Very preterm infants are at an increased risk of cognitive, motor, and behavioural impairments resulting from diffuse white matter injury and accompanying axonal deficits involving the thalamus, basal ganglia, cerebral cortex, brainstem and cerebellum¹. Neuroimaging studies have demonstrated that poor functional outcome in very preterm infants is related to reduced thalamic and basal ganglia volumes^2,3, which appear to persist into childhood and adolescence⁴. However, it is currently not known whether these volumetric deficits are associated with microstructural changes in tissue integrity. Quantitative relaxometry enables the assessment of subtle changes in tissue integrity associated with alterations in the MR relaxation times T1 and T2, and may provide greater sensitivity to microstructural alterations than volumetric methods. The purpose of this study was to investigate differences in cerebral microstructure in very preterm children and adolescents in comparison to their term-born peers, using quantitative MR relaxometry.

Participant Group: Very pre-term group: Thirty one children and adolescents (mean age 12, range 10-16) born ≤ 32 weeks of gestation, with no evidence of periventricular leukomalacia or haemorrhagic infarction on neonatal ultrasound and no diagnosis of cerebral palsy or developmental delay at the routine follow-up assessment between the ages of four and eight years. Controls: thirty-one term-born children and adolescents, with no history of perinatal complications, and no evidence of any neurodevelopmental illness (e.g., ADHD). Control participants were group matched to the very preterm participants with regard to sex and age.

MRI Acquisition: MRI scans were acquired with a 3T GE MR750 scanner using an 8 channel receive-only head coil. Quantitative MR relaxometry was performed using the driven equilibrium single pulse observation of T1 with high-speed incorporation of RF field inhomogeneities (DESPOT1-HIFI) method.⁵ The acquisition comprised a series of 3D sagittal spoiled gradient recalled echo (SPGR) images with a range of flip angles (3, 4, 5, 6, 7, 9, 13, and 18 degrees), with echo time (TE) = 2.38 ms, repetition time (TR) = 5.8 ms, field of view = 220x 220 mm, matrix 256x256. A sagittal inversion-recovery prepared SPGR image volume (IR-SPGR) was also acquired (TE = 2.38 ms, TR= 5.8 ms, inversion time = 450 ms, flip angle= 5 degrees) to allow correction for B1 inhomogeneities.

MRI Data Analysis: The SPGR and IR-SPGR data for each participant were linearly co-registered to correct for intra-session motion using the FSL linear registration tool FLIRT⁶, and skull stripped with the BET brain extraction tool⁷. A B1 map was calculated for each participant using the DESPOT1-HIFI method⁵, and then smoothed with a 6mm median filter using fslmaths. Quantitative T1 maps were then calculated from the variable flip angle SPGR images, correcting for B1 inhomogeneities with the smoothed B1 map⁵. The total intracranial volume was calculated with freesurfer⁸. T1 maps for each participant were normalized for voxelwise group analysis using the registration methods implemented in the FSL TBSS pipeline⁹. Each T1 map was aligned to the most representative T1 map in the cohort, and an age-appropriate template was then derived from the average of the normalized T1 maps. Voxelwise statistical analysis of the T1 data was performed with FSL randomize to test for differences in T1 between very preterm and term-born participants, controlling for age and gender. A statistical threshold of p<0.05 was applied after family wise error (FWE) correction for multiple comparisons using threshold free cluster enhancement (TFCE).

Very preterm children and adolescents did not differ from their term-born peers in mean age, gender, or total brain volume. No difference was observed in the volume of the thalamus or caudate between very preterm children and adolescents and their term-born peers. Very preterm children and adolescents showed significantly increased T1 relaxation times in the caudate and thalamus and decreased T1 relaxation times in insula cortex and amygdala/hippocampus (p<0.05, FWE corrected, controlling for age and gender).Quantitative T1 relaxometry reveals microstructural alterations in the caudate, thalamus, and insula in very preterm children and adolescents, even in the absence of overt volumetric differences. These results highlight the vulnerability of basal ganglia, thalamic and cortical structures to neonatal brain injury, in keeping with the notion of an “encephalopathy of prematurity”¹. Quantitative relaxometry may therefore play an important role in evaluating the subtle microstructural changes associated with prematurity.