Paediatric developmental venous anomalies (DVAs): how often do they bleed and where?

Abstract

Developmental venous anomalies (DVAs) are anomalies of venous drainage and considered a low-flow malformation. Studies evaluating natural history and risk factors for intracranial haemorrhage in the paediatric population are rare. We evaluate clinical and radiological features, risk factors and outcomes of paediatric DVAs. A retrospective study was conducted over a 10-year period between 2004 and 2014. Medical records, imaging and prospective databases were reviewed. Three-hundred-and-three radiological studies in total were evaluated. Fifty-two children (20 boys and 32 girls [median age: 6years] were identified with DVAs. Their age distribution was as follows: 1.9% neonates (< 1month), 11.5% infants (1month to 1year), 30.8% 1-5years, 30.8% 5-12years and 25% 12-16years. The majority (92.3%) presented with asymptomatic DVAs identified incidentally. Overall, anatomical distribution revealed predilection for frontal region (42.3%) with other common sites being posterior fossa (17.3%) and basal ganglia (13.5%). Temporal (11.5%), parietal (9.6%) and occipital (5.8%) were the remainder. Associated cavernous malformations (CMs) were present in 3/52 (5.8%), and no DVAs were associated with aneurysms or arteriovenous malformations (AVMs). Three patients had more than one DVA. There were three deaths unrelated to DVAs over median follow-up of 3.8years. Four patients (7.7%) suffered DVA-related intracranial haemorrhage presenting with neurological deficits. The ages of the children with DVA-related haemorrhages were 21days, 2years and 6months, 7years and 1month and 11years and 7months. Left-sided DVA haemorrhages predominated (3/4, 75%). The relative risk of a cerebellar DVA haemorrhage compared to its supratentorial counterpart was 5.35 (OR 6.8, 95% CI 0.8-58). DVA-related haemorrhage is sevenfold greater in our paediatric cohort compared to adults and is significantly associated with cerebellar location and cavernous malformations. There were no haemorrhages over a median period of 3.8years of prospective follow-up.