The Need for Rigorous Evidence on Medication Use in Preterm Infants

Abstract

APPROXIMATELY 200 000 INFANTS BORN ANNUALLY IN the United States require admission to a neonatal intensivecareunit for treatmentofprematurity, costing more than $26 billion per year. Preterm infants are at substantial risk of death or developing serious morbidity that can affect them for life. Unlike treatments used in other fields of medicine, most medications administered to preterm infants lack convincing data to support their safety and efficacy with more than 90% not approved by the US Food and Drug Administration (FDA) for the prescribed indication. No new medications have substantially improved outcome for preterm infants since the introduction of antenatal corticosteroidsandsurfactant15 to20yearsago. Infantsadmitted to the neonatal intensive care unit may be exposed to more than60separatedrugs,with themostpremature infants receiving the greatest number of medications. Serious adverse drug reactions from single or multiple agents can significantly increase mortality and serious morbidity resulting in shortand long-term adverse consequences. It is important to solve this knowledge gap and define systems for drugs to be adequately studied. Without this, each preterm newborn is essentially being enrolled in an uncontrolled and unapproved clinical trial that will not yield data of substantial value. Congress passed the Best Pharmaceuticals for Children Act (BPCA) in 2002 (reauthorized in 2007 and 2012) to stimulate the study of pediatric therapeutics. The BPCA authorizes the FDA to offer drug manufacturers an extra 6 months of patent protection (exclusivity) for testing their products in children, although no specific age range is mandated. In the past 14 years, the FDA has updated the labeling of 434 drugs following completed studies in children, but only 1 involved premature infants. This is primarily due to certain disincentives to conducting research in preterm infants. Many institutional review boards and the FDA struggle with this vulnerable population, with some suggesting that it is unethical to conduct any research in newborns. Preterm infants also represent a relatively small market and can develop permanent injuries, which can be associated with large malpractice awards whether the adverse outcome is caused by the drug or not. Many companies understandably limit their studies to older children, an approach that still permits extended exclusivity under the BPCA. The BPCA also requires the National Institutes of Health to prioritize therapeutics in need of study (eg, pediatric formulations), to sponsor necessary pediatric clinical trials, and to submit thedata to theFDAfor labelingchanges. While the National Institute of Child Health and Human Development has effectively prioritized and sponsored a number of clinical trials in neonates, no funds have actually been appropriated for this program. This severely limits the number, type, and quality of studies that can be performed. The decision to use a drug in neonates is often based on a number of factors, including the clinical impression of the prescribing physician, an expert opinion, studies in older children, or a pilot study in newborns. In the absence of definitive information, it is imperative that initial safety and pharmacokinetic studies are followed by trials of sufficiently rigorous design and power to establish whether the benefits of using the drug in neonates exceed the risk. Because of the considerable morbidity and mortality intrinsic to preterm infants, and their complex physiology, randomized, masked, placebo-controlled trials are essential and should be designed to assess the superiority of one drug over another (vs no treatment at all). Although initial studies can use short-term outcomes, these studies must be followed by continued surveillance at least until the child reaches school age to more accurately define outcomes. For example, caffeine was initially shown to significantly reduce apnea, bronchopulmonary dysplasia, and cerebral palsy in preterm infants. However, follow-up at 5 years of age demonstrated that caffeine was no longer associated with significantly improved rates of survival without disability. Determining long-term responses to therapy administered to infants shortly after birth will be challenging, expensive, and will require commitment and novel approaches by all key stakeholders. The most important required change is that unapproved medications used in preterm infants should be assessed in randomized controlled trials to establish shortand longerterm safety and efficacy. Rigorous evaluation of complex therapies used in infants and children is possible. For ex-