Intracranial Hemorrhage and Vascular Lesions in the Neonate


Germinal Matrix Hemorrhage–Intraventricular Hemorrhage

Incidence

Germinal matrix hemorrhage–intraventricular hemorrhage (GMH-IVH) mainly occurs in premature infants, and the risk is higher with decreasing maturity. A 2003 study, however, also showed that IVH, sometimes associated with a thalamic hemorrhage, can be seen in full-term infants, and this can be associated with a sinovenous thrombosis. The first studies using computed tomography (CT) and ultrasonography were performed between 1978 and 1983 and showed an incidence of 40%-50% in infants with birth weights less than 1500 g. In the 1990s and more recently, several groups noted a decline in the incidence to approximately 20% of infants with very low birth weight, but this decline has not been confirmed by others, and the decrease in severe GMH-IVH shown in a cohort of the Vermont Oxford Network was of borderline statistical significance. In a recent study from the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network in 34,636 infants with a gestational age between 22 and 28 weeks, a decline in the incidence of severe IVH was seen from 19%-15% with the following trends between 1993 and 2012: A significant reduction was seen for those with a gestational age of 26-28 weeks (adjusted relative risk [RR] for the change per year, 26 weeks 0.987) (95% confidence interval [CI] 0.974-0.999), 27 weeks 0.964 (95% CI 0.971-0.997), and 28 weeks 0.973 (95% CI 0.958-0.989).

Since the early 1980s, the incidence of intraparenchymal hemorrhage has also shown a decline, although this decline was not found in all studies. The average incidence for intraparenchymal hemorrhage is now 5%-11%. An even lower incidence of 3% was reported in a French population-based cohort. The decrease in the incidence of GMH-IVH is mainly attributed to the increased use of antenatal corticosteroids and postnatal use of surfactant.

Timing of the Hemorrhage

Accurate timing of GMH-IVH is possible only when sequential ultrasonography is performed. A diagnosis of a hemorrhage of antenatal onset can be made only when the first ultrasonography is performed on admission, within a few hours after delivery. Many studies have shown that almost all hemorrhages develop within the first week after birth, and many of them within the first 48 hours after birth. Progression of a GMH-IVH over 1-2 days is not uncommon, and this applies especially to an IVH progressing to an intraparenchymal hemorrhage. Impaired venous drainage of the medullary veins in white matter leading to venous infarction is the most likely underlying mechanism for this progression (see section on Neuropathology ). Only approximately 10% of cases of GMH-IVH occur beyond the end of the first week, in contrast to periventricular leukomalacia, wherein late onset is not uncommon.

Neuropathology

Pathologists have noted GMH-IVH developing after hemorrhage in the subependymal germinal matrix, a structure that is most prominent between 24 and 34 weeks of gestation and that has almost completely regressed by term. Germinal matrix tissue is abundant over the head and body of the caudate nucleus but can also be found in the periventricular zone. More recently, magnetic resonance imaging (MRI) has confirmed just how extensive this tissue is in preterm infants. The germinal matrix contains neuroblasts and glioblasts that undergo mitotic activity before migrating to other parts of the cerebrum. Bleeding into the caudothalamic part of the germinal matrix is predominant, but using a new MRI sequence, susceptibility weighted imaging (SWI), it has been noted that many cases also have bleeding into the temporal or occipital germinal matrix outer zones. Using this MRI sequence, it has also been observed that preterm infants with GMH-IVH have a higher variability in anatomy of subependymal veins which may be a predisposing factor for GMH-IVH. The germinal matrix receives its blood supply from a branch of the anterior cerebral artery known as the Heubner artery . The rest of the blood supply is derived from the anterior choroidal artery and the terminal branches of the lateral striate arteries. Vessels in the germinal matrix are primitive and cannot be classified as arterioles, venules, or capillaries and are often referred to as the immature vascular rete . Venous drainage of the deep white matter occurs through a fan-shaped leash of short and long medullary veins through which blood flows into the germinal matrix and subsequently into the terminal vein, which is positioned below the germinal matrix. The anatomic distribution of parenchymal lesions associated with GMH-IVH suggests venous infarction resulting from obstruction of this vein.

Germinal Matrix Hemorrhage

Most GMHs arise in the region of the caudate nucleus. Although GMHs at the level of the caudate nucleus are identified with ultrasonography, MRI using the SWI sequence has shown that GMH also occurs in the germinal matrix in the roof of the temporal horn. The size of the GMH changes with the maturity of the infant: The less mature the infant, the larger the GMH. The site also varies with maturity, with occurrence over the body of the caudate nucleus in the less mature infant and over the head of the caudate nucleus in the more mature infant. A GMH can result in suppression of cell proliferation of the human ganglionic eminence. When the GMH is followed longitudinally with ultrasonography, a subependymal cyst is seen as a sequel after several weeks, and this cyst is often still present at term-equivalent age and is better visualized with cranial ultrasonography than with MRI.

Intraventricular Hemorrhage

Hemorrhages occurring in the germinal matrix often rupture through the ependyma into the lateral ventricle and are then referred to as an IVH . These hemorrhages can vary considerably in size and, if large, can lead to acute distension of the lateral ventricle. Large clots can be present, seen as casts at postmortem examination or with cranial ultrasonography, and these can sometimes change position with repositioning of the infant. The blood can fill part of or the entire ventricular system, spreading through the foramen of Monro, the third ventricle, the aqueduct of Sylvius, the fourth ventricle, and the foramina of Luschka and Magendie to collect eventually around the brainstem in the posterior fossa. Clot formation can lead to outflow obstruction at any level, but most commonly at the level of the aqueduct of Sylvius or more diffusely at the level of the arachnoid villi. More gradual progressive ventricular dilation occurs especially in those with a large GMH-IVH; this is known as posthemorrhagic ventricular dilation. This can be transient or persistent. In a small group, it can be rapidly progressive and noncommunicating, usually because of obstruction at the level of the aqueduct of Sylvius or the outlet foramina of Luschka and Magendie. Unilateral outflow obstruction at the level of the foramen of Monro can lead to unilateral hydrocephalus ( Fig. 53.1 ). In cases of communicating progressive posthemorrhagic ventricular dilation, the increase in ventricular size is more gradual and is considered to be caused by an obliterative arachnoiditis resulting from blood collecting in the subarachnoid spaces of the posterior fossa, leading to an imbalance between cerebrospinal fluid (CSF) production and reabsorption.

Fig. 53.1, A, Cranial ultrasonography, coronal view, shows posthemorrhagic ventricular dilation after a large left-sided hemorrhage. B, Note isolated enlargement of the left occipital horn on the parasagittal view. C, The magnetic resonance image, T2-weighted, spin-echo sequence still shows evidence of enlargement of the left occipital horn and enlargement of the frontal extracerebral space. Myelination of the posterior limb of the internal capsule (PLIC) is symmetric and seen as low signal intensity.

Parenchymal Hemorrhage

The most severe type of hemorrhage involves the parenchyma. This type of lesion occurs in approximately 3%-15% of all hemorrhages. A study by Dudink et al. described the different veins involved in this type of lesion. Direct extension into the parenchyma from pressure of blood in the ventricle is now considered unlikely. Some still take the view that all parenchymal hemorrhages are originally ischemic in origin, with any bleeding being a secondary complication. However, most agree that a unilateral parenchymal lesion accompanying GMH-IVH is most often caused by the presence of the GMH leading to impaired venous drainage and venous infarction. Gould et al. showed that the ependyma remained intact, indicating that there had not been an extension of the preceding IVH. These lesions almost invariably show a moderate to large IVH on the ipsilateral side. The parenchymal hemorrhage is mostly unilateral but can occur in both hemispheres and is associated with a worse neurodevelopmental outcome. A score was introduced, taking into account whether the lesion is unilateral or bilateral, the presence of a midline shift, and the number of regions that are involved. The appearance of the intraparenchymal hemorrhage has changed since the early 1990s. Instead of a unilateral globular hemorrhagic lesion in continuity with the lateral ventricle and evolving into a single porencephalic cyst, a smaller, triangular parenchymal lesion can be seen with the tip of the triangle toward the lateral ventricle. There may be partial or even no communication of the parenchymal hemorrhage with the lateral ventricle, and an evolution into a few cystic lesions, separate from or partially in communication with the lateral ventricle, can be seen after several weeks ( Figs. 53.2 and 53.3 ). This type of lesion is sometimes classified as unilateral periventricular leukomalacia (PVL), but in view of the later MRI appearance, with very focal gliosis instead of diffuse gliosis, this appears to be incorrect. With the use of sequential ultrasonography, the underlying problem can be better understood because unilateral cystic lesions occurring after an ipsilateral IVH are more likely to be caused by a venous infarction rather than cystic PVL. It is possible that improvement in neonatal care is associated with the change in appearance of the unilateral parenchymal hemorrhage seen over time.

Fig. 53.2, A, Cranial ultrasonography, coronal view, shows intraventricular hemorrhage and parenchymal hemorrhage not in communication with the lateral ventricle. B, Three weeks later, a single cystic lesion is present after resolution of the hemorrhage. C, At term, the cyst is no longer seen.

Fig. 53.3, A and B, Cranial ultrasonography, coronal view, shows intraventricular hemorrhage and parenchymal hemorrhage not in communication with the lateral ventricle. C, At term age, multiple cysts are seen adjacent to the lateral ventricle.

Intracerebellar Hemorrhage

Cerebellar hemorrhage was always considered uncommon, but the increased use of routine MRI in preterm infants has shown that this condition is more common than previously thought. This problem is especially common in very immature infants. The condition has been reported in 2.5% of high-risk preterm infants, but with the increased use of neonatal MRI, the diagnosis is more often made, and the reported incidence is now reported to be between 2% and 19%. The high incidence includes all punctate lesions, which are not seen with ultrasonography, which is unable to recognize lesions less than 4 mm, but only with MRI. Ultrasonography performed through the mastoid window is more successful in making the diagnosis, but is still inferior to MRI ( Fig. 53.4 ). Cerebellar hypoplasia without an apparent cerebellar hemorrhage has also been reported as a common sequel of severe immaturity. Reduced cerebellar volumes were only shown on three-dimensional MRI in preterm infants at term-equivalent age in association with supratentorial pathology, such as hemorrhagic parenchymal infarction, intraventricular hemorrhage with dilation, and periventricular white matter damage. More recent studies have confirmed significantly slower cerebellar growth at term-equivalent age in the presence of severe supratentorial IVH, but no reduced cerebellar growth in the presence of associated white matter injury.

Fig. 53.4, A, Cranial ultrasonography, axial scan through the temporal bone, shows a large right cerebellar hemorrhage. B, The magnetic resonance imaging (MRI) scan, T2-weighted, spin-echo sequence still shows evidence of blood (low signal intensity) at term equivalent age, as well as atrophy of the right cerebellar hemisphere.

Pathogenesis

The precise nature and origin of GMH-IVH remain uncertain. Although some groups have suggested that the capillaries in the germinal matrix do not rupture easily, others have suggested that they can because the vessels are immature in structure with little evidence of basement membrane protein and they are of relatively large diameter. Early neuropathologists considered subependymal bleeding to be entirely venous in origin. This theory was rebutted by Wigglesworth and Pape, who suggested, on the basis of their injection studies, that capillary bleeding was more prominent than terminal vein rupture. There has been a return to the concept that most parenchymal hemorrhages are caused by venous infarction or by reperfusion injury after an ischemic insult. An anatomic analysis of the developing cerebral vasculature did not show precapillary arteriole-to-venous shunts.

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here