Tract-Based Spatial Statistics in Preterm-Born Neonates Predicts Cognitive and Motor Outcomes at 18 Months

Abstract

Background and purpose: Adverse neurodevelopmental outcome is common in children born preterm. Early sensitive predictors of neurodevelopmental outcome such as MR imaging are needed. Tract-based spatial statistics, a diffusion MR imaging analysis method, performed at term-equivalent age (40 weeks) is a promising predictor of neurodevelopmental outcomes in children born very preterm. We sought to determine the association of tract-based spatial statistics findings before term-equivalent age with neurodevelopmental outcome at 18-months corrected age.

Materials and methods: Of 180 neonates (born at 24-32-weeks' gestation) enrolled, 153 had DTI acquired early at 32 weeks' postmenstrual age and 105 had DTI acquired later at 39.6 weeks' postmenstrual age. Voxelwise statistics were calculated by performing tract-based spatial statistics on DTI that was aligned to age-appropriate templates. At 18-month corrected age, 166 neonates underwent neurodevelopmental assessment by using the Bayley Scales of Infant Development, 3rd ed, and the Peabody Developmental Motor Scales, 2nd ed.

Results: Tract-based spatial statistics analysis applied to early-acquired scans (postmenstrual age of 30-33 weeks) indicated a limited significant positive association between motor skills and axial diffusivity and radial diffusivity values in the corpus callosum, internal and external/extreme capsules, and midbrain (P < .05, corrected). In contrast, for term scans (postmenstrual age of 37-41 weeks), tract-based spatial statistics analysis showed a significant relationship between both motor and cognitive scores with fractional anisotropy in the corpus callosum and corticospinal tracts (P < .05, corrected). Tract-based spatial statistics in a limited subset of neonates (n = 22) scanned at <30 weeks did not significantly predict neurodevelopmental outcomes.

Conclusions: The strength of the association between fractional anisotropy values and neurodevelopmental outcome scores increased from early-to-late-acquired scans in preterm-born neonates, consistent with brain dysmaturation in this population.

Fig 1.

Participant flow chart. Neonate data (180 very preterm-born infants of <32 weeks' gestation, with 1 or 2 MR imaging scanning sessions including DTI) are separated into 3 groups: group 1, 75 neonates with only 1 early scan near the time of birth (median postmenstrual age at scanning, 32 weeks) (all 75 neonates have neurodevelopmental follow-up data at 18-month corrected age); group 2, 78 neonates with 2 scans, both early (PMA, 32 weeks) and at term-equivalent age (PMA, 39.7 weeks) (75 neonates have follow-up data); group 3, 27 neonates with a late scan (PMA, 39 weeks) (16 neonates have follow-up data).

Fig 2.

TBSS analysis of term scans (PMA of 37–41-weeks). Top: Mean FA map (red-yellow) demonstrating the significant positive linear association between cognitive scores on the Bayley-III and FA in the territory of the medial prefrontal cortex (left), the genu of the corpus callosum (middle), and portions of the inferior fronto-occipital fasciculus (right, P < .05, corrected for multiple comparisons). The mean FA skeleton is shown in green. Bottom: FA (R = 0.3, P = .03), AD (R = −0.1, P = .2), and RD (R = −0.2, P = .03) values from the significant clusters in the FA map. Spearman ρ correlation and an α level are set at .05.