Role of the advanced MRI sequences in predicting the outcome of preterm neonates

AIM The aim of the project is to evaluate the role of advanced MRI sequences (susceptibility weight imaging (SWI), diffusion tensor imaging (DTI), and arterial spin labeling (ASL) perfusion) in detecting early changes that affect preterm neonatal brain, especially in those patients without lesions at conventional MRI or with small brain injuries (i.e. low grade germinal matrix-intraventricular hemorrhage (GMHIVH)), and to correlate these subtle brain abnormalities with neurodevelopmental outcome at 24 months. METHODS Since November 2015 until June 2017, 287 preterm neonates and 108 term neonates underwent a 3T or 1.5T MRI study at term corrected age (40±1 weeks). SWI, DTI and ASL sequences were performed in all neonates. SWI sequences were evaluated using both a qualitative (SWI venography) and quantitative (Quantitative Susceptibility Map analysis (SWI-QSM)) approach. DTI data were analyzed using a Tract-Based Spatial Statistics analysis (TBSS). ASL studies were processed to estimate Cerebral Blood Flow (CBF) maps. Perinatal clinical data were collected for all neonates. Neurodevelopmental data were evaluated at 24 months in 175 neonates using 0-2 Griffiths Developmental Scales. RESULTS The analysis performed on SWI-venography revealed differences in subependymal veins morphology between preterm and term neonates with normal brain MRI, with a higher variability from the typical anatomical pattern in preterm neonates. The same analysis performed in preterm neonates with GMH-IVH revealed that the anatomical features of subependymal veins may play a potential role as predisposing factor for GMH-IVH. Moreover, the SWI-QSM analysis revealed a greater paramagnetic susceptibility in several periventricular white matter (WM) regions in preterm neonates with GMH-IVH than in healthy controls. This finding is likely related to the accumulation of hemosiderin/ferritin following the diffusion of large amounts of intraventricular blood products into the WM, and it is also supposed to trigger the cascade of lipid peroxidation and free radical formation that promote oxidative and inflammatory injury of the WM in neonatal brain after GMH-IVH. The TBSS analysis confirmed that microstructural WM injury can occur in preterm neonates with low grade GMH-IVH even in the absence of overt signal changes on conventional MRI, with different patterns of WM involvement depending on gestational age. Moreover, the distribution of these WM microstructural alterations after GMH-IVH correlates with specific neurodevelopmental impairments at 24 months of age. Finally, the analysis of brain perfusion at term-corrected age revealed lower CBF in preterms with sub-optimal neuromotor development, reinforcing the hypothesis that impaired autoregulation of CBF may contribute to the development of brain damage in preterm neonates. CONCLUSION Advanced MRI sequences can assist the standard perinatal brain imaging in the early diagnosis of preterm neonatal brain lesions and can provide new insights for predicting the neurodevelopmental trajectory. However, detailed and serial imaging of carefully chosen cohorts of neonates coupled with longer clinical follow-up are essential to ensure the clinical significance of these novel findings.