In the United States, more than 10 percent of babies are born prematurely, with two percent of live births occurring before the seventh month of gestation. These extremely premature infants are at high risk of brain injury linked to cerebral palsy and other neurological deficits. Work appearing in the June 25 Journal of Neuroscience from the lab of Frances Jensen, a neurologist at Children’s Hospital Boston and an HMS professor of neurology, uncovers a potential new therapy for treating brain injury in premature infants.
According to co–first author Simon Manning, a neonatologist at Brigham and Women’s Hospital and a research fellow in the Jensen lab, “While medical advances have improved survival for those extremely premature babies, we haven’t made very big inroads into preventing cerebral palsy, the handicap that is most commonly associated with being born very prematurely. There really is no specific pharmacological treatment to prevent these major neurological sequelae.”
Due to age-specific physiological differences, especially in the brain, physicians cannot apply to infants knowledge and treatment gathered from research on adults. And because the brain changes so quickly during early development, major differences also arise between prenatal and neonatal periods. One such difference is the brain’s response to injury.
“What the literature has told us,” said Jensen, “is that the pattern in premature infants is very different from that even just a few weeks later at term, such that the same stimulus, commonly lack of oxygen or lack of blood supply in the premature brain, gives a very different pattern of injury than in the term brain.” In babies born at term, these insults often cause seizures and strokes with resulting cortical damage. The same insults in a premature brain cause more damage to the white matter that lies beneath the cortex.
Jensen’s group investigated the most vulnerable white matter cell type, oligodendrocytes (OLs). These glial cells have small processes that wrap around axons when they mature, much like electrical insulation, allowing the nerve cells to conduct signals much faster. During development, OLs are exquisitely sensitive to the excitatory neurotransmitter glutamate, which tends to build up at abnormally high levels during brain injury. OLs in the premature brain express high levels of glutamate receptors, in some cases much higher than at any later stage in life. So oxygen deprivation or decreased blood flow, which causes glutamate levels to climb, leads to neuronal overexcitation. The result is excitotoxic cell death of OLs in the white matter of the premature brain. Long-term white matter injury, termed periventricular leukomalacia (PVL), is the predominant pattern seen in this population and is associated with cerebral palsy and cognitive deficits.
When Jensen and her colleagues asked why OLs are uniquely vulnerable to hypoxic or ischemic injury in premature infants, they started by examining glutamate receptors. In their previous experiments, blocking the AMPA subtype with the drug topiramate provided only partial protection from PVL. Since OLs were recently shown to express another subtype, NMDA receptors (NMDAR), Jensen’s group asked whether they were in the critical place at the critical time to induce PVL. They found that OLs highly express NMDARs during the PVL window in both rat and postmortem human tissue.
“There was no actual human data until now to validate working animal models of PVL,” said Delia Talos, a research fellow in the Jensen lab and HMS instructor in neurology who is the co–first author with Manning. “We have found similar expression patterns for NMDARs in rat and human brain tissue at the cellular and regional levels.”
To test the NMDAR contribution to PVL susceptibility, Jensen’s group treated immature rats for two days with memantine, an NMDAR blocker, following hypoxia and ischemia. Memantine treatment resulted in far less white matter damage.
Developed at Children’s nearly 20 years ago, memantine already is FDA-approved to treat Alzheimer’s disease. The drug “has binding kinetics such that it has a very fast on–off binding rate,” Jensen said. Memantine therefore blocks the NMDARs only briefly, minimizing interference with normal physiological NMDAR functioning. Since the drug does not effectively block NMDARs under normal conditions, it does not have the negative side effects seen with previous NMDAR blockers. Jensen’s group previously verified this limit in the immature rat, showing that memantine treatment does not cause neuronal damage or impair learning and memory at the behavioral or cellular level.
Adding to its clinical relevancy, memantine treatment in the current study improved long-term outcomes as well. In both humans and rats, cortical brain thinning usually follows PVL. Jensen’s group showed that memantine prevented the cortical thinning that occurs in immature rats two weeks after brain hypoxia-ischemia, again demonstrating the drug’s protective effect against PVL, now including both acute and chronic damage.
Importantly, memantine has great potential as a PVL treatment because, as this study demonstrates, it is effective when given after the insult. Though it is difficult to predict which premature babies will get hypoxic-ischemic brain injury, new technologies give Jensen hope that posttreatment will become highly plausible. “There’s a technology being developed here at Children’s Hospital … called near-infrared spectroscopy (NIRS), which measures brain oxygenation. That could be used as a screening tool to determine babies at risk for brain injury,” Jensen said. It would be an improvement over the current technology—magnetic resonance imaging—because “there are only so many times you can move the child to an MRI … whereas NIRS can be done at the bedside through probes placed on the skull.”
In the meantime, pretreatment may also be an option. According to Manning, babies born before 30 weeks gestation are highly at risk for PVL and may benefit from pretreatment with memantine.
Brain injury in premature infants remains a serious concern. The parallel demonstration of a new therapeutic target both in the animal model and in human tissue, amenable to a drug that is already in use in another human disease process, raises the hope that researchers are another step closer to being able to treat these vulnerable patients.
Conflict Disclosure: The authors declare no conflicts of interest.
Funding Sources: The National Institutes of Health Grants and the William Randolph Hearst Foundation, in addition to core support from the Mental Retardation Developmental Disorders Research Center
