Cerebral Perfusion, Oxygenation, Electrical Activity

Overview

The specific aim of the research proposal in preterm infants with IVH and PHH who require placement of an Omaya reservoir or a shunt is to determine if decreasing ventricular volume improves, middle cerebral artery flow, cerebral oxygenation, and cortical neuronal electrical activity. To accomplish this aim, we will simultaneously perform the following evaluations prior to shunt placement or prior to and after routine CSF aspiration from reservoir in: 1. middle cerebral artery velocity time integral and resistive index using Doppler ultrasonography 2. cerebral oxygenation using near infrared spectroscopy (NIRS) 3. background neuronal electrical activity using an EEG. In addition, we will measure serial CSF concentration of neuroproteins, S100B, GFAP, NSE, TGF-ß, and IL-6, as evidence of ongoing neuronal damage and correlate the concentration with cerebral perfusion and activity as measured above.

Very low birth weight infants are at risk for developing intraventricular hemorrhage (IVH). Post hemorrhagic hydrocephalus (PPH) is a major complication of IVH and contributes to long-term developmental delays. Progressive PHH often requires management of ventricular dilation by adjusting CSF fluid volume. There are 3 methods employed to acutely decrease CSF volume: 1) serial lumbar puncture, 2) open drain system (continuous CSF removal), and 3) reservoir system. Serial lumbar puncture is only effective if there is communicating hydrocephalus. The open drain system is infrequently used as it is cumbersome and there is a relatively high risk of infection. The method most commonly used to manage CSF volume is the reservoir system where an in-situ drain is connected to a subcutaneous (Omaya) reservoir which is periodically aspirated by needle puncture through the scalp. Because the Omaya system is closed, intraventricular volume and, thus, pressure must necessarily rise prior to and decrease after CSF aspiration. Ventricular dilation is controlled by the frequency and volume of CSF aspiration. When hydrocephalus continues to be a problem despite removal of CSF, a ventricular-peritoneal (VP) shunt is placed. Approximately 50% of infants with hydrocephalus treated with removal of CSF resolve their hydrocephalus and do not require VP shunt placement. Placement of a VP shunt is difficult in extremely preterm infants due to increased risk of ulceration around the shunt site and the high protein concentration in CSF which can occlude the valve in the VP shunt requiring revision. Thus, hydrocephalus is usually treated with serial removal of CSF to allow for identification of those infant's whose hydrocephalus resolves over time. However, timing and method of CSF management is controversial because the effect of increasing hydrocephalus on cerebral perfusion, oxygenation, electrical activity, and neuronal damage has not been established. Serial removal of CSF causes a change in intracranial volume/pressure that can be potentially transmitted to intracranial vessels. The caliber of cerebral vessels may be modified by this balance between intravascular and intracranial pressure, and if the caliber should change, the blood flow characteristics should also change. In cerebral veins and capillaries where intraluminal pressure is low, high intraventricular pressure may significantly affect blood flow and can cause venous stasis. NIRS measures cerebral oxygenation in capillaries and veins. Intracranial arteries and arterioles may be somewhat less affected, except when intraventricular pressure greatly increases. The resultant decrease in the arterial supply can affect tissue perfusion. Arterial flow can be measured by Doppler ultrasonography. Thus, there is concern that during periods of increasing or fluctuating ventricular size cerebral arterial perfusion may be compromised and further cerebral injury may result. Intracranial pressure is further influenced by the plasticity/deformability of the immature brain and the easy expansibility of the cranial vault due to the presence of sutures and open fontanelles. There is indirect evidence from experiments in animals that ventricular distention itself may cause secondary brain injury. Thus, axonal stretching and disruption secondary to progressive ventriculomegaly is be associated with gliosis. Periventricular vascular distortion and compression may decrease cerebral blood flow causing ischemic injury to periventricular white matter. Inflammation and repair may interfere with CSF flow. Little research is available that helps answer the primary question involved in clinical management of ventricular dilation in premature infants: Are there relationships between ventricular enlargement, cerebral perfusion, brain oxygen delivery, and on-going cerebral damage? Doppler ultrasonography has been applied in premature infants to characterize post delivery changes in arterial and venous cerebral blood flow velocities.