Infantile Posthemorrhagic Hydrocephalus

Published on 13/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 3230 times

CHAPTER 187 Infantile Posthemorrhagic Hydrocephalus

Yin C. Hu, Shakeel A. Chowdhry, Shenandoah Robinson

Intraventricular hemorrhage (IVH) is a leading cause of infantile hydrocephalus. The management of hydrocephalus in these infants is critical because the initial neonatal treatment can have lifelong implications for both shunt survival and the child’s neurological outcome. An understanding of the unique challenges in caring for these infants is essential to their proper management. IVH and subsequent posthemorrhagic hydrocephalus (PHH) arise from different pathophysiologic mechanisms, pose different clinical and management issues, and suggest different prognoses in preterm versus term infants. Thus, they are discussed separately.

Intraventricular Hemorrhage in Preterm Infants

Epidemiology

The incidence of preterm birth continues to rise in the United States, with 12.7% of infants born before 37 weeks’ estimated gestational age (EGA), representing an almost 20% increase over the past 15 years.¹ Preterm infants are stratified by birth weight or EGA, or both. For classification purposes, the EGA of preterm infants is not rounded up. For example, an infant born at weeks is classified as a 26-week EGA infant.² The survival of all preterm cohorts has risen dramatically over the past few decades owing to improvements in perinatal medicine, but many of these infants remain at risk for neurodevelopmental deficits. According to the 2007 report from the Institute of Medicine’s Committee on Understanding Preterm Birth, preterm infants account for 64% to 75% of infant mortality, 42% to 47% of children with cerebral palsy, 27% of children with cognitive deficits, 37% of children with visual impairments, and 23% of those with hearing impairments.³ Neurodevelopmental outcomes are discussed in more detail later.

The prevalence of IVH varies, depending on the availability and quality of neonatal intensive care, and only a small subset of patients requires neurosurgical intervention. The incidence of IVH increases inversely with decreasing birth weight or EGA.⁴ In a study of infants born in the mid-1990s weighing less than 1500 g, 22% had IVH, and one fourth of those infants had progressive ventricular dilation.⁵ Among survivors with progressive ventricular dilation, one third required surgical intervention.⁵ Although infants born more recently continue to suffer IVH, fewer require surgical intervention. A recent collaborative study from the National Institutes of Health (NIH) Neonatal Research Network following more than 6000 preterm infants born at less than 1000 g between 1993 and 2002 showed that one third suffered IVH, and one third of those with IVH developed progressive ventricular dilation.⁶ Ten percent of those with IVH (3% of the total weighing <1000 g) required shunt insertion for symptomatic PHH, a decrease from 15% in older studies.⁶ IVH is graded on Papile’s scale from I to IV (Table 187-1).⁷ Although only 1% of infants with grade I or II IVH in this recent NIH study required shunt insertion, 18% of infants with grade III and 29% with grade IV IVH required a shunt.⁶ As perinatal care continues to improve, the proportion of preterm neonates who require shunt insertion will likely continue to decline.

TABLE 187-1 Papile’s Classification of Preterm Intraventricular Hemorrhage on Ultrasonography

GRADE	DESCRIPTION
I	Isolated germinal matrix hemorrhage
II	Intraventricular hemorrhage without ventricular dilation
III	Intraventricular hemorrhage with ventricular dilation
IV	Intraparenchymal plus intraventricular hemorrhage

Because less than 10% of infants with IVH develop PHH and require a shunt, strong collaboration between the neurosurgical team and the neonatology team is required to obtain neurosurgical consultation in a timely manner and to avoid unnecessary nonsurgical consults. Intravenous access in cranial veins should be avoided in any preterm infants who might develop hydrocephalus.

Pathophysiology of Germinal Matrix Hemorrhage and Posthemorrhagic Infarction

The germinal matrix is a transient structure in the subependyma of the ventricular walls that generates neural cell progenitors for the overlying cortex, primarily from 8 to 28 weeks’ EGA. The germinal matrix typically completes involution by 32 weeks’ EGA. The proliferative germinal matrix is a metabolically active area with friable, immature vessels prone to hemorrhage.⁸ Hemorrhage can be limited to within the germinal matrix (grade I), but about 80% of germinal matrix hemorrhages rupture through the ventricular wall into the ventricle (grade II, III, or IV; see Table 187-1). Grade IV is also called periventricular hemorrhagic infarction.⁹ Venous infarction is believed to result from the germinal matrix hemorrhage rather than direct expansion of the hemorrhage into the parenchyma.

