Twelve years ago an estimated 80 percent of medications lacked labeling for use in children because they had only been tested for U.S. Food and Drug Administration approval in adults. A big legislative push in 2002 led to great strides in getting more drugs tested in children so that now about half of medications lack directions. But more work is needed, according to a recent policy statement by the American Academy of Pediatrics.
Despite the 500 changes made in pediatric drug-labeling updates since 2002, doctors still must rely only on clinical experience and existing scientific literature to prescribe thousands of commonly used medications in children “off-label”—without specific guidance from the FDA. The new statement published February 24 addresses off-label use as well as obstacles related to insurance coverage and generic use.
The first hurdle is the term itself. “Off-label doesn’t mean the drug is experimental or doesn’t work or that there’s no evidence for it,” says Kathleen Neville, a pediatrician at Children’s Mercy Hospital in Kansas City, Mo., and the policy statement’s lead author. Very common drugs such as ampicillin and many children’s cancer drugs, for example, are prescribed off-label. “It just means there’s not labeling yet,” usually because studies have not been done specifically with children for FDA approval. A drug will only have labeling for use in kids if the FDA has specifically determined the drug’s safety and effectiveness for particular conditions in children based on the data provided during the drug’s approval process or through post-licensure submissions to the FDA. The label, or package insert, lets doctors know how to prescribe the medication, but labels may also simply alert clinicians that the drug is ineffective or unsafe in children.
The FDA can only initially approve a drug for indications and populations that were tested during the drug’s clinical trials. After the lengthy, expensive approval process, further research or clinical use may reveal the drug can treat other conditions or other groups—such as children—not in the original trials. Antiepileptic drugs, for example, are often prescribed for psychiatric disorders based on evidence of their effectiveness. But these uses, or directions, are only added to the label if the drug is resubmitted to the FDA. The AAP statement’s first recommendation therefore advises doctors to base off-label drug use on “sound scientific evidence, expert medical judgment or published literature whenever possible.”
Barriers to prescribing for children
Yet, it is still “the gold standard to have a pediatric label that says it’s approved from birth to 17 years of age,” says Anne Zajicek, chief of the Obstetric and Pediatric Pharmacology and Therapeutics Branch at the National Institute of Child Health and Human Development. Children are not miniature adults—their physiology has different needs—so it requires more time, money and research to determine the safety, effectiveness and indications of drugs for children. “But it’s been difficult to get that information,” she says.
“Because many of these drugs have been around for so long, there’s not as much concern about safety, but you don’t really know until you look,” Zajicek says. What’s not known can be dangerous. It took 10 years, for example, to learn that the antibiotic chloramphenicol can cause potentially fatal gray baby syndrome—a serious side effect causing vomiting, gray-colored skin, a limp body and possible heart failure­­—in newborns when dosing doesn’t take into account their underdeveloped liver functions. Yet, “doing pediatric trials are far more complicated than doing adult trials,” she says. The trials require testing within subgroups of newborns, infants, school-age children and teenagers, and the obstacles become especially hard for newborns when it’s unclear what outcomes researchers should use. A study last month found that among almost 400 drugs used in newborns, not even half of the 24 recent drug label changes made for newborns were even commonly used. Withholding medications, however, would be unethical: Babies born to heroin addicts, for example, require methadone even though it hasn’t been approved for use in newborns.
When data do exist, it comes from post-licensure studies or clinical use. Even then, some fields have more data available than do others. “Child psychiatry is a field where we have the least amount of evidence-based or practice guidelines,” says Scott Carroll, a child psychiatry professor at the University of New Mexico in Albuquerque. “When I say we fly by the seat of our pants, that’s an understatement.”
The primary challenge to testing psychiatric drugs in children, as with drugs in other specialties, is that dosing cannot simply be scaled down. “A teenager’s metabolism is about double by body weight, so my lithium dosing would blow most adults away, but that’s just to get the drug barely therapeutic for a teenager,” Carroll explains. Yet teens need only the same dosages as adults for antidepressants because the drug acts on the body differently than, say, lithium; meanwhile younger children require only about half an adult dose. Child psychiatrists will often preferentially use older drugs because there’s more clinical data available. “What is approved is so spotty, it’s not even considered first-line treatment sometimes,” Carroll says.