\[1\] Critically low superior vena cava flow measured by Doppler ultrasonography in the first 24 hours of life has been associated with intraventricular hemorrhage in premature infants born before 30 weeks gestation.\[2\] Using pulsed Doppler ultrasonography, the resistive index can be measured which is a inversely related to blood flow. In older infants with established hydrocephalus, cerebral blood flow resistive index appeared to be a good indicator of increased intracranial pressure.\[3\] However, in a review article, the value of Doppler indices alone in predicting increased intracranial pressure was strongly questioned.\[4\] In a recent article, the resistive index of the anterior cerebral artery decreased significantly after CSF drainage in infants with PHH. \[5\] Although Doppler ultrasonography is easy to perform, its role in detecting significant changes in cerebral perfusion associated with increased CSF volume and ventricular dilation is not yet established. Near infrared spectroscopy (NIRS) is a portable non-invasive technique used to measure regional cerebral oxygen saturation in cerebral blood.\[6\] NIRS has been used to evaluate effect of head \[7\] and body position \[8\] on cerebral hemodynamics in preterm infants, and the effect of certain treatments, such as surfactant administration \[9\] and suctioning on conventional or high-frequency ventilation.\[10\] NIRS measurement of cerebral oxygenation and hemodynamics shows impairment in those neonates at risk for developing severe brain ischemia.\[11\] In premature infants with PHH, CSF fluid aspiration was associated with a significant increase in cerebral perfusion, cerebral blood volume, and oxidative metabolism.\[12,13\] In a small animal model of acute hydrocephalus, NIRS measurement of global cerebral blood flow using oxygen as a tracer was highly correlated to cerebral blood flow measured by radioactive microspheres over a wide range of increased intracranial pressure.\[14\] As intracranial pressure increased, NIRS measurement and absolute measurement of cerebral blood flow decreased. NIRS measured cerebral perfusion appears to directly reflect absolute cerebral blood flow and is sensitive enough to detect significant changes in cerebral perfusion which occur with evacuation of CSF in infants with PHH. Amplitude-integrated EEG (aEEG) is a device used for neurologic surveillance. The cortical electrical activity is transformed into a single signal that represents overall electrocortical background activity of the brain. Specifically the upper and lower voltage margins and the amplitude of the tracing and the presence of seizure activity can be evaluated. In addition, the presence of sleep-wake cycles, a rhythmic sinusoidal variation in amplitude, can be evaluated. Infants \> 34 weeks gestation have predictable aEEG patterns. The signal depends on gestational age and postnatal age and standards for preterm infants have been reported.\[15,16\] In a case report, 2 preterm infants with PHH demonstrated abnormal sleep-wake cycles and markedly decreased cerebral electrical activity with increasing ventricular enlargement prior to clinical signs of increased intracranial pressure. In one infant, aEEG normalized after VP shunt placement.\[17\] To date, there are no clinical studies on the effect of aspiration of CSF on aEEG in preterm infants with PHH. Biomarkers for cerebral injury have been used to predict severity of injury and long-term outcome, identify patients early at risk for poor neurologic outcomes, and evaluate effectiveness of therapeutic interventions. Neuronal specific enolase (NSE), a marker for neuronal damage, and S100B, secreted by astrocytes and a marker of glial/neuronal injury, have been shown to be increased in children after traumatic brain injury.\[18\] In preterm infants with PHH, S100B and glial fibrillary acid protein (GFAP), a structural protein in astrocytes, were significantly elevated in those infants who had brain parenchymal lesions and poor neurological outcomes.\[19\] These biomarkers are quickly metabolized and thus ongoing elevation would indicate ongoing damage. Transforming growth factor beta (TGF-ß), produced by fibroblasts and released into the CSF after injury, stimulates production of extracellular protein which could result in PHH due to a permanent obstruction to CSF flow.\[20\] Cytokine IL-6 is a marker of inflammation. There are no longitudinal studies on the effect of aspiration of CSF on these markers of cerebral damage in preterm infants with PHH.