Translocator protein PET imaging in temporal lobe epilepsy: A reliable test-retest study using asymmetry index

Translocator protein (TSPO) targeting positron emission tomography (PET) imaging radioligands have potential utility in epilepsy to assess the efficacy of novel therapeutics for targeting neuroinflammation. However, previous studies in healthy volunteers have indicated limited test-retest reliability of TSPO ligands. Here, we examine test-retest measures using TSPO PET imaging in subjects with epilepsy and healthy controls, to explore whether this biomarker can be used as an endpoint in clinical trials for epilepsy.

Five subjects with epilepsy and confirmed mesial temporal lobe sclerosis (mean age 36 years, 3 men) were scanned twice-on average 8 weeks apart-using a second generation TSPO targeting radioligand, [¹¹C]PBR28. We evaluated the test-retest reliability of the volume of distribution and derived hemispheric asymmetry index of [¹¹C]PBR28 binding in these subjects and compared the results with 8 (mean age 45, 6 men) previously studied healthy volunteers.

The mean (± SD) of the volume of distribution (V_T), of all subjects, in patients living with epilepsy for both test and retest scans on all regions of interest (ROI) is 4.49 ± 1.54 vs. 5.89 ± 1.23 in healthy volunteers. The bias between test and retest in an asymmetry index as a percentage was small (-1.5%), and reliability is demonstrated here with Bland-Altman Plots (test mean 1.062, retest mean 2.56). In subjects with epilepsy, V_T of [¹¹C]PBR28 is higher in the (ipsilateral) hippocampal region where sclerosis is present than in the contralateral region.

When using TSPO PET in patients with epilepsy with hippocampal sclerosis (HS), an inter-hemispheric asymmetry index in the hippocampus is a measure with good test-retest reliability. We provide estimates of test-retest variability that may be useful for estimating power where group change in V_T represents the clinical outcome.