Spinal Care, Neurology, Spinal Medical Advice and Information.

Cerebellar Hemorrhage


Cerebellar hemorrhage in the neonatal period is a rare disease. It has recently become recognized as a frequent, life threatening complication in premature infants requiring intensive care. Autopsy results of one series of patients showed cerebellar hemorrhage in 21% of neonates who were between 26 and 34 weeks gestation. In a similar study of patients at autopsy there was a reported occurrence of 25% of infants between 26 and 28 weeks gestation. The gravity of the lesion presumably is related, in large part, to its location in the posterior fossa. Posterior fossa hemorrhage may represent the only site of intracranial pathology and prompt surgical evacuation of the clot may result in a favorable prognosis.

Early identification of intracerebellar hemorrhage is clearly of major importance because the lesion can cause fatal brainstem compression and because it is reasonable to believe that intervention could be life saving. The clinical features usually remain nonspecific. The infant usually is the subject of a delivery which has been complicated by the use of forceps or breech extraction. Initially the infant is thought to be well but usually develops lethargy and irritability within the first few days of life. This is accompanied by respiratory irregularities, tense anterior fontanelle, and increasing head circumference. Investigation usually reveals a falling hemoglobin level and a blood stained CSF.

These findings seem to result from blood loss in the posterior fossa with forward displacement of the cerebellum and brainstem. Reports of this pathologic process have noted nystagmus, skew eye deviation, 6th nerve palsy, and 7th nerve palsy, as well as depression of primitive reflexes and hypotonia. The interval between presentation and diagnosis can vary from hours to weeks. It appears that while in the term infant the hemorrhage is secondary to birth trauma, in the premature baby the pathology may be different. In the premature infants with cerebellar hemorrhage, they share the common features that put all premature infants at risk for intracranial hemorrhage - prematurity, birth asphyxia, hypoxia, and hypercarbia.

Grunnet and Shields suggest that the cerebellar germinal plate, present in infants up to 30 weeks in gestation, may represent a site predisposed to hemorrhage. Thus, intracerebellar hemorrhage in the premature baby, like subependymal hemorrhage, may involve failure of autoregulation of cerebral blood flow in highly vascularized tissue. There is one report that implicates the use of straps for facemask ventilation as a possible cause of intracebellar hemorrhage. Regardless of the specific cause, however, clinical abnormalities may not become manifest until aqueductal or 4th ventricular obstruction or rupture occurs, allowing a latency of several days.

Although the diagnosis of intracerebellar bleeding may prove elusive on clinical grounds, early recognition remains essential for prompt intervention. In the past, detection depended upon suggestive ventriculography coupled with negative subdural taps. Lumbar puncture provides only nonspecific information and the procedure may be actually contraindicated in the case of cerebellar pathology because of the risk of tonsillar herniation. Computed tomography has successfully demonstrated cerebellar hemorrhage in premature infants.

It should be pointed out that the posterior fossa may be difficult to visualize on CT with axial views alone because a supratentorial hemorrhage may mask an infratentorial hematoma. Coronal views may be more helpful in this situation. More recently a number of reports have suggested that transfontanel real time ultrasonography may be a very effective means for identifying intracerebellar hemorrhage. Comparative studies of CT and ultrasound are necessary to determine if ultrasound is more effective or as effective as CT in the identification of neonatal intracerebellar hemorrhage. This is of significant importance because portable real time ultrasound may be much preferable to CT for diagnosis in the premature infant because the former modality obviates the need for transport, a dangerous procedure for the critically ill infant.

Two approaches to infant cerebellar hemorrhage have been used and reported in the literature. The first is a conservative approach in which the patient as long as it is stable, is observed without surgical intervention. Often by the time the hemorrhage is diagnosed there is no further extension of the bleeding. The second approach is a surgical approach in which the hematoma is evacuated in order to decompress the posterior fossa. This is shown to be very successful in a number of cases. Ultimately the approach should be determined by the progression of the disease and the size and extent of the bleed with resulting hydrocephalus.