Multiple factors contribute to the germinal matrix’s propensity to hemorrhage, and the pathophysiology is incompletely understood. Because of their immaturity, preterm infants have impaired cerebral autoregulation. Systemic fluctuations in blood volume, flow, and pressure that commonly occur in preterm infants are thus transferred to the friable germinal matrix without the buffer of cerebral autoregulation. In addition, the germinal matrix has lower levels of fibronectin and other extracellular matrix components than do other areas of the late-gestation brain, and this likely contributes to the propensity to hemorrhage.¹⁰ An experimental study suggests that prenatal betamethasone treatment may enhance fibronectin levels and thus decrease the risk of hemorrhage. This correlates with the decreased risk of IVH observed clinically with prenatal betamethasone treatment.¹¹ Although neonatologists have worked aggressively with obstetricians to minimize stress to preterm infants during the peripartum, the risk for IVH most likely reflects a combination of genetic and environmental factors. After IVH occurs, the absorption of cerebrospinal fluid (CSF) is impaired, likely caused by a combination of decreased absorption across the arachnoid villi into veins and from ependyma into parenchyma.⁸

IVH can occur in association with periventricular leukomalacia (PVL), injury to the developing white matter. PVL can occur as both deep cystic white matter lesions and diffuse noncystic loss of oligodendrocytes with associated astrocyte gliosis.¹² Cystic PVL is visible on cranial ultrasonography (US), and most attention has focused on the white matter injury. Observations during the past decade from magnetic resonance imaging (MRI) and autopsy studies have shown that injury to the developing central nervous system affects both white and gray matter.¹³ The combined injury has been designated encephalopathy of prematurity, which more accurately reflects the widespread nature of the injury not only to the cerebral gray and white matter but also to the diencephalon, brainstem, and cerebellum.¹⁴ PVL is about 10-fold more common than IVH among infants weighing less than 1500 g at birth. At least half of these infants have imaging signs of PVL, but only 5% have IVH.¹⁴

Clinical Presentation and Diagnostic Evaluation

Most IVH occurs within the first 72 hours of life, when most preterm newborns are quite unstable (Fig. 187-1). IVH is readily diagnosed by bedside cranial US. Transport to obtain computed tomography (CT) or MRI can cause additional stress and potential injury in these very fragile patients. CT scans also involve radiation, and repeated exposure of the preterm brain to radiation should be minimized if possible. In general, CT scans should be reserved for the evaluation of diffuse, life-threatening cerebral hemorrhages when MRI is not feasible. US provides a reliable means of identifying ventricular dilation and IVH. If possible, all preterm newborns weighing less than 1500 g should have a cranial ultrasound examination within the first 48 hours of life, and then weekly or biweekly as needed. PVL can also be assessed by US. Although the interreader reliability for grades III and IV IVH is excellent, the reliability for grades I and II IVH and for PVL is less.¹⁵ These limitations of cranial US rarely present a problem for neurosurgical management (Fig. 187-2).

FIGURE 187-1 Typical imaging findings in posthemorrhagic hydrocephalus from germinal matrix hemorrhage after preterm birth at 27 weeks’ gestation. Cranial ultrasonography (US) in the coronal (A) and axial (B) planes at 12 hours of life shows minimal abnormalities. Repeat US 2 days later reveals bilateral grade IV intraventricular hemorrhage (C and D), also termed periventricular hemorrhagic infarction. The patient did not develop signs of symptomatic hydrocephalus as a neonate. At term equivalent, magnetic resonance imaging shows a paucity of white matter, suggesting the diffuse form of periventricular leukomalacia (E). Also note the small cerebellum, a typical finding among former preterm infants (F). At 8 months of age he developed symptomatic hydrocephalus, with an enlarging head circumference and dilated ventricles observed on CT (G). Several months later, a baseline CT scan shows well-decompressed ventricles, but the paucity of white matter remains (H, arrow).

FIGURE 187-2 Progressive ventricular dilation on cranial ultrasonography (US) in the neonatal period. Initial US of this infant born at 27 weeks’ gestation shows bilateral grade II intraventricular hemorrhage (IVH) (A). Although distinguishing between the lower grades of IVH on cranial US may be difficult for the average observer, the progressive ventricular dilation, with retraction of the germinal matrix clot, at 2 weeks (B) and 3 weeks (C) is readily apparent. After B, serial lumbar punctures were initiated for an enlarging head circumference with splayed sutures, but they failed to control the hydrocephalus. After C, a ventriculosubgaleal shunt was inserted.

