years of age) who needed treatment for arrhythmias. The other was a study of 25 children (in the same age range) using an ascending-dose titra- tion design with three dose levels. A pediatric pharmacokinetic analysis is sometimes embedded in a safety and efficacy study. For example, for the investigation of zoledronic acid (Zometa) for osteogenesis imperfecta, the pharmacokinetic study was part of the clinical safety and efficacy study (as allowed by the written request) (Vaidyanathan, 2008). One ethical rationale for this approach is that the study would have the prospect of benefit. As described earlier in this chapter, the methods of population phar- macokinetics can minimize the burden on child research participants, for example, by collecting fewer samples per participant from a larger study population (CDER/CBER, 1998a, 1999b; Howie, 2010). This approach has ethical as well as practical and economic advantages in certain situations. As discussed in Chapter 7, the lack of pediatric pharmacokinetic studies may contribute to unsuccessful efficacy trials. For example, FDA requested safety and efficacy studies but not a pharmacokinetic study for the use of albuterol sulfate inhalation (Ventolin HFA) aerosol to treat asthma in chil- dren ages birth up to 2 years and 2 years up to 4 years. The clinical reviewer concluded that the studies did not show efficacy and that the dose chosen for the studies might not have been optimal (Wang, 2008). Other Modifications in Trial Design Among other advances in strategies for designing clinical studies, adap- tive trial designs are potentially helpful in pediatric drug studies. These strategies allow certain changes in trial design based on planned analyses of data collected at interim points during a trial. As described in FDA guid- ance, such changes may make studies “more efficient (e.g., shorter duration, fewer patients), more likely to demonstrate an effect of the drug if one ex- ists, or more informative (e.g., by providing broader dose-response infor- mation)” (CDER/CBER, 2010a, pp. 1–2). For example, as dose-response data accumulate during the course of a trial, analyses may indicate a lack of response or unanticipated adverse reactions for a particular dose; further use of that dose can then be stopped. To cite another example, an interim analysis may suggest the need to adjust the sample size upward or down- ward, thus avoiding either an unnecessarily large sample or a statistically underpowered study that will not provide adequate evidence about a drug’s efficacy. The FDA guidance stresses the importance of careful application of these techniques to avoid the introduction of bias that compromises the validity of study results.

OCR for page 43
58 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN One example of an adaptive design in pediatrics is seen with clopido- grel (Plavix), which was investigated under BPCA for treatment of neonates and infants with cyanotic congenital heart disease palliated with a systemic artery-to-pulmonary artery shunt. The event-driven trial design included three interim analyses conducted by an independent statistician associated with the data-monitoring committee for the study. The design would have allowed the early discontinuation of the trial if the interim analyses showed a definite efficacy advantage (or a safety concern) for the test drug (Chen, 2010). As it turned out, neither the interim nor the final analyses supported efficacy. FDA also cited problems with the sponsor’s approach to certain aspects of the research that might have compromised the potential of the study to demonstrate efficacy. Attempts have been made to devise trial architecture that is more ac- ceptable to children and their families and that will thereby encourage enrollment of children, which is a persistent challenge for research spon- sors. Parents are particularly averse to enrolling their children into clinical trials in which the children may be exposed to long courses of a placebo (Caldwell et al., 2003). One example of alternative trial architecture is the randomized with- drawal design. It has been used for a number of trials of biologic therapies for juvenile arthritis (Lovell et al., 2000, 2008; Ruperto et al., 2008). In this design, all subjects are enrolled into an open-label phase in which all subjects receive study medication. Only those participants who show a response go on to further study (which makes this an example of an en- richment design). Those responding are then randomized to continue with active therapy or to be switched in a blinded fashion to placebo (i.e., with- drawn from active therapy). The main study endpoint is the proportion of participants in the two arms who maintain a response (or, conversely, the proportion who have a disease flare). This study architecture is favored by some parents and investigators since the children randomized to placebo may be switched back to active therapy (in an open-label fashion) as soon as a disease flare occurs; in this way, prolonged exposure to placebo is minimized. Other study architectures that aim to maximize enrollment and mini- mize exposure to placebo include randomized dose comparison designs, the randomized placebo phase design (Feldman et al., 2001; Abrahamyan, 2011), and crossover and multiple-crossover designs. Infrastructure for Research in Pediatric Therapeutics The kinds of challenges outlined above have prompted efforts to create and maintain research resources to support drug studies that appropriately accommodate developmental variability. These resources include

