General movements and magnetic resonance imaging in the prediction of neuromotor outcome in children born extremely preterm

BackgroundExtremely preterm (EPT) birth is a major risk factor for brain injury and neurodevelopmental impairment. Reliable tools for early prediction of outcome are warranted.AimTo investigate the predictive value of general movements (GMs) at “fidgety age” for neurological outcome at age 30 months in EPT infants, both in comparison and in combination with structural magnetic resonance imaging (MRI) at term equivalent age (TEA).Study designFifty-three infants born < 27 weeks of gestation were included prospectively. MRI was performed at TEA and images were evaluated for white and grey matter abnormalities. GMs were assessed at age 3 months corrected (“fidgety age”).Outcome measuresNeuromotor outcome was assessed at age 30 months corrected. Children were classified as having a normal neurological status, unspecific signs, or cerebral palsy (CP).ResultsAbnormal GMs were a common finding, seen in 32% (17/53) of infants. Of these, six infants (11%) had definitely abnormal GMs. Four infants (8%) had a diagnosis of CP at follow up. Definitely abnormal GMs were significantly associated to CP at 30 months (Fisher's Exact test p = 0.03, sensitivity 50%, specificity 92%). Moderate-severe white matter abnormalities on MRI were more strongly associated with CP (Fisher's Exact test p < 0.001, sensitivity 100%, specificity 98%) than GMs. Combining GMs with MRI-findings at TEA increased the predictive specificity to 100% (Fisher's Exact test, p = 0.005), whereas sensitivity remained unchanged.ConclusionsThe presence of definitely abnormal GMs was predictive of CP: prediction was significantly enhanced when the GMs assessment was combined with findings from MRI obtained at TEA.