Most preterm infants develop symptomatic hydrocephalus over several days, with a gradual shift in their clinical course. Infants with IVH should be observed closely with daily measurement of the occipitofrontal circumference. An increase of 0.5 to 1 cm/day for 2 to 3 consecutive days often suggests symptomatic hydrocephalus. The anterior fontanelle in very small infants is often quite large; usually the shift from a soft, slightly bulging fontanelle to a tense, full one can be appreciated. Widening of the space between the bone edges along the sagittal and coronal sutures (splayed sutures) is often a reliable finding on daily examination to assess the development of symptomatic hydrocephalus. Preterm infants with symptomatic hydrocephalus can have episodes of spontaneous apnea or bradycardia. Other symptoms and signs of increased intracranial pressure include refractory seizures, lethargy, and impaired upward gaze (“sunset” phenomenon). In addition to the daily clinical examination, serial ultrasound studies every few days are useful to assess for additional hemorrhage or the development of ventricular dilation. Rarely, preterm infants experience acute clinical deterioration due to a rapid onset of increased intracranial pressure. In these patients, more extensive and severe intracerebral hemorrhage may have occurred, often in association with other systemic problems such as sepsis and coagulopathy. Urgent interventions may be needed to sustain these infants, but the heroic nature and poor long-term outcome of these interventions should be discussed with the parents.

Hydrocephalus ex vacuo refers to ventricular dilation without increased intracranial pressure. In the mid-1990s, 24% of surviving infants with IVH had ventricular dilation without progression.⁵ Hydrocephalus ex vacuo reflects encephalomalacia, or a lack of adequate parenchymal brain growth; the implications for developmental outcomes are discussed later. Infants can have transient symptomatic hydrocephalus that spontaneously resolves. Distinguishing hydrocephalus ex vacuo from symptomatic hydrocephalus can be challenging. With continued observation, the type of hydrocephalus typically declares itself.

Treatment

Preterm infants are challenging patients. Their relatively poor nutrition, immature immune system, and other comorbidities make them less than ideal surgical candidates. For example, many preterm infants have anemia of prematurity treated with erythropoietin, red blood cell, and platelet transfusions. Even with scrupulous attention to minimizing blood loss during surgery, a preterm neonate may require a transfusion after a surgical procedure. Because grades III and IV IVH tend to occur in the smallest and sickest infants, they are more likely to have other confounding illnesses. Interventions for symptomatic hydrocephalus are offered in a stepwise progression to identify patients in whom permanent CSF diversion is not required. This process allows surgery to be avoided in all but the few who definitely need it. It also shows the family that every reasonable nonsurgical measure has been tried. This is important because neither CSF shunts nor endoscopic third ventriculostomy offers highly effective lifelong treatment without complications.

Nonsurgical Treatment

A commonly used treatment paradigm begins with serial lumbar punctures. In many cases, serial lumbar punctures can be used as a temporizing measure until the infant is older and more medically stable and thus a better surgical candidate. Ideally, lumbar punctures should be initiated as soon as progressive ventricular dilation is observed.¹⁶ This may reduce the need for a permanent shunt. The fourth ventricle is sometimes difficult to visualize by cranial US, and the tentative diagnosis of aqueductal stenosis may be suggested by the radiologic interpretation. As long as there is no large posterior fossa hemorrhage, most preterm infants can safely undergo lumbar puncture. Fortunately, progressive ventricular dilation after IVH spontaneously resolves in a significant proportion of infants. Infants with transient hydrocephalus should be followed for the late development of symptomatic hydrocephalus. The chance of surgical intervention becoming necessary continues to decline for these infants over time. If serial lumbar punctures are inadequate and the patient is a relatively stable candidate for surgery, a temporary CSF diversion device is used to delay the need for definitive surgical treatment as long as possible. A meta-analysis of repeated lumbar or ventricular punctures failed to show that CSF removal improved outcomes compared with no intervention.¹⁷ This study, however, included ventricular taps, which have a higher rate of complications and poorer outcomes; it also included patients without progressive ventricular dilation. Serial CSF removal would not be expected to improve the outcome for children with hydrocephalus ex vacuo.

