Timing of intervention in posthemorrhagic ventricular dilatation in preterm infants

Introduction and objectives: There is still significant variation in the management of post-hemorrhagic ventricular dilatation (PHVD). Recent evidence recommends cerebrospinal fluid (CSF) drainage as soon as Levene’s Ventricular Index (LVI) surpasses the 97th centile, which was shown to improve neurodevelopmental outcomes and reduce the need for a ventriculoperitoneal (VP) shunt. This study aimed to assess the timing of intervention in PHVD in a level 3 Neonatal Intensive Care Unit (NICU), its impact on the need for a VP shunt, and the presence of neurological sequelae at two years of age. Methods: A retrospective, single-center study was conducted, comprising preterm infants who developed PHVD. Ventricular dilatation was quantified in cerebral ultrasounds and details of interventions were obtained from clinical records. Infants were categorized into three groups depending on the neurological sequelae present at two years: surviving without sequelae, surviving with sequelae, and death. Results: Among the 22 infants diagnosed with PHVD, 59% required CSF drainage, and all the patients received initial intervention after LVI crossed the p 97 + 4 mm line (mean 7.5 mm above the 97th centile), at a mean postmenstrual age (PMA) of 30.6 weeks (± 2.7). Ventricular stabilization occurred after lumbar punctures (LPs) in 23% (3/13); 15% (2/13) died after temporizing neurosurgical procedures; 62% (8/13) required a VP shunt, at a median PMA of 38.9 weeks (IQR 37.0-41.3). Neurological sequelae (delayed motor development, cerebral palsy, epilepsy, and/or visual impairment) were less likely to occur in infants not requiring CSF drainage (p < 0.05), although no significant difference was found between ventricular width at first intervention and the need for a VP shunt or outcomes. Discussion: This cohort, treated before international guidelines were revised, received intervention later than what is now recommended. Infants who received intervention were more likely to have neurological sequelae than those that did not require intervention. Actively considering intervention as soon as the Ventricular Index (VI) surpasses the 97th centile will certainly make it possible to lower the intervention threshold to not more than p 97 + 4 mm, as currently recommended.