Impacts of perinatal factors on white matter outcome at 8 to 10 years by diffusion tensor imaging

1303 

Abstract Submission 

Injoong Kim^1,2, Omar Azrak², Mark Foster², Emil Cornea², Sang Kyoon Park², Martin Styner², John Gilmore²

¹Department of Radiology, Veterans Health Service Medical Center, Seoul, Republic of Korea, ²Department of Psychiatry, University of North Carolina Chapel Hill, Chapel Hill, NC, United States

Injoong Kim  
Department of Radiology, Veterans Health Service Medical Center|Department of Psychiatry, University of North Carolina Chapel Hill
Seoul, Republic of Korea|Chapel Hill, NC, United States

Omar Azrak  
Department of Psychiatry, University of North Carolina Chapel Hill
Chapel Hill, NC, United States

Mark Foster  
Department of Psychiatry, University of North Carolina Chapel Hill
Chapel Hill, NC, United States

Emil Cornea  
Department of Psychiatry, University of North Carolina Chapel Hill
Chapel Hill, NC, United States

Sang Kyoon Park  
Department of Psychiatry, University of North Carolina Chapel Hill
Chapel Hill, NC, United States

Martin Styner  
Department of Psychiatry, University of North Carolina Chapel Hill
Chapel Hill, NC, United States

John Gilmore  
Department of Psychiatry, University of North Carolina Chapel Hill
Chapel Hill, NC, United States

Postnatal brain development is a complex process that is under active investigation. Diffusion tensor imaging (DTI) can reveal altered white matter microstructure by using the orientation and integrity of white matter tracts. Here, we aimed to investigate how common neonatal indicators of health (birthweight [BW], gestational age [GA], head circumference at birth [HC]) and brain development are related to DTI-based measurement of white matter microstructure in 8 to 10 years old near-term-born children. Recent study (Nivins et al, 2023) showed how these indicators predict brain volume and white matter microstructure in the same age group, with only BW showing predictive power for white matter microstructure via a traditional tract-based spatial statistics (TBSS) analysis. Building on these results, the goal of the current study was to replicate their findings in a separate sample via an alternative, more sensitive DTI analysis framework.

From the University of North Carolina's Early Brain Development Study (EBDS) database, we selected participants with MRI scans collected within 8 to 10 years of age on a 3T Siemens Tim Trio (b=0 [13], 300 [8], 700 [32], 2000 [64 unique gradients]; 2x2x2mm resolution) and perinatal clinical information. Ninety-two children born at ≥ 30 weeks' gestation age were included (male, 42 [45%]; median age, 10 [range 8-10.9 years]). A study-specific quality control protocol was applied to all raw DTI data using the dmriprep (Dubos et al, 2023) module in the DMRIPlayground toolkit (https://github.com/NIRALUser/DTIPlayground). A study-specific DTI atlas was created via the dmriatlas module in DMRIPlayground. Fifty-three major white matter tracts were determined in that DTI atlas space via propagation and automated tracking (Ngattai Lam et al, 2018) of the EBDS pediatric DTI atlas (Short et al, 2022). Diffusion tensor metrics (fractional anisotropy [FA], axial diffusivity [AD], radial diffusivity [RD]) were extracted at evenly spaced points (arc lengths) along each fiber tract. Statistical analysis was performed via FADTTS (Zhu et al, 2011), covarying for gender, yielding tract-wise global as well as local p-value maps. For each tract, a multivariate analysis was first performed combining all three DTI metrics, followed by a posthoc analysis for each FA, RD and AD separately.

Among the three factors (BW, GA, HC), BW showed the most widespread significant associations (p<0.05) with white matter diffusion metrics globally (28 out of 53 white matter tracts) and locally (see Figure 1). GA (15 out of 53) and HC (16 out of 53) also showed significant associations though with a lower number of tracts affected. Gender was also highly associated with the diffusion metrics globally in most white matter tracts.
In the posthoc analysis, we found surprisingly far more correlations for all three perinatal factors with AD and RD than with FA. AD and RD associations with BW, GA and HC were found in 18, 8, 9 and 17, 8, 9 out of 53 white matter tracts, respectively. For FA, 5, 4, 3 out of 53 white matter tracts showed significant associations.

In comparison with previous study (Nivins et al, 2023), we show that neonatal measures of birthweight, gestational age, and head circumference at birth are all significantly predictive of white matter microstructure at age 8-10 years across many fiber tracts, with a focus on central white matter locations. These findings might hold the potential to offer substantial insights into the intricate relationship between perinatal factors and the development of the brain's white matter during childhood.
Please note that the presented analysis is based on a mid-size sample (n=92) scanned on a single 3T scanner. We are currently in the process to extend this study to include all EBDS subjects scanned at the ages of 8-10 years, increasing the sample size to over 300 subjects acquired on 3 different 3T Siemens scanners.

Normal Brain Development: Fetus to Adolescence ¹

White Matter Anatomy, Fiber Pathways and Connectivity ²

Development

MRI

PEDIATRIC

Tractography

White Matter

WHITE MATTER IMAGING - DTI, HARDI, DSI, ETC