A Comparison of MRI Scoring Systems Predictive Ability for Two-Year Neurodevelopmental Outcome in Neonatal Encephalopathy

Abstract

Background: MRI scoring systems are used in the perinatal period to predict later neurodevelopmental outcome in infants with neonatal encephalopathy (NE). Our objective was to compare the predictive ability of 3 MRI scoring systems for neurodevelopmental outcome at 2 years of age in infants with NE. Methods: We retrospectively analysed MRI scans from term-born neonates with NE in the first two weeks of life between February 2011 and June 2015 using three scoring systems: Barkovich, NICHD-NRN and Weeke. Neurodevelopmental outcome was assessed using Bayley Scales of Infant and Toddler Development, third edition (Bayley-III) 2 years later. After adjusting for postmenstrual age at MRI scan and therapeutic hypothermia, the relationship between scoring systems and Bayley-III composite scores was assessed by regression analysis and logistic regression was used to calculate the predictive value of MRI scoring systems for adverse outcome (Bayley-III score of <85 in any domain). Findings: Of the 136 infants recruited, MRI was performed in 99 infants. 66 infants (cooled n=49; non-cooled n=17) were assessed at two years of age, of which 16 (24%) had adverse outcomes. NICHD-NRN and Weeke scores were associated with cognitive, language and motor scores and Barkovich scores were associated with cognitive, and motor scores. In predicting adverse outcome at 2 years, sensitivity and specificity was 0.56 and 0.82 for conventional-based Barkovich, 0.56 and 0.96 for Barkovich with DWI, 0.69 and 0.80 for NICHD-NRN, and 0.50 and 0.96 for Weeke. Subcortical brain injury was highly predictive of cognitive and motor deficits, as well as adverse outcome. MRI scores in cooled infants had lower predictive ability for adverse outcome compared with non-cooled infants. Interpretation: Our results show MRI scoring systems demonstrated high specificity, however, we were not able to identify which system best predicted adverse outcome at 2 years. Therapeutic hypothermia treatment seems to reduce the predictive value of MRI scores for neurodevelopmental outcome. Funding Statement: National Children’s Research Centre, Tallaght, Ireland. Declaration of Interests: None to declare. Ethics Approval Statement: Approval was granted by the Ethics Committee of the National Maternity Hospital, Dublin, and Children’s University Hospital, Temple Street, Dublin. Written informed consent was not required in this retrospective cohort study and all patient data were processed anonymously.