Donna Ferriero, MD
Research Program Injury Response in the Developing Nervous System
Over the past 14 years, my research has been directed at discovering the pathobiology of hypoxic-ischemic injury in the developing nervous system. This process differs markedly from that in the mature nervous system and through a series of investigations, both in vivo and in vitro, in animals and in humans. We have two prospective human research studies ongoing to delineate injury patterns and identify prognostic factors that will determine neurodevelopmental outcome after perinatal asphyxia. We have studied over 100 preterm (PREMRI Premature Asphyxia Magnetic Resonance Imaging) and 150 term (BAMRI Brain Asphyxia Magnetic Resonance Imaging) infants using state-of-the art magnetic resonance (MR) imaging modalities such as spectroscopy, diffusion tensor imaging, and perfusion mapping.
There is a widely held view that the neonatal brain is resistant to injury from neonatal seizures because animal models of neonatal seizures have been unable to detect cell death. Using MR spectroscopy, we have discovered that although there is no frank structural damage, there is indeed injury as measured by cellular metabolism MRS metabolites (NAA, lactate) and explains why these patients have poor neurodevelopmental outcome. Based on these findings, we have developed a paradigm for predicting outcome using early clinical markers (seizure scores) and MRS metabolites (n acetyl aspartate, lactate). The ability to predict outcome will enable investigators to select newborns for clinical trials for therapies for this devastating condition, and also will assist physicians and families caring for these families to prepare them for the future. We are currently involved in the first international clinical trial for severe perinatal asphyxia and I am on the scientific advisory board utilizing selective head cooling.
In our BAMRI term newborn population, we have found a striking correlation with cytokines found in the neonatal blood spot obtained for newborn screening for metabolic disorders mandated by the State of California Department of Health Services, and future development of cerebral palsy.
In our PREMRI study, we have found that conventional imaging modalities like head ultrasound do not identify the most common predictor of poor outcome in the premature infant, periventricular white matter injury.
Links with the animal studies:
The immature brain is uniquely and exquisitely sensitive to oxidative stress, which accounts for the cell death seen after perinatal asphyxia. As in all neurological diseases, there are vulnerable regions of injury and these regions are damaged in an age-dependent manner. For example, in the premature brain, we see injury in the periventricular white matter that reflect damage to the developing oligodendrocytes, but we also have preliminary data to suggest that subplate neurons are involved as well . In the term baby, we see a different pattern of vulnerability, with damage to the deep gray nuclei and this pattern of vulnerability can be used to predict neurodevelopmental outcome later in life as described above. This pattern may be due to the unique distribution of a network of cells with a rich supply of glutamate receptors that parallels the distribution of a population of neurons containing neuronal nitric oxide synthase that are uniquely spared after hypoxia-ischemia at this age. Since these cells are spared, they are available to signal relentlessly and release the free radical nitric oxide and oxidatively challenge the newborn brain. Selective destructive or targeted elimination of the nitric oxide synthase neurons results in protection in animal models, although pharmacological inhibition of the enzyme has not been satisfactory due to the non-selectivity of the compounds and to the fact that there is some depression of enzymatic activity immediately after the insult.
We have also shown that hypoxic-ischemic injury results in prolonged and delayed cell death both locally and in remote regions after the insult. These data have important consequences for future studies because they suggest that there is prolonged window of opportunity for the administration of neuroprotective agents.
In regard to the unique differences of the immature nervous system, we have shown that therapies used for the adult nervous system can not be assumed to work for the newborn brain. For example, overexpression of superoxide dismutase had been successful in treating adult stroke, but in the immature animal, this overexpression leads to more injury because the neonatal animals have less antioxidant reserve. This was the first example where therapy assumed to be neuroprotective for all ages actually had deleterious effects in the immature brain. Subsequently, other reports followed for other agents (Ikonomidou et al, Science 283, 1999).
We have recently developed a model of neonatal focal ischemia since neonatal stroke occurs in 1/4000 live births. The neonatal brain has a robust inflammatory response to ischemic injury and the blood brain barrier does NOT break down easily, contrary to the belief widely held. We have used MR perfusion imaging coupled with anatomical studies to document these findings. We have also demonstrated that stem cells in the subventricular zone migrate to the areas of infarction. These preliminary data offer exciting avenues for future studies for recovery after stroke.
We have coupled these studies with an ongoing human study called ReFINES (Risk Factors in Neonatal Stroke). In this study, we have identified over 70 children who have had stroke in the perinatal period, and we have accessed their neonatal blood spots through the California Department of Health Services in a similar manner to the asphyxia trial described above. We will screen for over 100 genetic mutations for thrombophilias, adhesion factor abnormalities, vascular wall abnormalities, etc. in an effort to identify potential risks. We will also screen parents who consent in the study. One advantage to studying children is that many of the confounders such as disorders like atherosclerosis, diabetes, and hypertension have not yet developed at the time of stroke. We have already described neonates with intracerebral venous thrombosis and will continue to identify this important cohort.
It is our goal to accurately describe the selectively vulnerable populations of cells after hypoxic-ischemic insults at the different stages of development, and in so doing predict outcome. Designing more accurate animal models based on these descriptions to study pathophysiology will enable use to more accurately target therapies. I have organized a team of investigators who are multidisciplinary and dedicated to this goal.
In particular we will continue to:
1) Map selectively vulnerable regions after neonatal insults at different developmental stages.
2) Correlate MR findings with neurodevelopmental sequalae- current projects include mapping corticospinal tracts with diffusion tensor imaging with neuromotor findings, and vision-motor with oculomotor development.
3) Correlate neonatal and parental cytokine and coagulation and vascular abnormalities with patterns of stroke and outcome.
4) Use MR modalities to predict outcome after perinatal asphyxia.
5) Use MR patterns to develop better animal models to study pathophysiology of hypoxia-ischemia.
6) Develop more advanced and safer imaging modalities for the developing nervous system that will enable the study of evolving damage with serial imaging.
7) Use genetic strains of mice that are resistant and susceptible to injury and explore genomics and proteomics for novel genes and proteins that may shed light on possible probes for signaling pathways for protection.
8) Develop appropriate therapies based on information gained from animal and human studies of the developmental profiles of injury and pathogenesis.