As observed in many areas of medicine in which definitive surgical treatment is not ideal, many nonsurgical interventions have been attempted to minimize the need for neurosurgical intervention. Unfortunately, these interventions have not been effective. Currently, medical therapy is not recommended for preterm infants with symptomatic hydrocephalus. Diuretic therapy, including acetazolamide and furosemide, is not effective in this population and may increase the risk of nephrocalcinosis and other complications.¹⁸ Likewise, intraventricular streptokinase is not effective in this population.¹⁹

Temporary Surgical Interventions

Some infants do not obtain adequate relief of symptomatic hydrocephalus with serial lumbar punctures, as defined by stabilization of the head circumference on a reasonable growth curve and stabilization or reduction of ventricular dilation on cranial US. In these cases, temporary shunts are used to treat symptomatic hydrocephalus and allow the infant to grow and recover from other complications of prematurity before undergoing definitive surgical therapy. Preterm infants have immature immune systems, fragile skin, and impaired wound healing. In addition, infants weighing less than 1500 g often have insufficient peritoneal absorptive capacity to handle the CSF volume. Also, in the subset of infants with transient symptomatic hydrocephalus that persists only for several weeks, temporary shunts avoid permanent shunt placement. Clinical judgment is crucial to balance the timing and risk of surgical intervention against the potential detriment of temporarily and inadequately treated hydrocephalus. For example, preterm infants with a history of bacterial infection of the blood or CSF are at higher risk of significantly poorer neurological outcomes than comparable infants without infection.²⁰ The need to weigh the potential detriment of delaying surgery for several days to allow infection to be eradicated against the risks associated with symptomatic hydrocephalus is not uncommon for pediatric neurosurgeons and neonatologists.

Ideally, the neurosurgical team begins interacting with the infant’s family when it is clear that serial lumbar punctures are failing. Because surgical intervention for this population is associated with a relatively high rate of complications, including potential shunt revisions and infections, the family needs to understand that all nonsurgical means were exhausted. Infants with symptomatic hydrocephalus have poorer neurodevelopmental outcomes than those without hydrocephalus.²¹ Therefore, the family needs to be advised that treatment of hydrocephalus is necessary to optimize brain development, but it does not mitigate the neurological deficits associated with preterm IVH.

Temporary ventricular CSF diversion can be achieved with either a ventricular access device ²²^,²³ or a ventriculosubgaleal CSF shunt.²⁴^,²⁵ A ventricular access device consists of a reservoir that caps a ventricular catheter, and the reservoir is accessed via a needle through the scalp every 12 to 48 hours to withdraw enough CSF to maintain a stable head circumference. Ventriculosubgaleal shunts are also composed of a ventricular catheter and reservoir, with a 4- to 8-cm piece of shunt tubing exiting the reservoir and ending in a subgaleal pocket. CSF collects in the subgaleal scalp pocket and is slowly reabsorbed. When the subgaleal shunt is functioning, the CSF forms a large, sometimes tense pocket in the scalp, but the fontanelle is soft and flat, and the cranial sutures are not splayed.

Both temporary shunt methods are widely used, and some institutions may have slightly better results with one method than the other. With both procedures, meticulous technique is essential to minimize complications. No randomized multicenter trial comparing the two methods has been performed. Temporary shunts can be used for several months if necessary. Both methods can provide adequate temporary relief from symptomatic hydrocephalus, and both are prone to infection, wound dehiscence, and catheter occlusion.²²^,²⁶ Some have proposed that subgaleal shunts provide more consistent ventricular decompression compared with the fluctuation in ventricular size that occurs with daily aspiration from a ventricular access device, but the daily fluctuation has not proved harmful. Specially trained nonphysician health care providers can safely and reliably perform shunt taps.

