Evidence for interictal blood–brain barrier dysfunction in people with epilepsy

Abstract Objective Interictal blood–brain barrier dysfunction in chronic epilepsy has been demonstrated in animal models and pathological specimens. Ictal blood–brain barrier dysfunction has been shown in humans in vivo using an experimental quantitative magnetic resonance imaging (MRI) protocol. Here, we hypothesized that interictal blood–brain barrier dysfunction is also present in people with drug‐resistant epilepsy. Methods Thirty‐nine people (21 females, mean age at MRI ± SD = 30 ± 8 years) with drug‐resistant epilepsy were prospectively recruited and underwent interictal T1‐relaxometry before and after administration of a paramagnetic contrast agent. Likewise, quantitative T1 was acquired in 29 people without epilepsy (12 females, age at MRI = 48 ± 18 years). Quantitative T1 difference maps were calculated and served as a surrogate imaging marker for blood–brain barrier dysfunction. Values of quantitative T1 difference maps inside hemispheres ipsilateral to the presumed seizure onset zone were then compared, on a voxelwise level and within presumed seizure onset zones, to the contralateral side of people with epilepsy and to people without epilepsy. Results Compared to the contralateral side, ipsilateral T1 difference values were significantly higher in white matter (corrected p < .05), gray matter (uncorrected p < .05), and presumed seizure onset zones ( p = .04) in people with epilepsy. Compared to people without epilepsy, significantly higher T1 difference values were found in the anatomical vicinity of presumed seizure onset zones ( p = .004). A subgroup of people with hippocampal sclerosis demonstrated significantly higher T1 difference values in the ipsilateral hippocampus and in regions strongly interconnected with the hippocampus compared to people without epilepsy (corrected p < .01). Finally, z ‐scores reflecting the deviation of T1 difference values within the presumed seizure onset zone were associated with verbal memory performance ( p = .02) in people with temporal lobe epilepsy. Significance Our results indicate a blood–brain barrier dysfunction in drug‐resistant epilepsy that is detectable interictally in vivo, anatomically related to the presumed seizure onset zone, and associated with cognitive deficits.