Intraventricular haemorrhage is associated with developmental delay and cerebral palsy. The likelihood of cerebral palsy and poor neurodevelopmental outcomes increases with the grade of haemorrhage. But what if there’s only a bleed on one side?
Bottom Line
Intraventricular haemorrhage is bleeding into the ventricles.
It is more common in premature neonates.
An increasing grade of IVH is associated with poorer neurodevelopment outcomes.
Bilateral Grade 4 IVH is associated with a poor prognosis, high rate of Ccerebral Palsy and low MDI and PDI scores.
Both Mehrer and Maitre found that infants with unilateral grade 4 lesions had median cognitive scores that approached the mean for the general population and better overall neurodevelopment outcomes.
Two recent articles looked at neurodevelopmental outcomes for neonates with unilateral or bilateral intraventricular haemorrhage.
Both papers looked at outcomes with respect to the laterality of the haemorrhage, for either all grades or grade 4 IVH. Firstly though, a brief background…
Intraventricular Haemorrhage (IVH) is bleeding in the ventricles, usually as a result of germinal matrix fragility from immature vessels under stress. It is much more common in premature neonates: ~35% at 25 weeks of gestation; ~18% at 29 weeks; and <10% at 31-32 weeks. In term neonates, severe IVH is rare. IVH is associated with poor neurodevelopmental outcomes.
| USS Finding | Outcome |
|---|---|
| Grade 4 + cysts | 80% have moderate-severe disability, 90+%have CP |
| Grade 4 | ~50-70% have CP / intellectual deficit / IQ<70 (30-60% will require VP shunt) |
| Grade 3 | 20-35% have CP / intellectual deficit / IQ<70 |
| Grade 2 | 10-15% have severe disability |
| Grade 1 | 4-5% have severe disability |
| Normal | 5-8% have significant motor defect, IQ <70, moderate-severe disability |
The first paper, by Maitre and colleagues, is Neurodevelopmental Outcome of Infants With Unilateral or Bilateral Periventricular Hemorrhagic Infarction (PVHI).
Who were the patients?
Patients were 69 infants with a birthweight <1500g. They had confirmed PVHI on early cranial USS and were born between 1998 and 2004.
Children were excluded for congenital viral infections, major congenital heart disease, genetic abnormalities, structural brain malformations, metabolic diseases and admission only for shunt insertion.
PVHI is defined as IVH plus haemorrhagic necrosis of the periventricular white matter, which can occur with or without periventricular leukomalacia (PVL).
It’s the same as a Grade 4 IVH as per our earlier definition.
What did they look at?
A comparison between those with unilateral and bilateral PVHI was undertaken using Bayley Mental Developmental Index (MDI) scores, examining motor and cognitive outcomes. The authors used binary outcome measures of MDI <70 or Psychomotor Developmental Index (PDI) <70. These were equivalent to 2 SD below the population mean and corresponded to significantly delayed development.
Were the groups similar?
Groups were relatively similar with the exception of a significantly lower gestational age for the bilateral PVHI group: unilateral (25.9 weeks) vs bilateral (24.6 weeks) (p=0.005).
What were the outcomes?
In models not adjusted for gestational age, outcomes appeared worse with bilateral PVHI. After adjustment, the odds ratios were not substantially different.
Notably, for infants with bilateral PVHI, diplegic and quadriplegic cerebral palsy (CP) predominated. In infants with unilateral PVHI, CP was mostly hemiplegic.
A further subgroup analysis examined the unilateral PVHI group to compare unilateral PVHI with and without PVL. The authors identified 2 distinct groups of infants with unilateral PVHI: one with cognitive indices approaching average scores and the other with profound delays.
Patients without PVL appeared to have better outcomes; of those with unilateral PVHI, 78% of infants with PVL had moderate to severe cerebral palsy versus 18% in the group without PVL. Those with PVL also had significantly higher rates of vision impairment and seizure disorders.
Infants with unilateral lesions had median cognitive scores that approached the mean for the general population.
Two-thirds of infants with unilateral PVHI have CP, but in a majority of these infants, CP is mild. Conversely, almost all infants with bilateral PVHI have very poor cognitive and motor outcomes.
This study was limited by the differences between gestation of unilateral & bilateral groups, and the retrospective nature. Despite the actual small numbers, this was one of the largest studies looking at this question.
The second paper is Mehrer‘s 2012 paper.
Who were the patients?
Inclusion criteria were at least one abnormal cranial USS – including any grade IVH, ventricular dilatation, parenchymal cysts, cystic PVL or porencephalic cysts. The most severe cranial USS finding was used to categorise the infants. The authors did, however, exclude all periventricular leukomalacia in addition to lethal congenital abnormalities, chromosomal abnormalities and a history of meningitis
What did they look at?
Their study aimed to determine whether the laterality of IVH (unilateral versus bilateral) was a predictor of neurodevelopmental outcome at 18–22 months in a cohort of ~166 ELBW (<1000g) infants with Grades I– IV IVH.
This study is assessing the effect of laterality on outcomes for all Grades of IVH
What were the outcomes?
The outcome measures were similar to Maitre’s study: cerebral palsy, PDI & MDI <70; however, blindness and hearing impairment were added.
Unlike in the Maitre study, regardless of statistical significance, laterality (unilateral vs bilateral IVH) remained in the models because it was directly related to the study hypothesis. This study had a bias towards lower grades of IVH; sample sizes decreased with increasing IVH grades.
For Grades I-II, outcomes are unchanged between unilateral and bilateral IVH. There was a difference in MDI & PDI scores between unilateral and bilateral Grade 4 IVH when controlled for a number of factors including sepsis, and maternal steroids, among others
This quote from Mehrer’s discussion nicely summarises their findings:
“[Maitre’s] findings are consistent with our study and suggest that the brain can adapt when there has been significant destruction of brain tissue on one side, but severe impairment usually results when there is parenchymal injury on both sides of the brain.”
The model is a multivariable logistic regression, and there are clearly a number of confounders to consider. The raw numbers of unilateral vs bilateral IVH at each grade are not stated – these could be quite small with respect to the conclusions drawn. In particular, I’d be interested to see the prevalence of a unilateral Grade 4 IVH without any contralateral haemorrhage, and whether the mechanisms are different from those of the remainder of the population with IVH.
References
Davies, Cartwright & Inglis. “Pocket notes on Neonatology 2E.” 2008. Elsevier. (3rd Ed available as iPhone application)
Uptodate – “Clinical manifestations and diagnosis of intraventricular hemorrhage in the newborn”