Various factors hinder adequate drug research for children. Concerns about ethics, harm and consent make it difficult to get institutional board approval or recruit participants for drug trials, and most pharmaceutical research is market-driven. “A lot of the diseases in children are less common than in adults, so doing studies in children doesn’t lead to blockbuster medications,” Neville says. “There’s not a lot of financial incentive.”
As recently as the 1970s, she says, many considered it unethical to study new drugs in children, resulting in a knowledge gap the industry has no financial interest in filling. Hence, two of the AAP statement’s recommendations encourage pediatricians to advocate for and participate in better pediatric drug research and a third urges journals to publish this research. Journals are often reluctant to publish “negative findings”—what doesn’t work—thereby hindering knowledge on effectively treating children if a drug that treats adults is worthless in kids.
Legislation paving the way
Addressing the dearth of data eventually got a legislative push. “It would be unthinkable that we as a society would functionally deny children access to therapies because we cannot get a product studied in children,” says Dianne Murphy, director of the FDA’s Office of Pediatric Therapeutics. Part of Murphy’s role there is to execute the Pediatric Research Equity Act (PREA), passed in 2003 to require pharmaceutical companies to study new drugs in children if the medication might be prescribed to minors. Companies can only receive PREA waivers if the medication treats a disease, such as prostate cancer, that doesn’t typically occur in children and teens.
Likewise, the Best Pharmaceuticals for Children Act (BPCA), enacted in 2002, updated a 1997 law that gave pharmaceutical companies six extra months of patent use if the company conducted clinical trials for children for drugs already on the market. But the 1997 law didn’t address the many generics in use, so the BPCA authorized the National Institutes of Health to fund studies on older pediatric drugs and prioritize what to test first, enabling unprofitable drugs to gradually receive pediatric labeling. Even after the BPCA, for example, it took 10 years before hydroxyurea, the only drug known to effectively treat sickle-cell anemia, was updated with pediatric labeling.
According to Zajicek, prioritizing pediatric drug studies involves a balancing act based on data that already exist versus what are still needed, what treatment offers the greatest public health benefits and what conditions are most severe. The NIH has already sponsored 17 clinical trials since the BPCA, with 96 drugs or therapeutics and 46 conditions prioritized.
More roadblocks to pediatric dosing
Yet obstacles remain, especially with insurance practices and generics’ use. “Just because a drug doesn’t have labeling for children doesn’t mean that an insurance company shouldn’t [cover it],” Neville says. This has been enough of a problem that the statement’s final recommendation explicitly addresses it, noting that institutions and payers should not use labeling status as the only criterion for using or covering a medication.
The same recommendation cautions against automatically filling prescriptions with less-expensive alternatives, such as generics. Some states  have laws requiring generics to be substituted for branded drugs unless doctors specifically request the branded drug be filled. Yet according to Adam Powell, a health economist who founded Payer+Provider Syndicate, generics are designed to fall within the range of 80 to 125 percent equivalence to the branded drug—but not to one another. So one generic could fall at the low end and another at the high end, with two generics being much further from one another than from the branded drug. Switching from one generic to another could involve a big change in the medication’s effects, not to mention other inactive ingredients that could differ. “There’s a small target between helping a person and killing a person, and the target shrinks for children since drugs are often dosed by weight,” Powell says. “If you combine the drug that has a very narrow therapeutic index [range] with the difficulty of properly dosing the drug for a child and the great potential differences between two generics, you get what may be a dangerous situation.”
Neville suggests parents ask doctors if a particular prescription should be filled with a brand-name drug or a generic. Yet without more data and better labeling, pediatricians may not always know the answer. “It requires pediatricians to understand every generic versus every trade drug, which is an impossible feat,” Neville says. “If you don’t even have enough data on the trade, then what do you do?”