Intracranial haemorrhage in the neonate

Incidence

Intracranial haemorrhage (ICH) can affect newborns of all gestational ages and often is clinically 'silent'. Germinal matrix haemorrhage and intraventricular haemorrhage (GM-IVH) is most common in the premature population.

Estimates of frequency have changed over the last 20 years. Currently, large series report a 15% prevalence in infants <32 weeks.

Newborn Services, Starship Child Health's data for the period 2015-2017 indicates an incidence of 13.0% for Grade 1-2 IVH in infants<30 weeks. Grade 3 and 4 IVH and haemorrhagic infarction was seen in 6% of infants<30 weeks.

The incidence of periventricular leukomalacia (PVL) in infants <32 weeks over the same period of time was <1%.

Routine screening for IVH is performed in infants <30 weeks or <1250g at birth.

Associated risk factors

• SGA status of newborn

• Maternal pre-eclampsia

• No antenatal steroids

• Asphyxia

• Male gender

• Outborn infant

• Antepartum haemorrhage

• Base deficit >10

Timing of IVH

• Most haemorrhage occurs within the first three days of life.

• "Late" haemorrhage (i.e., after three days of age) may be associated with pneumothorax and its restriction of venous return to the heart.

Diagnosis

• IVH is reliably diagnosed with ultrasonography.

• Parenchymal injury (ischaemia, petechial haemorrhage, haemorrhagic infarct) can be diagnosed with ultrasonography, but with less sensitivity.

• Parenchymal ischaemia, without haemorrhage, may not be evident even with transducers of high frequency and very good equipment.

• MRI is more sensitive but currently impractical as a routine investigation in preterm infants.

• CT scanning is more sensitive than ultrasonography in cases where extra axial (subdural, subarachnoid) or posterior fossa haemorrhage is suspected.

Indications for cerebral ultrasound scan

• Infants born <30 weeks or <1250g at birth.

• Clinical suspicion of IVH

• Monitoring of hydrocephalus

• MCDA twins

• Following a significant clinical event such as NEC, major collapse, abnormal head growth or unexplained sharp drop in Hb.

• Be aware that ultrasound is not a sensitive method of detecting cortical or brainstem injury as seen with neonatal encephalopathy, or for detecting evidence of cerebrovascular events or collections outside the brain. In infants suspected of other pathologies, MRI or CT should be considered.

Timing of cerebral ultrasound scans

• The first scan should be performed at 4-5 days of age

• A second ultrasound scan should be done at day 28 (looking at resolution of previous IVH, evidence of parenchymal injury, and for evidence of periventricular leukomalacia .

• A "discharge" head ultrasound scan should be done at "term" (36 weeks is often chosen as the most appropriate time).

• If an abnormality is found discuss timing for serial scans with SMO

Grading System

GM-IVH

Intraparenchymal haemorrhage (haemorrhagic infarction) may be localised or extensive. The most common regions to be involved are parietal and posterior frontal (see figure below) . Extensive haemorrhagic infarction is defined as involving two or more regions.

(ref: Bassan H, Benson CB, Limperopoulos C et al. Ultrasonographic features and severity scoring of Periventricular Hemorrhagic infarction in relation to risk factors and outcome. Pediatrics 2006:117(6);2111-18)

Periventricular Leucomalacia

Post haemorrhagic hydrocephalus

• Dilatation of the ventricle due to CSF should be recorded as an additional finding. It is not the same as Grade III IVH.

• Neurosurgical opinion should be sought for infants with signs of increasing hydrocephalus:

  • HC is crossing centiles or HC enlarging >1.5cm per week

  • Presence of a tense anterior fontanelle

  • Separation of cranial sutures

  • Symptoms of increased intracranial pressure – apnoea, vomiting, abnormal posturing

Prognostic information

Prediction of abnormal neuromotor function by cranial ultrasound³

Normal scan refers to absence of haemorrhage within the brain parenchyma or ventricles, cysts or ventricular dilation. The grade of IVH (intraventricular haemorrhage) is given according to the Papile classification. PVL indicates periventricular leukomalacia. Ventricular dilation indicates moderate to severe ventricular dilation not meeting the criterion for hydrocephalus. Hydrocephalus indicates massive ventricular dilation >4mm above the 97th centile. Pre-test probability refers to the prevalence of cerebral palsy based on the Epipage study (larroqueref) The likelihood ratio is the probability that a patient with cerebral palsy has a positive test (abnormal ultrasound result). Post-test probability is the probability that a patient with a specific abnormality on cranial ultrasound will have abnormal neuromotor function.

Bassan et al found anterior frontal involvement in survivors was more commonly associated with abnormal neurological examination at follow up. ²

Cerebellar haemorrhage

The overall incidence of cerebellar haemorrhage is low however, the impact on neurodevelopmental outcome is high. In a systematic review of isolated cerebellar haemorrhage 43-75% were severely delayed in ≥1 domain of cognition, motor, language, and/or behavioural development. With Cerebellar Vermis involvement 87-93% had severe neurodevelopmental impairment.⁵

Unilateral Periventricular Haemorrhagic Infarction

A retrospective study of infants from 3 centres born between 1998 – 2004 at 18-36 month follow up found infants with unilateral periventricular haemorrhagic infarction (PVHI) have a higher median Mental Developmental Index (82 vs 49) and Psychomotor Developmental Index (53 vs 49) than infants with bilateral PVHI. Infants with unilateral PVHI were less likely to have severe cerebral palsy OR 0.15 [95%CI 0.05 – 0.45] compared to infants with bilateral PVHI. ⁶