Abstracts

Rationale: Although intermittent light stimulation (ILS) has been used for decades to activate photoparoxysmal responses in people with epilepsy, there is little data on frequency-dependent regional cerebral blood flow (rCBF) changes during ILS. As the baboon is the most extensively studied animal model of photosensitivity, we investigated the rate dependence of ILS-induced brain activations in both photosensitive (PS) and control (CTL) animals. Methods: Eight baboons (5 PS; 3 CTL) underwent ILS at frequencies of 5 Hz, 10 Hz, 15 Hz and 25 Hz during concurrent H215O PET scans. The baboons were sedated using intravenous ketamine (5-6 mg/kg/hr) and paralyzed with vecuronium (0.1-0.3 mg/kg). The order of stimulations was randomized and two resting scans were performed before and after ILS. The images were analyzed using a combination of voxelwise statistical parametric images (SPIs) and correlation analyses. Group z-score images (SPI{z}) were obtained by comparing group-averaged resting scans and at each stimulation frequency. Activation and deactivation clusters were localized using each animal s MRI. Mean value normalized PET counts were obtained from volumes of interest (VOI) of >125 mm3 from the occipital, motor, posterior cingulate and orbitofrontal cortices and used for statistical parametric analyses. The mean counts were compared by two-tailed t-tests for differences in means. Results: Activations were localized in the primary visual cortex, orbitofrontal and posterior cingulate cortices in PS animals. Mean PET count values for each of these areas demonstrated significant differences between each ILS frequency and baseline conditions between the two groups. Further analysis revealed an increase in PET counts with increasing ILS rates, with the primary visual regions peaking between 10-15 Hz in the CTL group, while PS baboons demonstrated maximal brain activations at 25 Hz. Connected regions (posterior cingulate and motor cortex) also demonstrated significant differences between the PS and CTL groups; rCBF increases with ILS frequency, dropping off at higher frequencies for the CTL group, whereas the PS group exhibited decreased rCBF with increasing ILS frequencies in these regions. The orbitofrontal regions demonstrated significant differences between the PS and CTL groups at 25 Hz. Conclusions: In CTL baboons, peak activations of the visual cortex occurred at an ILS frequency of 15 Hz (compared to 10 Hz in humans). Occipital CBF peaked at 25 Hz in the PS baboons, which is the frequency most likely to activate a photoparoxysmal response. Nonetheless, the regional CBF decreased in orbitofrontal and motor cortices at the same frequency, implying cortical inhibition related to interictal or ictal epileptic discharges. These findings support the use of a baboon model to investigate the mechanisms underlying ILS-induced physiological and photoparoxysmal responses in humans.