Abstracts

Rationale: 30-40% of patients with epilepsy (PWE) continue to have seizures despite adequate therapy. In addition to seizures, cognitive deficits in PWE also contribute to poor quality of life. Efforts to develop treatments to improve both seizure control and cognition in PWE are ongoing. This study assesses how adjunct treatment with pharmaceutical grade cannabidiol (CBD; Epidiolex) affects seizure control and attention-related cortical activity during fMRI in patients with intractable epilepsy who are part of the UAB CBD Program. Methods: 11 PWE subjects (6 female, ages 16-62, mean illness duration = 20 years) underwent 3T fMRI at 2 time points: before starting CBD treatment and after achieving a stable intermediate dose of CBD (20 mg/kg or 25 mg/kg) for at least 2 weeks. At each time point, each patient performed 2 fMRI runs of a modified Eriksen Flanker task (as recommended by NIH CDE) to assess attention and inhibitory control (Fig. 1). Each run consisted of 6 blocks with congruent (CON) stimuli and 6 blocks of incongruent (INCON) stimuli. PWE were instructed to indicate the direction the center fish was swimming for each stimulus via button press using their right hand. Standard fMRI image processing, statistical modeling and visualization were performed using the AFNI software package. General linear modeling was performed to contrast the fMRI response to CON vs. INCON conditions for each subject, while accounting for head motion, MRI signal drift, and serial correlation of noise parameters. Regression analyses were performed to assess CBD treatment changes in attention-related cortical activity during fMRI, while accounting for age, illness duration and monthly seizure frequency. Changes in seizure frequency and Flanker task accuracy were assessed using paired samples t-tests. Results: The average time between fMRI visits was 11.5 weeks. PWE exhibited overall reduced monthly seizure frequency between visits (median (range): 14 (4-186) to 8 (0-64); p=0.07). Mean task accuracy also improved for both CON (33.2% to 43.2%) and INCON (31.8% to 38.3%) stimuli following CBD treatment, but these were not statistically significant changes (p=0.13 and p=0.30, respectively). Regression analysis showed deactivation in a number of regions to CON and INCON conditions at visit 2 compared to pre-CBD fMRI (p < 0.05 uncorrected, cluster threshold 10 voxels; Fig. 2A). Regression analysis also revealed a greater difference in fMRI response to CON vs. INCON conditions in the left inferior frontal gyrus at visit 2 compared to pre-CBD fMRI (p < 0.05 uncorrected, cluster threshold 10 voxels; Fig. 2B). Conclusions: PWE experienced overall improved seizure control while on CBD, with 8/11 patients exhibiting reductions in monthly seizure frequency by over 50%. PWE also showed increased accuracy on the Flanker task that paralleled increased deactivation to CON and INCON conditions following CBD treatment, which may reflect improvement in attention and inhibitory control processes. These preliminary results in a small number of heterogeneous patients suggest the potential for positive cognitive effects of CBD that are associated with corresponding fMRI signal changes. Funding: State of Alabama