Transcranial Stimulation During Sleep to Improve Cognition in Epilepsy

Aim 1: Determine whether sleep enhances learning across a range of cognitive domains in healthy subjects. Aim 2: Determine whether low-frequency transcranial stimulation (TCS) delivered during slow wave sleep (SWS), compared to sham stimulation, enhances learning outcomes as indexed by a complete neuropsychological battery of tests in epilepsy patients and healthy control subjects. Aim 3: Determine whether low-frequency TCS delivered during SWS, compared to sham stimulation, enhances sleep architecture associated with enhanced memory consolidation (ie. increased coherence of slow wave activity and increased frequency of sleep spindles). Aim 4. Determine whether low-frequency TCS during sleep results in a more distributed memory representation, as suggested by increased hippocampal-perirhinal connectivity on fMRI in human subjects. Aim 5. Determine whether the frequency of interictal activity during sleep in epilepsy subjects is associated with the degree of cognitive benefit conferred by SWS. These studies will provide critical pilot data on whether non-invasive brain stimulation protocols previously tested in healthy subjects can be extended to epilepsy patients for potentially therapeutic cognitive benefits.

Epilepsy and cognitive dysfunction Recent investigations suggest that more than 60% of patients with epilepsy suffer cognitive impairment, such as memory, attentional and executive dysfunction (Elger et al., 2004, Lin et al., 2012). The severity and profile of cognitive impairment are heterogeneous and depend on the epilepsy syndrome, age of onset, seizure control, frequency of interictal activity, structural brain abnormalities, and antiepileptic medication use (Elger et al., 2004, Helmstaedter and Witt, 2012, Zeman et al., 2012). Impairment in declarative memory is the most common cognitive complaint in patients with temporal lobe epilepsy and other localization-related epilepsies (Oyegbile et al., 2004, Hoppe et al., 2007, Helmstaedter and Witt, 2012). Three types of memory deficits associated with epilepsy have been described: transient epileptic amnesia, accelerated long-term forgetting and remote memory loss (Butler and Zeman, 2008, Zeman et al., 2012). In particular, accelerated long-term forgetting - defined as a difficulty in retaining memories that were initially able to be acquired - has been recently reported to be more prevalent than previously recognized (Witt et al., 2012). Current therapeutic approaches include management of antiepileptic drugs, monitoring for side- effects, and decreasing seizure and interictal frequency. However, even when optimized, treatments have shown only limited effectiveness and do not modulate the underlying pathophysiology (Meador, 2011, Lin et al., 2012). Indeed, despite the prevalence of cognitive comorbidity among epilepsy patients, no effective therapies exist. Sleep-related memory consolidation There are two major steps in learning: 1) attending to and encoding the information and 2) stabilizing or consolidating the transiently coded information (Squire, 1992). These steps involve different physiological mechanisms, different brain states and different types of interactions among hippocampal-neocortical circuits. Acquiring new information can be affected in either the acquisition or consolidation phase. Therefore, strategies to enhance memory formation can target either of these steps. Sleep, increasingly recognized for its critical role in memory consolidation, provides an intriguing opportunity for therapeutic intervention. Memory consolidation, the second stage in the learning process, is the activity of transforming short-term memory traces into stable long-term representations (Dudai, 2012). One of the predominant views is that consolidation occurs via hippocampal-neocortical interaction during the brain's 'offline' state of sleep (Buzsaki, 1989; Diekelmann and Born, 2010). The key electrophysiological pattern responsible for this interaction is the hippocampal sharp wave ripple (SPW-R), during which the spatio-temporal patterns of neuronal firing associated with preceded learning are replayed in compressed time scales (O'Neill et al., 2010). Learning affects both the incidence and neuronal contents of SPW-Rs in subsequent sleep. Conversely, modification of SPW-Rs by various means affects the consolidation process. For example, selective elimination of SPW-Rs in rodents impairs learning as dramatically as surgical removal of the hippocampus, even though such experimental manipulation leaves other aspects of post-learning sleep unaltered (Girardeau et al., 2009). In addition to SPW-Rs, two other SWS-related oscillations have been implicated in memory consolidation: the amplitude and coherence of slow oscillations and sleep spindle density are increased when SWS is preceded by hippocampally-dependent declarative memory tasks (Gais et al., 2002, Molle et al., 2004) or procedural skill learning (Huber et al., 2004). SPW-Rs are modulated by sleep spindles (12-16 Hz), which are in turn orchestrated by slow (0.5-1.5 Hz) and ultraslow (0.1 Hz) oscillations (Steriade 1993, Sirota 2003, Isomura 2006, Molle 2006, Sullivan 2011). Such cross-frequency coupling suggests that a complex and distributed neuronal system coordinates neocortical-entorhinal cortex (EC)-hippocampal activity. Recently, it has been demonstrated that low frequency (0.01 - 0.1 Hz) blood-oxygen-level-dependent (BOLD) fMRI correlations can be enhanced during post-encoding waking rest periods relative to baseline rest periods with the inclusion of some forms of learning 7-11. Furthermore, we have demonstrated, across participants, that the magnitude of changes in hippocampal-cortical resting connectivity is related to subsequent associative memory performance 10. These spatial and temporal relationships provide a remarkable opportunity to influence hippocampal activity, and thereby memory consolidation, by affecting neocortical slow oscillations and/or thalamocortical sleep spindles (Ozen et al., 2010). Specifically, cortical rhythms can be biased by non-invasive stimulation or behavioral interventions, thereby potentially enhancing or interfering with learning. Indeed, several studies have demonstrated that SWS-related electrical patterns can be enhanced in healthy subjects by non-invasive brain stimulation techniques, such as TCS, applied during early sleep (Marshall et al., 2004, Marshall et al., 2006, Massimini et al., 2007). Specifically, low delta frequency (0.75 Hz) TCS during early SWS sleep increased endogenous oscillations and post-sleep gain of word-pair association memory, a hippocampal-dependent declarative memory task (Marshall et al., 2006, Marshall et al., 2011). Modulation of SWS, via TCS, may enhance learning outcomes across a range of cognitive domains for patients with epilepsy and other neurologic disorders. TCS: Background and safety issues. TCS is a method of noninvasive brain stimulation which delivers low amplitude current through scalp electrodes. Current parameters may be altered to deliver current in a unidirectional [ie. transcranial direct current stimulation (TDCS)], alternating [ie. transcranial alternating current stimulation (TACS)], or random manner [ie. transcranial random noise stimulation (TRNS)], to affect endogenous brain oscillations in different directions. Decades of research in both humans and animal models have shown that TCS can modulate brain activity - both to enhance and reduce cortical excitability (reviewed in Priori, 2003). Advantages of TCS include low cost, ease of administration, safety profile, and its noninvasive and painless nature. TCS is based on the application of a weak direct current to the scalp. Low amplitude (<2 mA) currents are applied via the scalp electrodes and penetrate the skull to enter the brain. Although there is substantial shunting of current in the scalp, sufficient current penetrates the brain to modify the trans-membrane neuronal potential as shown by two recent modeling studies (Miranda et al., 2006, Wagner et al., 2007), and thus, influence the level of excitability and modulate the firing rate of individual neurons. When TCS is applied for a sufficient duration, cortical function can be altered beyond the stimulation period (Nitsche and Paulus, 2001) and the direction of the cortical excitability changes depends on current orientation. Several well-conducted animal studies on the effects of TCS dating back to the 1950s and 60s demonstrate its ability to modulate brain function. These studies demonstrated that polarizing currents applied to the surface of the brain result in a modulation of the cortical activity. Surface anodal polarization of the cortex increases spontaneous unit discharges (Burns, 1954, Creutzfeldt et al., 1962) and initiates paroxysmal activity (Goldring and O'Leary, 1951), whereas cathodal polarization generally depresses these events. Low-level surface polarization was also shown to facilitate the acquisition of learned motor responses and to induce prolonged changes in patterns of evoked cortical unit discharges (Bindman et al., 1964). Furthermore, Purpura et al. (1965), studying pyramidal tract cells from cats, showed that prolonged periods of polarization may produce progressive membrane and post-synaptic potential changes as well as after-effects (Purpura and McMurtry, 1965). More recently, extracellular and intracellular studies in rats have shown that TCS can reliably entrain neurons in widespread cortical areas, including the hippocampus (Ozen et al, 2010). TCS offers several advantages as compared with other techniques of noninvasive brain stimulation, such as repetitive transcranial magnetic stimulation (rTMS). These include: (1) small size of the electrodes and stimulator, thus allowing portable use, (2) simple and non-expensive technique that can easily be translated for use in clinical practice, (3) durable effects - the modulatory effects of TCS last longer as compared to rTMS [e.g., 13 minutes of TDCS can change brain excitability for up to 2 hours (Nitsche and Paulus, 2001)], (4) more easily blinded in the setting of clinical trials (Gandiga et al., 2006), and (5) well-established safety profile (Liebetanz et al., 2009).

