Memantine, a non-competitive NMDA receptor antagonist, has attracted interest due the potential to reduce excitotoxicity-mediated memory deficits. Using a randomized, double-blind, placebo-controlled clinical trial design, Marimuthu et al. recruited epilepsy patients on AEDs and with subjective memory complaints (9). Active arm subjects (n = 29) received memantine 5 mg daily for the first 8 weeks and 10 mg daily for the next 8 weeks, while placebo arm subjects (n = 30) received a daily placebo for the full 16 weeks. Fifty-five of fifty-nine subjects completed the study (26 active and 29 placebo arm); subjects were between 18 and 55 years old, with different seizure types (GTCs = 39, focal onset impaired awareness = 9, absence seizures = 5, and one each with myoclonic and focal onset preserved awareness seizures). Twenty-one subjects were on AED monotherapy, while thirty four were on polytherapy. Cognitive and memory function were assessed at baseline, and at the end of the 4th, 8th, 12th, and 16th weeks: subjective symptoms were assessed at the same timepoints using the QOLIE-10-P. By week 16, subjects in the active arm experienced an increase in MMSE from 18.42 at baseline to 25.35, compared to an increase from 18.31 to 22.62 in the placebo group (p < 0.001). Wechsler memory scale (WMS) score showed a statistically significant improvement from 56.35 to 67.92 by the end of week 8 (p = 0.002) and to 91.35 (p < 0.001) by the end of week 16; for comparison, placebo arm WMS scores increased from 56.52 to 70.14 by the end of week 16. Within-group analysis between the memantine 5 mg and 10 mg daily periods showed a significant improvement in both MMSE and WMS by week 16 as compared to week 8. Particular areas of improvement in active arm subjects were immediate recall (logical memory), delayed recall (memory span) and visual reproduction memory by week 8, and statistically significant increase in associate learning by the end of week 12. Active arm subjects also reported improved subjective memory compared to placebo arm (p = 0.017 by week 8, and p = < 0.000 by week 16) and overall QOL improvement (p = 0.004 by week 8, p < 0.000 in weeks 12 and 16). Of note, cognitive and memory assessment was not adjusted for epilepsy type or mood.

Given potential cognitive benefits and likely low risk of increased seizures, two studies set out to evaluate methylphenidate as a cognitive enhancer in epilepsy (10, 11). The first was a double-blind, placebo-controlled single-dose 3-period crossover study in 31 adult epilepsy patients (24 with focal, 6 with generalized, and 1 with unclassified epilepsy) with cognitive complaints on a stable AED regimen (10). Subjects completed three study visits ~1 week apart; during each visit, they received either placebo, methylphenidate 10 mg or methylphenidate 20 mg (each subject received each of the three treatments once during the study). Following a 1 h wait for medication absorption and peak effect, subjects completed a neurocognitive battery to evaluate processing speed [Symbol Digit Modalities Test (SDMT)], immediate verbal recall [Medical College of Georgia Paragraph Memory Test (MCG)], and attention-related issues [Conners Continuous Performance Test Third Edition (CPT 3)]. After either dose of methylphenidate, subjects enjoyed a statistically significant cognitive benefit, with comparable benefit noted for both doses. Particular improvement was noted in processing speed and consistency in response speed [hit reaction time standard deviation (HRTSD) subtest of CPT 3]. A trend toward improvement was noted in the go/no-go d' subtest of CPT 3. This study did not evaluate mood, or include subgroup analysis based on epilepsy type.

The authors followed up with a second 1 month open-label trial completed by 28 adult subjects recruited from the crossover study (21 with focal, 6 with generalized, and 1 with unclassified epilepsy) (11). During the open-label phase, subjects started at methylphenidate 5 mg twice daily or 10 mg twice daily, and titrated, as tolerated, toward to 20 mg twice daily (goal dose). Fourteen healthy controls completed neurocognitive testing at the same intervals as epilepsy subjects; controls did not receive methylphenidate. Efficacy, tolerability and adverse effects were assessed on a weekly basis. Following the treatment month, subjects, and healthy controls (when appropriate) repeated the same neurocognitive battery used in the previous cross-over study, while mood was evaluated with the Beck Depression Inventory, 2nd edition (BDI-II), Beck Anxiety Inventory (BAI), and Apathy Evaluation Scale (AES). The QOLIE-89 assessed quality of life in epilepsy subjects only. While there was a significant improvement in both epilepsy subjects and controls across multiple cognitive variables, epilepsy subjects enjoyed greater improvement on measures of attention, and this difference was found to be clinically significant on post-hoc analysis. Epilepsy subjects also had significant improvement in quality of life. Both groups had improvements on psychiatric measures, but with no significant difference between the treated epilepsy patients and untreated healthy controls. Mood was not controlled for in the analysis of cognitive improvement, and subgroup analysis based on epilepsy type was not included. Interestingly, scores for the stimulant side effects and the adverse events profile decreased by visit 5 from baseline, which the authors suggest may reflect a general subjective sense of improvement due to either placebo effect or improved quality of life.

