Abstracts

Rationale: Overexpression of the multidrug efflux transporter P-glycoprotein (P-gp) at the blood-brain-barrier (BBB) is thought to be involved in pharmacoresistance in epilepsy by extruding antiepileptic drugs (AEDs) from their target site. We performed positron emission tomography (PET) scans with the P-gp substrate (R)-[11C]verapamil (VPM) in refractory and seizure-free mTLE patients and healthy controls. Additionally, refractory mTLE patients and healthy controls received the P-gp inhibitor Tariquidar (TQD). We hypothesize that VPM uptake at baseline and its increase after P-gp inhibition is reduced in refractory mTLE ipsilaterally relative to contralateral and to seizure-free patients as well as healthy controls.Methods: To date 5 healthy controls (4 fem, age 37-55 y), 6 seizure-free (4 fem, age 23-50 y) and 9 refractory (6 fem, age 18-56 y) mTLE patients with unilateral hippocampal sclerosis underwent a VPM PET scan. Refractory patients and controls underwent a second VPM PET scan 60min after i.v. infusion of 2mg/kg (n=11) or 3mg/kg (n=3) of TQD. The transfer rate constant from plasma to brain, K1, was estimated using a single-tissue compartment model with a VPM-in-plasma arterial input function. Analysis was performed on the first 10min of dynamic data containing limited radiolabeled metabolites. Regions were defined automatically using a brain atlas, and ratios of K1 values were calculated between a reference region (parietal cortex) and target regions. Parametric maps of K1 were generated using generalised linear least squares which was extended to include blood volume and used for SPM voxel-based analysis.Results: At baseline K1 values were significantly higher globally in both patient groups compared to controls (p<0.05) but refractory were not different from seizure-free mTLE patients (p=0.239). The peripheral metabolism of VPM was not different between patient groups (p=0.932) but significantly faster in patients than controls (p<0.005). Ratios of K1 values were significantly lower globally in refractory compared to seizure-free patients (p<0.01) (Figure 1). After TQD, VPM K1 increased significantly for temporal lobe regions in controls (60%) compared to mTLE patients (26%) (p<0.05). Preliminary voxel-based analysis reveals fronto-temporal reductions in patients ipsilateral to the seizure focus compared to controls before TQD, which is further accentuated following TQD (Figure 2).Conclusions: Lower brain uptake of the P-gp substrate VPM in refractory compared to seizure-free mTLE patients and the two-fold lower increase in VPM brain uptake after TQD in refractory mTLE patients compared to healthy controls support the hypothesis of P-gp overexpression in refractory epilepsy. Voxel-based analysis is suggestive of ipsilateral fronto-temporal accentuation of this effect, but larger cohorts are required to assess specific regional changes. The global differences raise the question whether epilepsy per-se, uncontrolled seizures or chronic treatment with AEDs stimulate a global rather than regional response of P-gp function.