Regional impairments of cortical folding in premature infants

Abstract

Objective: This study was undertaken to evaluate the influence of preterm birth and other factors on cerebral cortical maturation.

Methods: We have evaluated the effects of preterm birth on cortical folding by applying cortical cartography methods to a cohort of 52 preterm infants (<31 weeks gestation, mild or no injury on conventional magnetic resonance imaging) and 12 term-born control infants. All infants were evaluated at term-equivalent postmenstrual age.

Results: Preterm infants had lower values for the global measures of gyrification index (GI; 2.06 ± 0.07 vs 1.80 ± 0.12, p < 0.001; control vs preterm) and cortical surface area (CSA; 316 ± 24 cm(2) vs 257 ± 40 cm(2) , p < 0.001). Regional analysis of sulcal depth and cortical shape showed the greatest impact of preterm birth on the insula, superior temporal sulcus, and ventral portions of the pre- and postcentral sulci in both hemispheres. Although CSA and GI are related, CSA was more sensitive to antenatal and postnatal factors than GI. Both measures were lower in preterm infants of lower birth weight standard deviation scores and smaller occipitofrontal circumference at time of scan, whereas CSA alone was lower in association with smaller occipitofrontal circumference at birth. CSA was also lower in infants with higher critical illness in the first 24 hours of life, exposure to postnatal steroids, and prolonged endotracheal intubation.

Interpretation: Preterm birth disrupts cortical development in a regionally specific fashion with abnormalities evident by term-equivalent postmenstrual age. This disruption is influenced by both antenatal growth and postnatal course.

Figure 1

Average left and right midthickness cortical surfaces for term controls and preterm infants with average sulcal depth maps overlaid. The color bar shows the scale of the sulcal depths in millimeters.

Figure 2

Differences in gyrification index (GI) and cortical surface area (CSA) between preterm and control infants.

Figure 3

t Statistic maps of sulcal depth differences between control and preterm infants. The top row illustrates lateral and medial views, whereas the second row illustrates ventral and dorsal views for both the left and right hemispheres. The color bar shows the t values, with yellow and red indicating regions where controls have deeper sulci than preterm, whereas blues indicate regions where sulci are deeper in preterm infants. Black contours identify regions of significant (p < 0.025) difference between groups.

Figure 4

Anatomical slice views (for visualization) of an average T2-weighted image comprised of the 12 control infants used in this study and 20 preterm infants with little or no brain injury. The blue and red ribbons outline the average anterior commissure–posterior commissure midthickness cortical surfaces of the control and preterm infants, respectively. (A) A coronal slice through the insula and temporal lobe. Significant differences in size and shape can be observed. (B) A more posterior coronal slice through the temporal lobe and insula. The yellow arrow identifies the posterior portion of the superior temporal sulcus. (C) A horizontal slice. The more anterior yellow arrow identifies the precentral sulcus. The more posterior yellow arrow identifies the superior temporal sulcus.

Figure 5

Spatially normalized coordinate distance analysis for preterm and control infants. The top row (lateral and medial views) and bottom row (dorsal and ventral) show red–green–blue (RGB) maps for the left and right hemispheres that illustrate the differences between groups. The RGB maps show the directionality of the differences (red = medial–lateral shift, green = anterior–posterior shift, blue = superior–inferior shift). Because the majority of both hemispheres reach significance, for all maps, white contours encircle regions that were not significantly (p < 0.025) different between groups.