Evaluation of Macro- and Microstructural White Matter Changes in Hot Water Epilepsy

BACKGROUND AND OBJECTIVES: Hot water epilepsy (HWE) is a form of reflex epilepsy and usually lacks definitive magnetic resonance imaging (MRI) features. Although white matter abnormalities have been extensively investigated in patients with temporal lobe epilepsy or generalized epilepsies, very few studies have examined quantitative changes in reflex epilepsy. This study aims to explore the white matter changes in HWE. The objective is to investigate the white matter micro- and macrostructural changes in patients with HWE using diffusion tensor imaging (DTI) indices and three-dimensional structural-imaging-derived subcortical white matter volumes. METHODS: Eighty-eight subjects with HWE and 50 age- and sex-matched controls underwent high-resolution volumetric MRI and DTI. DTI and volumetric data were processed using tract-based spatial statistics and FreeSurfer, respectively. Between-group differences in DTI metrics and regional subcortical white matter volumes were assessed. Correlations between DTI metrics, white matter volumes, and epilepsy-specific clinical variables were also evaluated. RESULTS: Subjects with HWE showed significant alterations in DTI metrics indicative of impaired white matter integrity, including decreased fractional anisotropy and increased mean diffusivity, axial diffusivity, and radial diffusivity. These changes were evident both globally (whole-brain) and within specific supratentorial and infratentorial white matter tracts. Subcortical white matter volume loss was predominant in the thalamo-limbic, somatosensory, posterior cingulate, cuneus, and frontal regions. Positive correlations were observed between alterations in global fractional anisotropy and changes in macrostructural (subcortical) white matter volumes. CONCLUSION: Subjects with HWE exhibited global and region-specific white matter alterations at both micro- and macrostructural levels. Somatosensory subcortical involvement may be associated with the stimulus sensitivity characteristic of this disorder. Involvement of temporal-limbic and default mode network regions is consistent with the clinical semiology, while brainstem changes concur with the reflex nature of the disease.