White matter fiber morphology in persisting postconcussive symptoms and posttraumatic headache after pediatric concussion: a fixel-based analysis

Feiven Fan,Richard Beare,Sila Genc,Jesse S Shapiro,Michael Takagi,Stephen J C Hearps,Georgia M Parkin,Vanessa C Rausa,Nicholas Anderson,Fabian Fabiano,Kevin Dunne,Gavin A Davis,Franz E Babl,Vera Ignjatovic,Marc Seal,Vicki Anderson

DOI: https://doi.org/10.3171/2024.6.PEDS2499

2024-07-26

Abstract:Objective: Posttraumatic headache (PTH) represents the most common acute and persistent postconcussive symptom (PCS) in children after concussion, yet there remains a lack of valid and objective biomarkers to facilitate risk stratification and early intervention in this patient population. Fixel-based analysis of diffusion-weighted imaging, which overcomes constraints of traditional diffusion tensor imaging analyses, can improve the sensitivity and specificity of detecting white matter changes postconcussion. The aim of this study was to investigate whole-brain and tract-based differences in white matter morphology, including fiber density (FD) and fiber bundle cross-section (FC) area in children with PCSs and PTH at 2 weeks after concussion. Methods: This prospective longitudinal study recruited children aged 5-18 years who presented to the emergency department of a tertiary pediatric hospital with a concussion sustained within the previous 48 hours. Participants underwent diffusion-weighted MRI at 2 weeks postinjury. Whole-brain white matter statistical analysis was performed at the level of each individual fiber population within an image voxel (fixel) to compute FD, FC, and a combined metric (FD and bundle cross-section [FDC]) using connectivity-based fixel enhancement. Tract-based Bayesian analysis was performed to examine FD in 23 major white matter tracts. Results: Comparisons of 1) recovered (n = 27) and symptomatic (n = 16) children, and those with 2) PTH (n = 13) and non-PTH (n = 30; overall mean age 12.99 ± 2.70 years, 74% male) found no fiber-specific white matter microstructural differences in FD, FC, or FDC at 2 weeks postconcussion, when adjusting for age and sex (family-wise error rate corrected p value > 0.05). Tract-based Bayesian analysis showed evidence of no effect of PTH on FD in 10 major white matter tracts, and evidence of no effect of recovery group on FD in 3 white matter tracts (Bayes factor < 1/3). Conclusions: Using whole-brain fixel-wise and tract-based analyses, these findings indicate that fiber-specific properties of white matter microstructure are not different between children with persisting PCSs compared with recovered children 2 weeks after concussion. These data extend the limited research on white matter fiber-specific morphology while overcoming limitations inherent to traditional diffusion models. Further validation of our findings with a large-scale cohort is warranted.