Abstract / Summary
Background/Objectives: Extremely preterm infants remain at high risk of adverse neurodevelopmental outcomes, particularly in the presence of intraventricular hemorrhage (IVH) and associated brain injury. Term-equivalent cerebral magnetic resonance imaging (cMRI) provides a detailed assessment of structural brain injury and may provide anatomical information not captured by conventional IVH grading. We aimed to investigate whether the extent and regional distribution of cMRI-detected brain injury are associated with neurodevelopment at two years corrected age. Methods: This retrospective cohort study included extremely preterm infants born before 28 weeks’ gestation with IVH, term-equivalent cMRI, and neurodevelopmental follow-up at two years corrected age. Regional brain injury was assessed on T2-weighted cMRI using a Bassan-derived anatomical approach. Associations with neurodevelopment were examined using multivariable regression models. Results: A total of 137 infants were included in the regional cMRI analysis. The regional injury score ranged from 0 to 10. Each additional affected cerebral region was independently associated with lower cognitive (β −3.62; 95% CI −5.09 to −2.14), language (β −3.13; 95% CI −4.90 to −1.36), and motor scores (β −4.70; 95% CI −6.08 to −3.32), all p < 0.001, after adjustment for prespecified clinical and demographic covariates. Associations remained significant after adjustment for conventional IVH classification. Inclusion of the regional cMRI injury score significantly improved model fit for cognitive (p = 0.019), language (p = 0.006), and motor outcomes (p < 0.001). In exploratory region-specific analyses, frontal injury, particularly posterior frontal involvement, showed the largest effect estimates. Conclusions: Higher regional cMRI injury burden was associated with poorer cognitive, language, and motor outcomes in extremely preterm infants with IVH. Exploratory regional assessment suggested that frontal injury was associated with adverse neurodevelopmental outcomes. These findings support the value of considering both the extent and anatomical distribution of brain injury on term-equivalent cMRI when characterizing neurodevelopmental risk.