Neurological risk factors in infants having noncardiac surgery and their relationship to growth and neurodevelopmental outcome

© 2017 Dr Margaret MoranINTRODUCTION: With advancements and improvements in surgical techniques, neonatal and antenatal care over the last sixty years, the mortality following neonatal non-cardiac surgery (NCS) has fallen to less than five percent. Consequently, the focus has now shifted to morbidity and long-term outcomes with increasing concern that these children are at risk of neurodevelopmental (ND) impairment. Other high risk neonatal groups, including those born prematurely and those who require early cardiac surgery, are known to have increased risk of ND impairment and frequently have abnormalities detected on brain imaging in the newborn period. The neurobiological substrates for ND impairment following NCS are unclear and likely to be multi-factorial but may include brain injury. The effect of neonatal NCS on brain structure, growth and development has not been described. AIMS: The aim of the study was to investigate neurological risk factors, in particular brain injury, in infants requiring neonatal NCS, and their impact on growth and ND outcomes over the first two years of life. METHODS: Infants undergoing major NCS in the first two weeks of life at The Royal Children’s Hospital, Melbourne for a diagnosis of oesophageal atresia (OA), congenital diaphragmatic hernia (CDH) or abdominal wall defect (AWD), were enrolled in this case-controlled prospective study. Infants were excluded if they were born before 32 weeks’ gestation or if they had a known genetic anomaly associated with ND impairment. Magnetic resonance imaging (MRI) of the brain was performed, time and clinical status allowing, pre-operatively and repeated prior to hospital discharge along with an assessment of neurobehaviour using Prechtl’s general movements. Surviving infants were followed up at six, 12 and 24 months when growth and neurodevelopment were assessed using the Alberta Infant Motor Scale (AIMS) and/or the Bayley Scales of Infant and Toddler Development, 3rd Edition (BSID-III). Thirty-nine infants were matched for gestation at birth, gestation at time of post-operative MRI and gender with controls. Control infants were recruited from a neighbouring perinatal hospital as part of another research trial. RESULTS: Over 27 months, 54 infants were recruited with slight male preponderance of 61%. The mean (standard deviation [SD]) gestation at birth was 37.9 (2.3) weeks with a mean birth weight of 2.93 (0.67) kilos. Twenty-one of the participants had AWD, 17 had OA and 16 had CDH. Surgery was performed at a median (interquartile range [IQR]) age of 24 (10 – 53) hours with a mean (SD) duration of 133 (56) minutes. Four infants died prior to 24 month follow-up (mortality rate 7%), with three of the four dying before hospital discharge. Infants requiring neonatal NCS had poor growth in the first six months, in particular prior to hospital discharge. There was evidence of catch-up growth for weight and height by 24 months. However, head growth remained poor, in particular for those children with CDH. Compared to control infants, there was poor head growth in the first month of life in infants following NCS. At 24 months of age, infants who had NCS were lighter and tended to be shorter than matched controls. Pre-operative MR imaging of the brain, obtained in a small number of participants (n=7), revealed delay in brain maturation in six of the seven participants. Post-operative imaging, obtained in 51 participants, also revealed high rates of delay in brain maturation (47%), with delay in cortical folding more common than delay in myelination. Abnormalities on MR imaging were detected in 51% of cases with white matter signal abnormalities (WMSA) and ventricular dilatation seen most frequently at 18% and 27% respectively. WMSA were associated with delay in brain maturation and with intracranial haemorrhage. Compared to matched controls, infants undergoing NCS had smaller biparietal diameters, smaller transcerebellar diameters and larger ventricles. Infants requiring NCS were more likely to have delay in cortical folding [odds ratio (OR) 5.92 (95% confidence interval (CI) 1.91, 19.47), p < 0.001] and WMSA [OR 9.81 (95%CI 1.18, 446.25) p = 0.01]. Following NCS, infants had significantly lower scores than test norms on AIMS assessment at six and 12 months (p = 0.003 and < 0.001 respectively). BSID-III assessment was performed at mean (SD) age of 12.4 (0.8) months in 48 survivors and at a mean (SD) age of 27.0 (2.4) months in 45 survivors. Language and motor composite scores were lower than test norms at both 12 and 24 months, with p < 0.001 and 0.04 for language at 12 and 24 months, and p < 0.001 and 0.01 for motor, at 12 and 24 months. Participants were performing in the low average range, with mean (SD) scores of 91.5 (15.2) and 94.2 (14.4) at 12 and 24 months for language and 90.4 (11.4) and 94.4 (13.6) for motor. Mild developmental impairment at 24 months was found in one third of the group with composite scores greater than 1 SD below the mean in at least one domain, with 9% having delay in all three domains. ND impairment at 12 months was predictive of ND impairment at 24 months. Compared to matched controls, infants requiring neonatal surgery had lower composite scores for cognition, language and motor at 24 months. Similar to previous reports, the control group performed significantly better across all domains than test norms. Perinatal and peri-operative risk factors associated with ND impairment at 24 months were compromise at birth with an Agpar score of less than seven at one minute, markers of illness severity, feeding dysfunction and failure to thrive, in particular poor weight gain in the first six months. Markers of illness severity including a rising base excess in the first 24 hours post-operatively, longer duration of mechanical ventilation and hospital stay. Increasing left lateral ventricular size was the only MRI parameter that remained significant for ND impairment on multivariate analysis. Gestational age at birth had little impact on ND outcome at 24 months, for infants requiring NCS born at 32 weeks’ or more of gestation. CONCLUSIONS: Infants requiring early NCS are at increased risk of ND impairment at 24 months. Post-operative MR brain imaging revealed high rates of brain immaturity and abnormalities, including white matter signal abnormalities and ventricular dilatation. A clear-cut link between early brain injury and maturational delay and subsequent ND impairment was not found. Increasing illness severity, feeding dysfunction and failure to thrive are associated with ND impairment at 24 months