Giving Newborns Medicine Is A Dangerous Guessing Game. Can We Make It Safer?

By Megan Scudellari-01/04/2017

BOSTON — Two rows of plastic cocoons line the walls of the neonatal intensive care unit, sheltering babies so tiny, their little hands can’t wrap around their parents’ index fingers. Many have been treated with multiple medications in their short lives: antibiotics, anesthetics, narcotics, diuretics.

And most of the drugs coursing through their fragile bodies have not been approved for use in newborns.

That’s not just the case here at the Floating Hospital for Children at Tufts Medical Center. It’s a global problem. Pharma companies are afraid to test drugs on babies because they’re so vulnerable, and because the risk of liability is so high. Parents and doctors say they’re wary of enlisting newborns as “guinea pigs” in clinical trials.

The result: An estimated 90 percent of medications administered to newborns are not approved by the Food and Drug Administration for use in children so young. That means neonates — premature and full-term infants less than 28 days old — are routinely treated with drugs that are not adequately tested for safety, dosing, or effectiveness.

Despite this gaping hole in medical knowledge, infants admitted to a neonatal intensive care unit may receive up to 60 medications in their first month of life, said Dr. Jonathan Davis, chief of newborn medicine here at Tufts Medical Center.

Davis, a soft-spoken, baby-faced man with a smile for every child and parent, is a fierce advocate when it comes to changing all this. He’s conducting several clinical trials with the newborns in his NICU. And he’s helping other doctors around the world do the same.

In 2015, an FDA-funded nonprofit launched two global efforts to spur clinical trials in infants. Davis directs one of those initiatives, the

International Neonatal Consortium, known as INC. In November, INC published its initial contribution to the field — a guide to clinical trials in neonates, including information on trial design and data collection for investigators and study sponsors, and advice for drug regulators.

“We’ve got to do something,” said Davis, who has spent more than 35 years studying medications in infants and has published over 150 papers.

Without drug data for newborns, he said, “we can’t be certain which drugs, in which doses, to use when.”

Davis argues that the current system — doctors making decisions based on little more than anecdotes and intuition — essentially treats each sick newborn as an uncontrolled, unapproved study of one. The baby may or may not do well on the drug; either way, no data is collected and the result does not inform treatment of other infants around the world.

“Children are protected through research,” said Dr. Raphaël Rousseau, director of pediatric oncology drug development at Swiss pharmaceutical company Roche.

It’s technically more challenging to study drugs in neonates than older children, he acknowledged. But still, he said, “there is no real excuse not to do drug development in neonates.”

When mistakes are fatal

Some of the babies in the Tufts NICU are swathed in blue light to treat jaundice; others are draped with colorful knit blankets made by local volunteers. Behind a curtain in alcove number 18, first-time mother Aubrey Baptista nurses 2-week-old Elijah.

Elijah was born Oct. 28 with a heart defect requiring surgery. When doctors discovered his condition, Elijah was immediately put on an IV drip of a drug to help keep his heart valve open. During surgery, he received anesthesia and morphine. Post-surgery, doctors administered nitrous oxide, oxygen, and antibiotics.

At least, those are the drugs Baptista remembers being told about.

What she was not told, at any point, was that the FDA hadn’t reviewed the safety or efficacy of those drugs, in those doses, for babies like Elijah. Doctors simply told her, “This is what we’re doing.”

“There wasn’t much choice,” she told STAT.

Over the last 25 years, the FDA has approved only two drugs that significantly improved neonatal survival: surfactant and nitric oxide for respiratory conditions.

Most neonatologists do not tell parents about the lack of medical evidence for the drugs they use, said neonatologist Dr. Matthew Laughon of the University of North Carolina at Chapel Hill. They just prescribe.

Because there is little alternative.

“When you’re sitting there at the bedside, and you’re looking the parents in the eyes, and the baby is dying or really sick, to just stand there and say, ‘We don’t have anything to do because it hasn’t been proven’ —well, that’s challenging,” said Laughon. “If we never gave drugs because they were off-label, we wouldn’t have any drugs.”