OCR for page 43
59 CHILDREN’S GROWTH AND DEVELOPMENT • clinical investigators knowledgeable about developmental pharma- cology and other features of pediatric research; • physical facilities that accommodate children of different ages and their parents; • trial design and data analysis strategies tailored to pediatric trials; • suitable administrative structures, including systems that support the multisite networks often required for pediatric studies to enroll sufficient numbers of children; and • child-focused research ethics programs that include individuals with extensive experience in conducting or evaluating clinical re- search involving children. Although the actions are limited in scope, considering the need, NIH has taken some steps to develop a better infrastructure for pediatric clini- cal trials. In 1994, the National Institute of Child Health and Human Development (NICHD) established the first national network for pediatric pharmacology (NICHD, 1998). Later, it supported the creation of the Pediatric Pharmacology and Therapeutics Research Consortium. The an- nouncement of funding opportunities for the latter noted the need “to ad- dress knowledge gaps that may be responsible for failed [pediatric] efficacy trials” (NIH, 2008, unpaged). In 2010, NICHD announced a contract for Duke University to create the Pediatric Trials Network to develop a stron- ger infrastructure for clinical trials in support of the institute’s BPCA pro- gram, which focuses on high-priority studies of off-patent drugs (Berezny et al., 2011). (See Chapters 3 and 6 for a description of NICHD’s role in BPCA and in setting priorities for pediatric therapeutic research, including neonatal research.) Within the Clinical and Translational Science Awards program (which aims to speed the pace at which laboratory discoveries lead to effective treatments), a working group has focused on ways to accelerate progress in pediatric research. For the 2011 meeting of the Pediatric Academic Societ- ies, the group helped organize a session on the BPCA. The session featured presentations of strategies for developing better predictors of outcomes in pediatric drug studies (CTSA CCHOC, 2011). Disease-focused initiatives also play a role in supporting drug stud- ies for pediatric health conditions. For example, the Children’s Oncology Group (COG), created in 2000 through the merger of four smaller groups, is an international cooperative that each year conducts dozens of clinical trials with NIH and industry funding. Because cancer care for children is more concentrated in research institutions than is adult care, approxi- mately 90 percent of children with cancer in the United States are treated in COG institutions. The group’s cooperative research strategy has achieved relatively high rates of enrollment in trials of cancer therapies (50 to 60

OCR for page 43
60 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN percent of all eligible children and 90 percent of children under age 5 years) (O’Leary et al., 2008). Even so, achieving sufficient enrollment is often a challenge. The group places a priority on the early assessment of a drug’s potential and the timely ending of unpromising trials so that limited re- sources—including research participants—can be most effectively allocated. The Cystic Fibrosis Therapeutic Development Network, which is affili- ated with the Cystic Fibrosis Foundation (CFF), has been an innovator in advocacy group efforts to stimulate focused drug discovery, translational, and clinical research. The network is a subset of specialized research centers drawn from a larger network of clinical care centers; it has expanded from 18 to 80 centers in recent years (CFF, undated). Although not specific to pediatric studies, FDA’s initiatives to advance regulatory science have the potential to improve such studies. As defined by FDA, regulatory science is “the science of developing new tools, standards, and approaches to assess the safety, efficacy, quality, and performance of FDA-regulated products” (FDA, 2011a). As part of the initiative for de- veloping and refining clinical trial designs, endpoints and biomarkers, and analytic tools, the agency described needs to • continue to refine clinical trial design and statistical methods of analysis to address issues such as missing data, multiple endpoints, patient enrichment, and adaptive designs; • identify and evaluate improved clinical endpoints and related bio- markers for trials in areas where optimal endpoints are lacking (e.g., efficacy and safety endpoints for osteoarthritis in humans and animals, for gene therapy, for ophthalmic indications, for tumor vaccines, and for stem cell-derived therapies); • develop novel trial designs and endpoints for special needs (e.g., small trials for orphan indications, designs and endpoints for pedi- atric trials including neonatal trials); • continue to refine the use of modeling and simulation in clinical trial design to enhance the effectiveness of clinical studies; [and] • continue development and refinement of tools and approaches for assessing benefit/risk (FDA, 2011a, pp. 11–12). In some instances, as in the third bullet above, FDA explicitly notes the relevance of initiative elements to pediatric studies. To the extent that those involved in implementing the initiative for clinical trials consider developmental issues and solicit pediatric expertise, it should in the future yield improvements in the value of pediatric studies requested under BPCA or required under PREA.

