Impaired development of the cerebral cortex in infants with congenital heart disease is correlated to reduced cerebral oxygen delivery

Abstract

Neurodevelopmental impairment is the most common comorbidity associated with complex congenital heart disease (CHD), while the underlying biological mechanism remains unclear. We hypothesised that impaired cerebral oxygen delivery in infants with CHD is a cause of impaired cortical development, and predicted that cardiac lesions most associated with reduced cerebral oxygen delivery would demonstrate the greatest impairment of cortical development. We compared 30 newborns with complex CHD prior to surgery and 30 age-matched healthy controls using brain MRI. The cortex was assessed using high resolution, motion-corrected T2-weighted images in natural sleep, analysed using an automated pipeline. Cerebral oxygen delivery was calculated using phase contrast angiography and pre-ductal pulse oximetry, while regional cerebral oxygen saturation was estimated using near-infrared spectroscopy. We found that impaired cortical grey matter volume and gyrification index in newborns with complex CHD was linearly related to reduced cerebral oxygen delivery, and that cardiac lesions associated with the lowest cerebral oxygen delivery were associated with the greatest impairment of cortical development. These findings suggest that strategies to improve cerebral oxygen delivery may help reduce brain dysmaturation in newborns with CHD, and may be most relevant for children with CHD whose cardiac defects remain unrepaired for prolonged periods after birth.

Figure 1

Cerebral oxygen delivery (CDO2) demonstrates a positive association with grey matter volume (a) and whole brain gyrification (b). These trends persist after indexing CDO2 per 100 ml brain volume (c and d). Regional cerebral oxygen saturation has a limited positive relationship with gyrification index (e). Abnormal mixing lesions and left sided lesions demonstrate a significantly lower gyrification index, while right-sided lesions are less affected (f).

Figure 2

Gyrification index differences between newborns with complex congenital heart disease (open blue) and healthy controls (solid orange), in the whole brain (a) and exploratory regional analysis (b–e). The cortical surface visualisation (f) demonstrates regions where gyrification is reduced in newborns with congenital heart disease compared to healthy term controls, from the left lateral side (i) and from above (ii); colours represent p values from multivariate general linear models, using postmenstrual age as a covariate; no multiple comparisons correction has been performed in this visualisation.

Figure 3

Demonstration of the calculation of gyrification index. (a) Original description of gyrification index in histology setting, (b) Neonatal brain-extracted T2 volume, (c) Automatic segmentation, (d) Pial surface mesh (green) and superficial surface (red) created from the segmentation, used to calculate the gyrification index. Figure 3a reproduced from The human pattern of gyrification in the cerebral cortex, Zilles, K., Armstrong, E., Schleicher, A. & Kretschmann, H. J. Anat. Embryol. (Berl). 179, 173–179 (1988). Copyright Springer-Verlag 1988. With permission of Springer.

Figure 4

Phase contrast measurements of the cerebral vessels in the neonatal brain. The plane is planned from a 3D non-contrast angiogram in both coronal (a) and sagittal planes (b), aiming for the C4 segment of the internal cerebral arteries (ICA) where all three vessels are running approximately parallel). Following the scan, regions of interest are drawn around the three major cerebral vessels: left (green) and right (red) ICAs, and basilar artery (blue), and these regions are propagated through the cardiac cycle. Flow curves can then be derived across the cardiac cycle (d).