We will apply oscillating current at a slow frequency of 0.5-1.5 Hz during early sleep, which is rich in slow waves [ie non-REM sleep]. The peak intensity of stimulation which allows for optimal phase entrainment will be determined in pilot studies. However, peak stimulation intensities will not exceed 2 mA (as discussed above). The current will be applied over the left and right prefrontal cortex (F3, F4), corresponding to the predominant region of slow oscillations, during the onset of deep sleep to the first REM episode (early non-REM-rich sleep).

The EEG and stimulation electrodes will be placed as in the stimulation sessions, but stimulation will not be administered.

A non invasive brain stimulation which delivers low amplitude current through scalp electrodes by a Transcranial Stimulator

RAVLT, Yellow Cab Navigation Task, Working Memory tasks, Procedural Memory tasks, Insight based tasks.

Inclusion Criteria: Age between 18-70 years Must score 22 or more on the MoCA. Must be able to provide informed consent. Must be fluent in English. Diagnosis of focal epilepsy Exclusion Criteria: Patient has a progressive or unstable neurological or systemic disease Patient has a history of severe traumatic brain injury or prior brain surgery with skull defect Contraindictations to TCS, including metal in the head or implanted brain medical devices Pregnancy Any implanted electrical medical device, including pacers and implanted cardiac defibrillators History of schizophrenia, schizoaffective disorder, other psychosis, rapid-cycling bipolar illness, alcohol/drug abuse within the past year History of dementia History of known sleep disorder Additional Exclusion Criteria Ictal Focus over the F3 or F4 field Clinical or electrographic evidence of frequent nocturnal seizures, as determined with recent (<2 year) ambulatory EEG. Generalized epilepsy Epilepsy subjects will be identified and consented from the NYU Comprehensive Epilepsy Center. In addition to the inclusion and exclusion criteria above, epilepsy subjects must meet the following criteria: