Fetal intracranial hemorrhage refers to bleeding that occurs antenatally from a blood vessel into the ventricles, subdural space, or parenchyma of the brain. Whereas neonatal hemorrhage is a relatively common occurrence, affecting 40% to 60% of infants delivered before 32 weeks of gestation, fetal intracranial hemorrhage is quite rare. Factors that may place the fetus at risk for intracranial hemorrhage include alterations in maternal blood pressure, a maternal seizure disorder, placental abruption, specific medication or substance exposure (such as warfarin or cocaine), severe abdominal trauma, hereditary coagulation disorders (Komlósi et al., 2005) or alloimmune platelet disorders (Kuhn et al., 1992; Sherer et al., 1998; Lynch et al., 2002). Coagulation disorders associated with fetal intracranial hemorrhage incude factor V Leiden (Komlósi et al., 2005), Factor X mutations (Herrmann et al., 2005), prothrombin G20210A mutation, protein C or S deficiency, antithrombin III deficiency, and antiphospholipid antibody syndrome (Lynch et al., 2002).

Three types of intracranial hemorrhages can occur: intraventricular (or periventricular), intraparenchymal, and subdural. Intraventricular or periventricular hemorrhages are the most common, emanating from small vessels within the subependymal germinal matrix before 33 weeks of gestation (McGahan et al., 1984). The pathogenesis of intraventricular hemorrhage (IVH) is related to fragility of the capillary bed of germinal matrix, a disproportionate amount of total cerebral blood flow to the periventricular area, and the lack of autoregulation of cerebral blood flow in the fetus or premature infant. IVH is the most common form of intracranial hemorrhage seen in the neonate and is categorized in severity on a four-point scale, with the most severe fourth grade including parenchymal involvement (Lynch et al., 2002).

Intraparenchymal hemorrhages may be identified as distinct echogenic areas within the cerebral tissue, with or without displacement of the underlying ventricular or outer surfaces of the brain. As these hemorrhages evolve, they become hypoechoic and flattened as the hematoma liquefies. Residual changes may include development of a porencephalic cyst (see Chapter 21) or ventricular enlargement. Clots in the ventricular system are seen as bright echogenic areas that are similar to choroid plexus.

Subdural hematoma generally presents as fetal macrocephaly, with separation of the skull from the cerebral cortex. Hyperechoic and hypoechoic areas are identified, filling the space between the brain gyri and the skull. The presence of cerebral edema may cause acoustic enhancement of brain gyri. Catanzarite et al. (1995) have reported on the diagnosis of fetal subdural hemorrhage. They recommended imaging the fetal head in the axial or coronal plane at one level of the sylvian fissure. Normally the separation between the cortex and the inner table of the skull is less than 4 mm. Subdural bleeding can be seen as a collection of echodense material outlining the cortex and separating the sylvian fissure from the inner table of the skull. An additional report on fetal subdural hematoma came from Rottmensch et al. (1991), who identified enhanced echogenicity of the sulci as an indication of brain edema. These authors stated that brain function cannot be predicted sonographically except in the most severe cases.

In a retrospective study of pregnant women in whom a diagnosis of fetal intracranial hemorrhage was made, Achiron et al. (1993) identified five fetuses between 26 and 36 weeks of gestation at the time of diagnosis. Hyperechoic lesions were shown in the brain parenchyma and lateral ventricles in three of five fetuses. Transvaginal sonography permitted the enhanced visualization of ventriculomegaly in one fetus and periventricular leukomalacia in the second fetus. Of the five fetuses studied, two growth-restricted fetuses died after birth, and of the three survivors, two had normal long-term development. One infant developed hydrocephalus, suffered from neurodevelopmental retardation and eventually died at age 7 months (Achiron et al., 1993). In another series, Vergani et al. (1996) described their experience with six cases diagnosed at the University of Milan and 35 cases identified from the English literature. Of the 41 cases of fetal intracranial hemorrhage reviewed, 20 had isolated IVH, 13 had parenchymal hemorrhage, and eight had subdural or subarachnoid hemorrhage. These investigators devised a prognostic scoring system for IVH. A favorable outcome was seen in 50% (6 of 12) fetuses observed with Grade II IVH, but none of the fetuses with the equivalent of a Grade III IVH (blood filling the ventricles) had a favorable outcome.

In one study performed at a single institution, the University of Milan, six cases of antenatal intracranial hemorrhage were detected in 6651 fetuses studied. This suggests an incidence of almost one per 1000 fetuses (Vergani et al., 1996). An unusually high incidence of fetal subdural intracranial hemorrhage has been documented in Pacific Islanders who emigrate to New Zealand (Becroft and Gunn, 1989). The cause for this high incidence is unknown, but it suspected to be due to traditional forms of abdominal massage used to change fetal position in utero.

In one study, Sims et al. (1985) reviewed the clinical features and neuropathology of intracranial hemorrhage in 433 consecutive stillbirth autopsies. The authors identified 25 cases of periventricular or intraventricular hemorrhage or gliosis. Of the 25 cases, 10 had IVH alone, 5 had IVH plus an additional intraparenchymal hemorrhage, 5 had parenchymal hemorrhage only, and 5 had gliosis, indicating a remote neurologic injury. Other studies have shown that IVH is present in 6% of stillbirths at autopsy (Minkoff et al., 1985).

Immune thrombocytopenic purpura is the most common autoimmune disorder affecting pregnant women, with an incidence of 1 in 1000 to 1 in 10,000 pregnancies. Fewer than 1% of these women will have babies in whom an intracranial hemorrhage develops. In contrast, severe fetal thrombocytopenia commonly occurs with alloimmune thrombocytopenia, which develops as a result of maternal sensitization to fetal platelet antigens. This occurs in 1 in 2000 to 1 in 5000 pregnancies. Of these cases, approximately 10% to 30% are at risk for intracranial hemorrhage (Johnson et al., 1997; Bussel et al., 1988). This makes alloimmune thrombocytopenia the most common cause of severe thrombocytopenia in fetuses, and the most frequent cause of intracranial hemorrhage in term neonates (Bussel et al., 2005; Berkowitz et al., 2007).

Sonographic diagnosis of fetal intracranial hemorrhage was first reported in 1982 in a woman with recurrent episodes of pancreatitis and a fetal death at 29 weeks of gestation (Kim and Elyaderani, 1982). Subsequently, the majority of reported antenatally detected cases of intracranial hemorrhage have occurred during the late second and third trimesters (Catanzarite et al., 1995; Elchalal et al., 2005).

Intracranial hemorrhage is identified as an echogenic area within the ventricles or brain parenchyma. Blood clots are usually echogenic (Figures 17-1 and 17-2) and evolve to characteristic echolucent cystic areas. The sonographic criteria for an intracranial hemorrhage include the presence of an echogenic mass, unilateral or bilateral ventriculomegaly with the presence of intraventricular echogenic foci, germinal matrix, echogenicity, or the presence of periventricular hypoechoic lesions (cysts). In approximately 50% of cases of IVH, sonographic findings will be bilateral (Elchalal et al., 2005).

The commonly used IVH classification system in neonatal imaging has been applied to prenatal sonographic findings, with some modifications (Elchalal et al., 2005). Grade I IVH refers to hemorrhage limited to the subependymal matrix; Grade II refers to no more than 50% filling of the ventricles and lateral ventricular enlargement of no more than 15 mm; Grade III involves more than 50% ventricular filling together with significant ventriculomegaly; Grade IV refers to intraventricular bleeding that also involves the periventricular parenchyma.

The major considerations in the differential diagnosis include identifying other potential causes of intraparenchymal or intraventricular masses, as well as identifying potential causes for hemorrhage. It is important to rule out the presence of a prominent choroid plexus. This may present as a densely echogenic mass within the ventricles, mimicking blood clot. The choroid plexus can occupy a large proportion of ventricular space, but this is a normal finding with an expected normal prognosis (Kim and Elyaderani, 1982). Other causes of an echogenic midbrain mass include intracranial neoplasm or infection. A case of malignant glioblastoma multiforme presenting with the appearance of significant intracranial hemorrhage in a 32-week fetus has been reported (Sell et al., 2006).

Considerations in the differential diagnosis of the cause of hemorrhage include coagulation disorders, such as factor V Leiden (Komlósi et al., 2005), Factor X gene mutations (Herrmann et al., 2005), prothrombin G20210A mutation, protein C or S deficiency, antithrombin III deficiency, and antiphospholipid antibody syndrome (Lynch et al., 2002). Fetal thrombocytopenia due to maternal immune thrombocytopenia or maternal platelet antigen incompatibility should also be considered, with the latter being a much more likely cause of significant fetal thrombocytopenia and hemorrhage (Bussel et al., 2005; Berkowitz et al., 2007). Maternal TORCH infection and maternal medication or substance use are also causes of fetal intracranial hemorrhage.