Early postnatal cranial ultrasonography linear measures to predict neurodevelopment at 2 years in infants born at <30 weeks’ gestation without major brain injury on sequential imaging

© 2020 Rocco CuzzillaBackground: Infants born very preterm – less than 32 weeks’ gestational age (GA) – are at risk of long-term adverse neurodevelopment related to perinatal brain injury and aberrations in brain growth and maturation. Neuroimaging within the newborn period can aid clinicians to identify high-risk infants for developmental surveillance and early intervention therapy. Cranial ultrasonography (cUS) is the most frequently used neuroimaging modality for preterm infants because it is widely available, portable, and repeatable. Whilst early and sequential cUS can reliably detect most major preterm brain injury, it remains less sensitive than brain magnetic resonance imaging (MRI) for the more prevalent, diffuse white matter injury associated with very preterm birth and adverse neurodevelopment. Declining rates of major preterm brain injury further diminishes the sensitivity of cUS to predict neurodevelopmental outcomes. In the absence of major brain injury, and where brain MRI is not readily accessible, there is a need to improve the prognostic utility of cUS, not least to better understand why some very preterm infants without major brain injury seen on cUS later develop motor and cognitive impairments. Sequential cUS, from the first weeks after birth and up to term-equivalent age (TEA), affords an opportunity to explore the prognostic utility of early postnatal brain growth as a potential marker of neurodevelopment in very preterm infants. Objectives: In my thesis, I aimed to explore the utility of early postnatal cUS linear measures of brain size and brain growth to predict neurodevelopment at 2 years in infants born at less than 30 weeks’ GA and without major brain injury. My first study aimed: (1) to assess the reproducibility of linear measures of brain tissue and fluid spaces made from cUS performed as part of routine clinical care; (2) to evaluate early postnatal brain growth using sequential cUS linear measures made from birth and up to TEA; and (3) to explore perinatal predictors of early postnatal brain growth. My second and third studies aimed to examine the associations of early postnatal cUS linear measures with (4) neurobehaviour at TEA and (5) neurodevelopment at 2 years, respectively. Methods: Participants comprised 144 infants born at less than 30 weeks’ GA, and 201 term-born controls, at a single centre between January 2011 and December 2013. Infants with major brain injury seen on cUS, or congenital or chromosomal abnormalities known to affect neurodevelopment, were excluded. Linear measures of brain tissue and fluid spaces were made from cUS performed as part of routine clinical care from the first weeks after birth and up to TEA. Perinatal data were collected prospectively. Neurobehaviour was assessed at TEA using Prechtl’s Qualitative Assessment of General Movements (GMs) and the Hammersmith Neonatal Neurological Examination (HNNE). Neurodevelopment was assessed at 2 years using the cognitive, language, and motor scales on the Bayley Scales of Infant and Toddler Development, 3rd Edition, and risk of adverse neurodevelopment at 2 years was defined by a composite of either: cerebral palsy, any developmental delay, deafness, or blindness. Assessors were blinded to the clinical course of participants. Results: 429 scans were assessed for 144 infants. Several linear measures showed excellent reproducibility. All measures of brain tissue increased with postnatal age, except for the biparietal diameter (BPD) which decreased within the first postnatal week and increased thereafter. GA greater than or equal to 28 weeks at birth was associated with slower growth of the BPD and ventricular width compared with GA less than 28 weeks. Postnatal corticosteroid administration was associated with slower growth of the corpus callosum length (CCL), transcerebellar diameter (TCD) and vermis height. Sepsis and necrotising enterocolitis were associated with slower growth of the TCD. Larger brain sizes, assessed by linear measures of the BPD at 1-week and 2-months, and CCL at 1-month, were associated with lower risk of a suboptimal HNNE and abnormal GMs at TEA, respectively. Faster positive growth rates of the CCL and TCD between birth and 1-month were related to lower risk of abnormal GMs at TEA. A larger measure of the right anterior horn width (AHW) at 1-month, and faster positive rates of increase of the left and right AHW between 1- and 2-months, were related to lower risk of abnormal GMs and a suboptimal HNNE at TEA, respectively. Larger measures of brain tissue at 1-week, 1-month, and 2-months were related to higher cognitive and language scores, and larger measures of brain tissue at 2-months were related to lower risk of adverse neurodevelopment. Faster growth of the TCD between 1-month and 2-months was related to a higher cognitive score and lower risk of adverse neurodevelopment. Unexpectedly, larger measures of fluid spaces at 1-month and 2-months were related to higher language and motor scores, larger measures of fluid spaces at 1-week and 1-month were related to lower risk of adverse neurodevelopment, and a faster increase of the interhemispheric distance between 1-week and 1-month was related to a higher language score and lower risk of adverse neurodevelopment. Conclusions: Early postnatal brain growth in infants born at less than 30 weeks’ GA can be evaluated using sequential linear measures made from routine cUS, and is associated with perinatal predictors of long-term development. Linear measures of larger brain tissue from the first weeks after birth are related to lower risk of atypical neurobehaviour at TEA and better cognitive and language development at 2 years. My thesis provides evidence to support associations of early postnatal cUS linear measures of brain size and brain growth with neurodevelopment at 2 years in infants born less than 30 weeks’ GA and free of major brain injury. The relationships between larger fluid spaces and better neurodevelopment warrant further exploration