Evaluation of TMS to address cognitive deficits experienced by patients with epilepsy has been limited. In their randomized, double-blinded, sham-controlled, parallel-design clinical trial evaluating antiepileptic effects of repetitive TMS (rTMS) on refractory epilepsy patients with malformations of cortical development (MCD), Fregni et al. included cognitive changes as a secondary outcome (12). Twenty-one subjects with drug-refractory, non-surgical epilepsy were randomized to receive 1 Hz rTMS treatment (n = 12) or sham treatment (n = 9), and underwent daily treatment sessions for 5 consecutive days. Two subjects were on monotherapy, and 19 on polytherapy. In the treatment arm, there was a significant reduction in seizure frequency of up to 72% (p = 0.003) at week 2; this effect attenuated with time but did persist until the last check-in at week 8 (58% seizure frequency decrease, p = 0.001). Analysis of EEG data revealed a statistically significant decrease in epileptiform discharges in active arm subjects at week 2, which largely washed out by week 8. Cognitive assessment included digit span (forwards and backwards), simple reaction time, and the Stroop test at baseline, after treatment, and 8 weeks after treatment. The 14 subjects able to complete the Stroop showed significant improvement in performance at both post-treatment time-points, but no change in either digit span (n = 14) or simple reaction time (n = 19). Subjects reported short-term increase in energy and social interaction (subjective measures of cognition) at week 2, but not at weeks 4 or 8; analysis of cognitive improvement was not controlled for changes in energy or social interaction.

Transcranial Direct Current Stimulation applies a weak electrical current to the scalp, and thus modifies neuronal membrane potentials in the underlying brain tissue, either by depolarization (anodal tDCS) or hyperpolarization (cathodal tDCS) of the resting potential. Two studies have assessed the effect of tDCS on memory in patients with epilepsy. Del Felice et al. performed a randomized sham-controlled cross-over study in 12 subjects with TLE and mesial temporal lobe sclerosis, in which partially sleep-deprived subjects were treated with slow-oscillatory tDCS (sotDCS) prior to napping to evaluate whether sotDCS affects sleep spindles and enhances memory consolidation (13). Subjects underwent neuropsychological assessment and the Montreal Cognitive Assessment (MOCA) prior to active or sham treatment, followed by a period of rest (60 min of sleep or 105 min in bed, whichever occurred first). Active treatment consisted of excitatory depolarizing slow-oscillating anodal stimulation (fluctuating current between 0 and 250 mA, 0.75 Hz frequency applied for 5 min with 1 min interblock interval, 30 min total) over the ipsilateral fronto-temporal region. Thirty minutes after awakening, verbal [Rey Auditory Verbal Learning Test (RAVLT)] and visuospatial recall (Rey-Osterrieth Complex Figure Test and Rey-Osterrieth Complex Figure B) were tested. Subjects self-administered the Beck Depression Inventory (BDI) and State Trait Anxiety Inventory (STAI 1 and 2) between the learning and recall memory testing. Subjects underwent two courses of treatment (one active and one sham), separated by at least 1 week. Investigators found changes to EEG characteristics of certain sleep spindles and increased total sleep time following active stimulation, as well as statistically significant differences on neuropsychiatric testing between the sham and active treatment: declarative memory improved following tDCS and napping, specifically with improved retention (p = 0.022), and there was a slight decrease in forgetfulness on visuospatial testing (p = 0.048). There were no significant changes in BDI, STAI 1 or STAI 2 scores.

In a 2016 double-blinded, sham-controlled, randomized, parallel-group study, Liu et al. (14) evaluated the efficacy of a 5 day tDCS course on depression and memory in 33 patients with well-controlled TLE. Subjects underwent fixed-dose 2 mA tDCS or sham treatment for 20 min every day for 5 days; the anode was placed over the left dorsolateral prefrontal cortex with the goal of increasing regional activity. Neuropsychological testing [Letters and Numbers Sequencing and Digits Span Test subtests of the Wechsler Adult Intelligent Scale-III (WAIS-III)], symptom inventories [BDI-II, Neurological Disorders Depression Inventory for Epilepsy (NDDI-E)] and a 20 min EEG were performed at baseline, prior to the first session of stimulation, after the last stimulation session, and at the 2 and 4 weeks post-stimulation follow-up sessions. The authors found no difference in the scores of any of the measures of working or verbal memory in active vs. sham arm subjects. Treatment arm subjects reported a statistically significant decrease in depressive symptoms compared to the sham arm subjects, though this difference disappeared by week 2. Interestingly, QOLIE-31 scores did not change between the active and sham arms. The study found no difference in seizure frequency, interictal discharge frequency, or the EEG spectral power in the active arm compared to baseline or the sham treatment arm.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.