Predicting Short- and Long-Term Neurodevelopmental Outcome Based on Term-Equivalent Age MRI in Extremely Preterm Infants

Abstract

Objectives Extremely preterm infants (born <28 weeks of gestation) are at high risk for brain abnormalities as a complication of preterm birth, which can lead to impaired neurodevelopmental outcome. Better prediction early in life may steer targeted interventions and improve outcome. Therefore, this study aims to relate brain abnormalities seen on term-equivalent age (TEA) magnetic resonance imaging (MRI) to neurodevelopmental outcome at two and five years of age in a large cohort of extremely preterm infants and to determine risk factors associated with outcome.

Study design Brain abnormalities and maturation of the white matter, cortical and deep grey matter, and cerebellum were scored on TEA-MRI in 352 infants born before 28 weeks of gestation, using the Kidokoro scoring system. Cognitive, motor, and behavioural outcome was assessed at two and five years of age using standardised tests (Bayley-III-NL, WPPSI-III-NL, MABC-2-NL, and CBCL). Associations between TEA-MRI and outcome were determined using univariable and multivariable linear regression analyses.

Results Cognitive and motor outcome at two and five years could be well predicted by the global brain abnormality score, even after adjusting for perinatal variables and maternal education (B = -.808, p = .005, B = -.704, p = .007, B = -1.299, p = <.001, B = -.207, p = .003, respectively). However, the prognostic value of TEA-MRI related to behavioural outcome was limited. At two years of corrected age, numerous perinatal variables which contributed to outcome were determined, while these played a less important role at five years of age. Maternal education contributed to multiple outcome measures at both ages.

Conclusion The prognostic value of TEA-MRI related to cognitive and motor outcome at two and five years of age is promising. However, TEA-MRI was limited in predicting behavioural outcome for both ages. Furthermore, maternal education played an important role in predicting outcome, suggesting more intensive support for lower educated parents may be necessary.