Direct ventricular taps with a needle through the fontanelle should generally be avoided as a routine treatment modality. Ventricular taps are reserved for life-threatening emergencies because they are associated with a markedly increased risk of infection and loculated hydrocephalus in childhood. Similarly, external ventricular drains have a much higher risk of complications in preterm infants than in older neurosurgical patients.²⁷ They have a higher risk of infection than temporary internalized shunt devices and have the same risk of catheter occlusion from debris in the CSF. A randomized trial comparing standard treatment with aspiration through a ventricular access device and a more aggressive protocol consisting of drainage, irrigation, and fibrinolytic therapy showed no difference in the number of patients who died or required permanent shunts, and those receiving more aggressive therapy were much more likely to have a secondary IVH.²⁸

Permanent Cerebrospinal Fluid Diversion Procedures

Permanent ventriculoperitoneal shunts are required in only a small minority of preterm infants who suffer IVH. As mentioned earlier, the family needs to understand that although shunts and other surgical procedures are lifesaving, they are not without risks and complications. These children are prone to cerebral palsy, epilepsy, cognitive delay, and behavioral abnormalities regardless of whether the CSF diversion is effective. Insertion of the permanent shunt is frequently delayed through the use of a temporary shunt as long as feasible. This likely decreases the risk of shunt infection.²⁹ Permanent shunt insertion at an older age in the neonatal period also likely decreases the need for shunt revisions throughout childhood. Among 79 preterm infants with grade III or IV IVH born before 32 weeks’ EGA, weighing less than 1500 g, and followed for a mean of 10 years, the shunt revision rate for those with permanent shunts inserted at 3 to 5 months of age (uncorrected) was about half (55%) the shunt revision rate for infants with permanent shunts inserted before 3 months (Robinson and colleagues, unpublished results). Some have advocated that low-pressure valves be inserted initially in preterm infants. However, in a retrospective study of children with infantile hydrocephalus of all causes, medium- and high-pressure valves inserted during infancy were associated with a lower shunt revision rate during childhood.³⁰ Many former preterm infants develop slit ventricle syndrome during childhood, which can be associated with more shunt revisions. The high-pressure valve may deter slit ventricle formation in some but not all patients.

Shunt Technique

Recent studies to minimize shunt complications have shown that consistent techniques and protocols are effective in reducing some risks, particularly infection. Despite our best efforts, however, infections and other complications are still likely in this vulnerable population with inadequately developed immune systems. Shunt insertion in small infants also requires meticulous attention to positioning, aseptic technique, and wound closure. Local anesthetic (0.1% lidocaine with epinephrine 1 : 1 million, up to 4 mL/kg body weight) infiltration of the incisions minimizes blood loss and the need to coagulate the fragile scalp. Both frontal and occipital approaches can be used, with equal rates of long-term success. Unless there is concern for infection, we typically place the permanent shunt through the same opening used for the temporary shunt. Care is taken during opening of the incision to preserve the galea to facilitate optimal wound closure and to avoid cutting through the dura near the fontanelle. We typically make a bur hole adjacent to an open cranial suture. To make a bur hole, the pericranium is coagulated, and the bone is removed with curets and Kerrison rongeurs. Hemostasis in the bone can often be achieved with coagulation. After the dura and pia are coagulated and incised, the ventricular catheter is inserted using standard landmarks, and CSF is collected for protein, glucose, cell count, and culture. For permanent shunts, the catheter should be as long as safely possible (e.g., 5 cm for frontal) to minimize the chance the child will outgrow it as the cerebral cortex expands. Many infants have relatively thin cerebral cortices, and care is taken to avoid overdrainage that may precipitate a subdural collection. Low-profile shunt hardware is ideal, and care should be taken to minimize suture knots that may protrude through the thin skin. In some patients it may be necessary to approximate the shunt reservoir to the pericranium while the wound heals. The choice of shunt components depends on the surgeon’s preference, and no components have proved more effective or less problematic than others.

For a ventricular access device, the only components are the reservoir and ventricular catheter, and the wound is closed after they are inserted and secured. For a ventriculosubgaleal shunt, the subgaleal pocket is created from near the catheter insertion site to the contralateral hemisphere, extending as far as possible and at least 8 cm in diameter. A 4- to 8-cm length of tubing is attached to the outflow side of the reservoir. Ideally, after the subgaleal shunt is inserted, the outflow tubing is temporarily occluded while most of the closing sutures are placed, to minimize overdrainage and collapse of the ventricles. The pocket fills with CSF after the wound is closed and remains full. The tubing in the subgaleal pocket is massaged a few times a day during the first few days to prevent early closure of the pocket.