OCR for page 43
61 CHILDREN’S GROWTH AND DEVELOPMENT SHORT-TERM STUDIES AND LONG-TERM CONCERNS Most studies used to support the approval of drugs by FDA are rela- tively short term, lasting for a few days, weeks, or months, even for drugs that are used for years in the treatment of chronic conditions such as asthma, diabetes, and autism. The scarcity of long-term studies of medica- tion effects is a concern for both adult and pediatric populations. For children, however, an added concern is how drugs used either acutely or chronically may affect growth and development or have late ad- verse effects. Even relatively short-term use may be associated with adverse effects years later. One reasonably well-understood example involves drugs that help save the lives of young children with cancer but create risks for later problems, including cognitive limitations, fertility impairment, or new cancers (NCI, 2011). Even when FDA identifies long-term growth and development or other safety issues, it may not include long-term studies in a written request or re- quire longer-term postmarket studies after approving use of a drug by chil- dren. For example, in requesting studies of the use of aripiprazole (Abilify) for treatment of schizophrenia in adolescents, FDA noted concerns about the effects of the drug on growth and development and encouraged but did not specify long-term studies (Behrman, 2003). Some time later, when the agency approved the drug for acute treatment of irritability associated with autism, it did require a long-term efficacy and safety study for maintenance treatment for the condition (Laughren, 2009a). The unclear risk-benefit ratio of the long-term use of some chronic medications may raise questions about when such agents should be started, particularly when the events that they are intended to avert would not be expected to occur for many years. Thus, in an editorial discussing statins and children, Stein (2007) suggested that “given the residual uncertainty of the impact on safety, growth, and sexual development in the younger age groups and the fact that clinical events do not appear until the mid to late 20s at the earliest, it would still appear prudent to delay the start of statin and other lipid-lowering drug therapy until the age and sexual development stage outlined by the recent AHA [American Heart Association] consensus statement” (p. 595). FDA must balance the benefits of facilitating the entry to the market of products showing short-term benefit against the risks of long-term harm. At the same time, it must consider the possibility that the incentives of BPCA may not be sufficient to attract positive responses from sponsors when a request involves a long-term study. Chapters 5 and 6 also note the need for long-term studies of drugs. Chapter 5 suggests that FDA could make greater use of its authority to require long-term safety studies when it approves a product for pediatric use.

OCR for page 43
62 SAFE AND EFFECTIVE MEDICINES FOR CHILDREN CONCLUSIONS This chapter has provided an overview of developmental pharmacology as a basis for designing, conducting, and evaluating pediatric drug studies. It has discussed how children’s growth and development may require al- terations in research strategies that are commonly used in conducting drug studies with adults. The exclusivity incentive and other features of BPCA and PREA ex- plicitly recognize and accommodate some distinctive features of pediatric research. Notably, with direction from Congress and on its own initiative, FDA has applied additional expertise in pediatrics and pediatric research to the oversight of pediatric study requests or requirements, discussions with sponsors about acceptable research designs, and appropriate review of submitted pediatric data (see Chapters 3 and 4). By employing sufficient expertise in developmental pharmacology and pediatric clinical research from the early stages of pediatric plan discussion through the review of sub- mitted studies, FDA increases the likelihood that studies will generate useful information to guide and improve clinical care for children of all ages. The next chapter moves from developmental variability and pediatric research to public policy. It builds on the overview provided in Chapter 1 to discuss BPCA and PREA in more detail.