Dots and spots: A retrospective review of T2-hyperintense white matter lesions in pediatric patients with and without headache.

Abstract

We aimed to determine if T2-weighted hyperintense white matter lesions (WMLs) on brain magnetic resonance imaging (MRI) occur more frequently in pediatric patients with migraine and other primary headache disorders compared to the general pediatric population. Small foci of T2 hyperintensity in the white matter are frequently identified on brain MRI during the workup of pediatric headache. Such lesions have been reported to be more common among adults with migraine versus adults without migraine; however, this association has not been well established in the pediatric population. We performed a retrospective cross-sectional single-center study of electronic medical records and radiologic studies, examining pediatric patients from 3 to 18 years old who underwent brain MRI between 2016 and 2021. Patients with existing intracranial disease or abnormalities were excluded. Patients with reports of headache were categorized. Imaging was reviewed to determine the number and location of WMLs. Headache-associated disability scores (Pediatric Migraine Disability Assessment) were noted, when available. Brain MRI of 248 patients with a diagnosis of headache (144 with migraine, 42 with non-migraine primary headache, and 62 with headache that could not be further classified) and 490 controls were reviewed. WMLs were encountered commonly among all study participants, with a prevalence of 40.5% (17/42) to 54.1% (265/490). There was no statistically significant difference comparing the number of lesions between each of the headache groups and the control group: migraine group versus control group median [interquartile range (IQR)], 0 [0-3] versus 1 [0-4], incidence rate ratio [95% confidence interval (CI)], 0.99 [0.69-1.44], p = 0.989, non-migraine headache group versus control group median [IQR], 0 [0-3] versus 1 [0-4], 0.71 [0.46-1.31], p = 0.156, headache not otherwise specified group versus control group median [IQR], 0 [0-4] versus 1 [0-4], 0.77 [0.45-1.31], p = 0.291. There was no significant correlation between headache-associated disability and the number of WMLs (0.07 [-0.30 to 0.17], rho [95% CI]). T2 hyperintense WMLs are common within the pediatric population and are not encountered more frequently in pediatric patients with migraine or other primary headache disorders. Thus, such lesions are presumably incidental and unlikely related to headache history.