Neurodevelopmental outcome of children with spina bifida aperta repaired prenatally vs postnatally: systematic review and meta-analysis.

Abstract

To assess the neurodevelopmental outcome of children with spina bifida aperta (SBA) treated prenatally as compared to those treated postnatally. We performed a systematic review of the literature in PubMed/MEDLINE, EMBASE, Web of Science and The Cochrane Library, comparing the neurological outcome of infants with SBA treated prenatally vs postnatally. Only randomized controlled trials (RCTs) and non-randomized prospective controlled studies were included. The primary outcome assessed was neurodevelopmental impairment at the age of 1 year or later. Secondary outcomes were preterm birth, need for ventriculoperitoneal (VP) shunt by 12 months of age, absence of signs of hindbrain herniation at the first postnatal magnetic resonance imaging (MRI) evaluation and independent ambulation evaluated at 30 months. Of 11 359 studies identified through the electronic search, six met the inclusion criteria and were assessed in full text and two, one RCT and one prospective cohort study, were ultimately included in the meta-analysis. Sensitivity analysis did not show any difference between the outcomes of the RCT alone and those of the pooled RCT and prospective cohort study. This allowed neurodevelopmental assessment of 213 children between 14 and 53 months of age. Neurodevelopment was assessed by the Bayley Scales of Infant Development II (BSID-II) mental development index corrected for chronological age, with a cut-off of ≥ 70 (representing no more than 2 SD below the mean). The presence of neurodevelopmental impairment was similar between children who underwent prenatal (25/105 (23.8%)) and those who had postnatal (30/108 (27.8%)) repair of SBA (odds ratio (OR), 0.82 (95% CI, 0.43-1.56); P = 0.54), although the risk of prematurity was higher in the prenatal-repair group (OR, 17.62 (95% CI, 7.60-40.87); P < 0.0001). For every two fetuses operated on before birth, there was, compared with those operated on after birth, one additional premature birth (number needed to harm = 2 (95% CI, 1-3)). The need for VP shunt placement by 12 months of age was lower in the prenatal-repair group (45/109 (41.3%)) than in children that had postnatal repair (93/112 (83.0%); OR, 0.14 (95% CI, 0.08-0.26); P < 0.0001). Data on neurodevelopmental impairment in children with a shunt were available only for patients from the prenatal-surgery group of the RCT; in this subgroup, the likelihood for impairment was similar between children who did (7/39 (17.9%)) and those who did not (4/48 (8.3%)) have shunt placement (P = 0.21). At first postnatal MRI evaluation, no signs of hindbrain herniation were detected in 28/88 (31.8%) children who were operated on prenatally compared with 4/89 (4.5%) who had postnatal repair (OR, 9.45 (95% CI, 3.12-28.64); P < 0.0001). Independent ambulation at 30 months was achieved by 41/109 (37.6%) children who underwent prenatal repair compared with 21/111 (18.9%) who had postnatal repair (OR, 2.59 (95% CI, 1.39-4.86); P = 0.003). The risk of neurodevelopmental impairment in infants with SBA was similar between those who underwent prenatal and those who had postnatal surgical repair, despite an increased risk of prematurity in the prenatally repaired group. Copyright © 2018 ISUOG. Published by John Wiley & Sons Ltd.