A substantial number of patients who require a permanent shunt may also require a gastrostomy tube; inserting the distal shunt away from the left upper quadrant of the abdomen may minimize shunt difficulties caused by the gastrostomy tube later in childhood. When passing the shunt tubing from the abdomen to the cranium, care should be taken to avoid the nipple and any central lines in the upper chest and to avoid penetrating the skin in the neck. Once the plastic sheath has penetrated the subgaleal space at the skull base, it can often continue to be passed over the cranium in the subgaleal space without the metal stylet to minimize pressure on the skull and the need for passing incisions. A full length of distal shunt tubing can be placed to minimize the need to lengthen the tubing in the future.

Endoscopic Third Ventriculostomy

Endoscopic third ventriculostomy is a very effective treatment for noncommunicating hydrocephalus in children older than 2 years. Its use in young infants and in communicating hydrocephalus remains controversial, however. Several studies have shown that endoscopic third ventriculostomy is not effective in more than half of infants with nonobstructive hydrocephalus.³¹ In a few circumstances, despite the relatively low chance of success, it may still be the best option available.

Complications

Neurological Outcome and Comorbidities

Preterm infants with IVH who require a shunt are more likely to have neurological deficits than those with IVH without PHH, suggesting that the need for a shunt is an independent risk factor for poor outcome.⁶ A small subset of preterm infants with PHH does not demonstrate significant neurological deficits at follow-up, and it is difficult to predict outcomes for children when they are infants. The likelihood of neurological deficits should be explained to the parents and caregivers to help them understand that the goal of a shunt is to optimize neurodevelopment. Shunt insertion and revision do not cause the neurological deficits associated with prematurity, but PHH and the need for a shunt clearly represent a more extensive central nervous system injury during development. For example, about two thirds (59% to 66%) of former preterm infants with grade IV IVH surviving at 2 years have motor deficits, and almost half (44% to 50%) have cognitive delay.³²^–³⁴ A study from the Netherlands showed that 80% of children with grade IV IVH and a shunt had cerebral palsy.²¹ Children with shunts can be expected to have more deficits in all realms. Among 79 children who were born at less than 32 weeks’ EGA, weighed less than 1500 g, required shunt insertion for PHH, and were followed for an average of 10 years, only 29% performed at the expected grade level, ambulated without an assistive device, and were free of seizures (Robinson and colleagues, unpublished results).

An inherent part of using standardized techniques to assess outcomes in young children is that studies of preterm infants tend to focus on severe, well-defined deficits present at about 2 years of age. Although this allows an accurate comparison of various interventions, the more subtle deficits that can significantly impact a child’s achievement are not well documented. For example, 26% of former preterm (<1500 g) infants have symptoms strongly suggestive of autism spectrum disorders,³⁵ yet these disorders are not routinely assessed in outcome studies. A recent study compared the intellectual function of children aged 6 to 16 years with shunted PHH who performed within 1 year of expected grade level and that of normal controls; the investigators found that the full-scale, verbal, and performance IQ in children with shunts averaged a full standard deviation below that of controls.³⁶ This study likely overestimates the performance of former preterm children with shunts, because only half the patients in the study were born preterm, and those not performing near expected grade level were excluded.

Intraventricular Hemorrhage in Term Infants

Epidemiology

IVH in term infants is rare. Among the more than 2675 full-term infants admitted to a neonatal intensive care unit from 2003 to 2005, approximately 15% had peri- or intraventricular hemorrhage.³⁷ At another institution, 66 full-term infants with any type of intracranial hemorrhage were identified during a 12-year period, and only 30% had IVH.³⁸ In contrast to the minority of preterm infants who have low-grade IVH, 71.7% of term infants had a grade I lesion on cranial US, and 27.7% had grade II. Only two infants had grade III IVH, and one had grade IV.³⁷

Pathophysiology

Similar to preterm infants, IVH appears to arise from both genetic and environmental factors in term infants. Male gender, poor Apgar score, respiratory distress syndrome, and hyperbilirubinemia increase the risk of IVH.³⁷ Similarly, another study showed that forceps were used during delivery in 45% of infants with IVH, Apgar scores were less than 9 in 90% at 1 minute and in 60% at 5 minutes, and intubation after birth was required in 35%.³⁸ Term infants are more likely to develop IVH from venous infarction due to hypercoagulopathy or from hemorrhage of the choroid plexus. In many cases a cause cannot be identified. An increased frequency of the PLA2 (platelet glycoprotein IIb/IIIa Leu33Pro polymorphism) allele was found in term infants with IVH compared with controls.³⁹ The exact mechanism is likely to be poorly understood in most patients.