Physicians often make treatment decisions by scaling down from how medications are used in adults. “We take it right out of the vial of an adult drug, dilute it down, and give it to the babies,” said Davis.

In multiple cases, that technique has led to disaster. “There have been some big, big, big mistakes in neonatology through the years when it comes to drugs,” said Laughon.

Those include the sudden deaths of preemies due to too-large doses of the antibiotic chloramphenicol in the 1950s; the fatal poisoning of infants from large amounts of benzyl alcohol, a preservative used to flush catheters, in the 1980s; and deaths from a preservative, propylene glycol, in a multivitamin given orally to premature infants in a dose intended for adults.

Infants are not tiny adults and should not be given drugs as if they were, said Catherine Sherwin, division chief of pediatric clinical pharmacology at the University of Utah School of Medicine, who studies pharmacology in neonates.

Newborns absorb, metabolize, and excrete drugs differently than adults. “Yet we haven’t done the studies to know exactly what those differences are,” Sherwin said. “We just know they’re different.”

Lackluster laws and shoddy trials

There have been two large legislative efforts to encourage pharmaceutical companies to increase the number of pediatric drug studies: the Best Pharmaceuticals for Children Act of 2002 and the Pediatric Research Equity Act of 2003. Both were made into permanent law in 2012 with the passage of the Food and Drug Administration Safety and Innovation Act.

Overall, the incentives and requirements within that legislation — carrots and sticks — worked. As of Oct. 31, 

651 drugs in the US have new or revised labeling for pediatric patients.

But the laws did not work for newborns.

Just 24 of the 406 labeling changes made as a result of that legislation affected neonates — and those were primarily on drugs that are used rarely, if at all, in NICUs in the United States, according to a 2015 analysis.

Even worse, the few labels that were changed for neonates didn’t add any beneficial new drugs to the NICU arsenal. “The most frequent drug labeling changes were [to state] that a drug was not effective or there was a safety concern,” said Laughon, who coauthored the study.

The most positive neonatal label changes have come from a National Institutes of Health initiative that put $25 million per year toward studying off-patent, older drugs in children — but that program’s funding runs out this year.

Pharmaceutical companies and institutional review boards continue to shy away from studying infants because they are fragile, cannot spare many blood samples, and are vulnerable to permanent injuries — injuries that, in the past, have been awarded large malpractice verdicts.

It’s also a small market, so pharmaceutical companies aren’t likely to make money by getting drugs approved for neonate use, Rousseau said. And few drug makers have dedicated pediatric teams.

“Any study that I’ve ever done that could potentially be sponsored by Big Pharma, but is in neonates, sends them running,” Sherwin said. “It’s hard to do, and it’s hard to get money to do it.”  

The majority of studies that have been done in neonates in recent years “were not able to establish efficacy,” wrote Dr. Susan McCune, deputy director of the Office of Translational Sciences at the FDA, who communicated with STAT via an emailed statement reviewed by the FDA’s press office.

It remains unclear whether that lack of success is due to the drugs themselves or trial design issues, she added. The latter is a major problem for the field. For example, many studies group all infants born two or more months before their due date into a single category, yet a 4-week-old infant born at 24 weeks gestation is not the same as a 1-day-old born at 28 weeks.

Last year, Davis and colleagues analyzed 25 proposed neonatal trial designs submitted by pediatric clinical trial groups and individual researchers, and found that six had fatal flaws — like requiring six EKGs on babies in the first day of life — that made it unlikely the trials would ever be successfully completed.

Davis hopes the new INC guidance will change that. At Tufts, he conducts about 10 neonatal trials at any one time. Among them: A clinical trial he’s been running for the past three years to test whether morphine or methadone, the two most common medications used in infants suffering from neonatal abstinence syndrome, is more effective.

Despite widespread use of both medications, doctors simply don’t know.

As Davis walks through the NICU, talking to nurses about new arrivals, Aubrey Baptista cradles Elijah, still with an IV port buried in his tiny hand for another round of antibiotics. Would she have enrolled her newborn son in a clinical trial if asked?

Baptista hesitates. “Because it’s life or death ... I don’t know. It’s scary enough as it is.”

Follow STAT on Twitter: www.twitter.com/statnews