Post hemorrhagic hydrocephalus and neurodevelopmental outcomes in a context of neonatal intraventricular hemorrhage: an institutional experience in 122 preterm children

Intraventricular hemorrhage (IVH) remains a serious complication in premature children, affecting approximately 20–30% of infants born < 29 weeks estimated gestational age (EGA) [1‐3]. In a few cases, IVH can occur in fetus during pregnancy or in children born at term. Improvements in obstetric care have led to an increase in survival and a decrease in the incidence of IVH in preterm newborns [4] secondary to the antenatal administration of corticosteroid and/or sulfate magnesium. Nevertheless, a correlation has been established between low gestational age at birth and the incidence and severity of IVH [5].

In preterm newborns, the physiopathology [6‐8] of bleeding is based on hemorrhagic transformation of hypoxia-ischemia in the vulnerable subependymal germinal matrix. This location is fed by rich terminal vascularization with an intense metabolism, immature at this step of brain development and highly sensitive to hemodynamic fluctuations. The invasion of bleeding in the ventricular system is responsible for post-hemorrhagic hydrocephalus (PHH) [9] due to the obstruction of cerebrospinal fluid (CSF) circulation and to the inflammatory response of the ependyma causing a loss of compliance and finally a decrease of CSF reabsorption. Moreover, white matter lesions due to intraparenchymal hemorrhage are responsible for alteration of oligodendrocytes and astrocytes, affecting the myelination and organization of the cerebral cortex.

Despite many treatment options, there is still no consensus on the management of PHH and very few data about neurodevelopmental outcomes and predictive factors of PHH [3, 10, 11] . The indication and the timing of surgical treatment [12, 13] remain challenging for the neurosurgeon and the neonatologist, as does the impact of IVH on the neurodevelopmental evolution of the child. The objective of this study was to evaluate the predictive factors of evolution to PHH in 122 newborns with neonatal IVH, and report their neurodevelopmental outcomes at 2 years.

In this study, based on the long-term outcomes of 122 newborns with neonatal IVH, we report a PHH rate of 18%. Our result is concordant with data in the literature [2, 17, 18] in which the PHH rate varies between 20 and 35%. Risk factors for PHH were high IVH grade and increased HC at diagnosis.

In a recent study [19] based on the outcomes of 97 infants with neonatal IVH, the PHH rate was of 35%, and the first significant risk factor for PHH was the grade of the initial bleeding. In this series, all infants with a permanent VPS had an initial bleeding grade of III or IV on the Papile classification. In our series, 6 of the 22 infants requiring VP shunt had an initial hemorrhage grade of 2 while most studies limited their inclusion criteria to grade 3 and 4. In another study [2] based on 42 infants with IVH and a PHH rate of 26%, the risk factors for onset of PHH were high IVH grade, late onset (later than 1 week after birth) of bleeding and < 30 WG. The absence of a direct relationship between gestational age at birth and PHH could be due to confounding factors and a higher mortality rate in extreme preterm births. We observed that a HC > + 2 SD at diagnosis was a risk factor for shunt dependence. This observation emphasizes the need for early management of PHH before the onset of ependyma lesions leading to a loss of compliance of the ventricles [13].

The type of CSF derivation device was not a discriminant risk factor for shunt dependence in our cohort. According to current data in the literature, two devices are recommended [12]: the ventriculo subgaleal shunt and the ventricular access device. The use of CSF washing was the subject of an important publication in the year 2003 [20]. The outcomes of this technique were discordant: a higher incidence of secondary bleeding but better neurodevelopmental outcomes at 2-year follow-up [21, 22]. According to a recent meta-analysis [12], there is not a sufficient level of evidence to recommend this strategy. Studies have been conducted to find an alternative to these strategies with the use for example, of iron chelator on animal models [23], to decrease inflammatory response and prevent the onset of hydrocephalus. These strategies could be applied to patients at risk of developing PHH. CSF biomarkers could be of interest to predict the onset of PHH in these young patients. For example, in a recent study, Morales et al. [24], demonstrated a strong association between the CSF level of amyloid precursor protein (APP) and ventricular size.

Concerning mortality, we report a rate of 36.9% defined as the rate of mortality during the 2 years of follow-up. In our study, risk factors for mortality were low gestational age at birth and high IVH grade. This rate is concordant with data in the literature [25]. Death was due to extra neurological causes in more than 50% of cases because of the onset of other complications inherent to prematurity (nosocomial infections, enterocolitis...) of children with a PHH.

Concerning motor outcomes at 2 years, 43 patients had a GMFCS ≥2. Risk factors for negative evolution were high IVH grade on ultrasound and increased cranial circumference at the time of hydrocephalus management. In a serie of 95 patients, De Vries et al. [13] reported motor impairment in 22% of patients with a PHH. In another study [11] based on 6000 patients, of the 40% who reached 2-year survival, 14% presented cerebral palsy. The prognosis was worse in patients with permanent VP shunt. In a previous study with 400 patients [26], the rate of motor impairment was 23%. As in our study, all these retrospective studies observed that the rate of cerebral palsy was elevated if we compared them to the rate of cerebral palsy in the cohorts of preterm infants regardless of the presence or absence of IVH [27]. However it was mentionned in several studies that the higher the grading of IVH, the higher the risk of cerebral palsy. This observation may help to explain the reduced cerebral volume and impaired developmental outcomes in patients with IVH.

In our cohort, 40 infants (51.9%) had an association of words at the age of 24 months. The impact of prematurity and IVH on school performance could not be evaluated in our study. A Dutch series [26] evaluated the neurodevelopmental outcomes of 484 preterm children born before 32 WG. In this cohort, at the age of 2 years, forty-five (15.3%) of the 294 survivors had a minor and 23 (7.8%) a major handicap. The presence of an IVH was associated with impaired neurodevelopmental outcome. The evolution of the same cohort was evaluated at the age of 14 years [28], school performance data were obtained for 278 of the 304 surviving adolescents. In this study, 129 adolescents (46.4%) performed normally, 107 (38.5%) were slow learners and 42 (15.1%) needed special education services. The presence of a perinatal IVH was the only factor, which was significantly asssociated with the need for special education. There was a fourfold risk of special education comparing patients with grade III/IV and patients without IVH. We report a sensorial deficit in 18 infants (23%) in our cohort. The presence of sensorial deficit is of interest and must be diagnosed early because it contributes to poor school performance.

Our study has some limits as it is a retrospective study collecting a high number of preterm infants born during a long period of 14 years during which the standards of care of preterm infants have changed. In our study, there was no difference in the rate of antenatal administration of corticosteroid or magnesium sulfate between groups of children with IVH with or without PHH. Both molecules have been associated with a lower rate of IVH. We can only observe that the rate of antenatal corticosteroid administration was low (52.7%) as well as the rate of antenatal magnesium sulfate (13.1%).

We conducted a study on 122 patients with a neonatal IVH. Among the 77 surviving patients at 2 years, 22 (18%) required a permanent VP shunt. Clinical evolution was favorable in 38 of the 77 survivors (49.4%). The risk factors for shunt dependence and impaired neurodevelopment were IVH grade and increased head circumference. We emphasize the need for close follow-up of these infants and early surgery in case of hydrocephalus. Among surviving patients, close attention must be given to neurodevelopment because of the risk of long-term consequences associated with this pathology. The development of biomarkers and medical therapeutic strategies may help to predict PHH and reduce its consequences.