Epilepsy is a neurological disorder that affects approximately 70 million people worldwide and is common in both children and adults (Ngugi et al., 2010). Despite the availability of anti-seizure medications, around one-third of patients have seizures that cannot be adequately controlled by medication alone. Recent drug development has not yielded many new solutions, with the rate of drug-resistant epilepsy remaining relatively stable for the past 30 years (Brodie, 2017). Even in cases where anti-seizure medication is effective, up to 17% of individuals experience limiting side effects from the medication (Chen et al., 2017). While epilepsy surgery is a good option for some patients with focal epilepsy, a third continue to have seizures despite surgery and it is generally not an option for those with genetic generalized epilepsy (GGE) (Baud et al., 2018). As a result, a significant portion of patients, particularly those with GGE, are left without effective treatment. Novel, and preferably non-invasive treatments, are urgently needed.

Transcranial electrical stimulation (tES) is a promising novel therapeutic approach for drug-resistant epilepsy (Yang et al., 2020; Simula et al., 2022). tES involves the application of a low-intensity electric current (typically <2 mA) to the brain via scalp electrodes. tES can be delivered using different waveforms, the most common being: (1) Direct current stimulation (tDCS), which is applied with a uniform, unidirectional current flowing from the anode to the cathode. While being an over-simplification, from results obtained from motor cortex stimulation, if the region of interest is under the anode (i.e., during anodal tDCS) it is broadly believed that this will result in a local increase of neuronal activity. Conversely, if it is under the cathode (i.e., during cathodal tDCS) it will lead to a decrease in neuronal activity (Bestmann and Walsh, 2017). Sham stimulation, in which the current is ramped up at the same rate as tDCS but then quickly turned off, is typically used as the control condition in investigations (Bestmann and Walsh, 2017).

Previous pre-clinical and clinical studies have shown that tDCS can be effective in reducing interictal epileptiform discharges (IEDs) and seizures in individuals affected by epilepsy (San-Juan et al., 2017; Simula et al., 2022). A recent systematic review of the use of tDCS in epilepsy demonstrated that tDCS in epilepsy is safe and led to a relevant seizure reduction in most clinical studies, though results varied greatly due to different stimulation paradigms (Simula et al., 2022). So far there exists only one double-blind, randomized, sham-controlled trial and almost all studies have been done in patients with focal epilepsy (Yang et al., 2020). Data on the application of tES in GGE is very limited, and to date it has been found to be ineffective (San-Juan et al., 2016). A recent meta-analysis of clinically established neurostimulation techniques such as vagus nerve and deep brain stimulation has shown a significant effect on seizure frequency in GGE (Haneef and Skrehot, 2023). While the underlying mechanisms of these techniques differ, they provide encouraging evidence to further investigate the use of tES in GGE. Current evidence from in-vitro and human studies assessing functional connectivity and using computational models indicate that the effects of tES are mainly achieved through the modulation of large brain networks, instead of focal brain activity (Simula et al., 2022). One possible target in patients with GGE may be the sensorimotor network, which has been shown to have greater network synchrony in the minute before epileptiform discharge onset (Tangwiriyasakul et al., 2018), in comparison to their healthy relatives (Tangwiriyasakul et al., 2019). This network has also been a frequent target of tES in studies outside of epilepsy, which provide existing protocols to build from (Violante et al., 2017; Mencarelli et al., 2020).

The primary objective of this study was to establish the feasibility and safety of using tDCS during fMRI in both healthy participants and patients with Juvenile Myoclonic Epilepsy (JME), a subtype of GGE (Hirsch et al., 2022). The secondary objective was to investigate the acute changes in brain connectivity within the sensorimotor network in both groups. Our hypotheses were that tES would (a) be low-risk and tolerable in both groups and (b) lead to altered connectivity in the sensorimotor network. To test these hypotheses, we applied an established protocol and analyzed network connectivity using measures of degree centrality to determine if network modulation might be feasibly measured via this approach.

We have met our primary objective regarding feasibility and tolerability: Anodal and cathodal tDCS were applied in healthy controls and three patients with GGE during fMRI without adverse events and were well tolerated by the subjects. In contrast to our results, a case study suggested a potential health risk of using tES in patients with GGE (San-Juan et al., 2016). We have found that phosphenes were routinely reported during tACS in our healthy controls. Computational head models of similar montages to the one used in our study have been reported in the literature and shown that the electric fields pass through the eye (Iacono et al., 2015). The electrode in the supraorbital area in our montage was close enough to the eye to allow for current to reach the retinas and induce phosphenes. Even montages with electrodes placed only on the occipital cortices are known to induce phosphenes (Lorenz et al., 2019). As explained in the Methods section we decided against using tACS in our patient group due to the potential health risks. Although tACS was not applied in the patient group in this study, owing to the potential risk of inducing seizures in photosensitive epilepsy patients, it has been shown to be an effective means to alter connectivity (Lang et al., 2019; Klink et al., 2020). In this context further investigation of tACS for this purpose should be considered further, utilizing a stimulation montage that can better target the motor network while avoiding stimulation of the visual cortex. Establishing the feasibility of applying tES simultaneously with fMRI in patients with GGE enables the investigation of changes in network connectivity caused by tDCS. This could have a potential therapeutic impact since network changes have been shown to reduce markers of epileptogenicity (Simula et al., 2022). At the same time data from recent years has strengthened the hypothesis that epilepsy is a network disorder (Bartolomei et al., 2017). tES has demonstrated the potential to reduce IEDs and seizures in patients with drug-resistant epilepsy through modulating large-scale brain networks (Simula et al., 2022). In our small sample, no IEDs were present in the EEG recordings prior to tDCS.

Regarding our secondary objective, we have shown that all healthy controls and patients had a significant decrease in degree centrality through anodal and cathodal tDCS. Patients with GGE had an overall lower degree centrality in the sensorimotor cortex than controls, though this could not be statistically assessed in such a small sample, and may be confounded by inter-scanner variability. Two recent systematic reviews on the use of tES in epilepsy reported only one case report in patients with GGE (Sudbrack-Oliveira et al., 2021), it is therefore difficult to compare our findings to preexisting literature. Studies using tES in other types of epilepsy with diffuse onset like Lennox–Gastaut or Rassmussen encephalitis have shown a significant seizure reduction (Auvichayapat et al., 2013; Tekturk et al., 2016). Currently, available studies using tES in epilepsy are overall highly heterogeneous regarding sample characteristics and methodology. For this reason, it is argued that conducting a meta-analysis would create biased effect sizes and estimations (Sudbrack-Oliveira et al., 2021), to date no meta-analysis exists.

Studies investigating tES during fMRI on brain networks of healthy controls lack consensus regarding its efficacy in modulating network function (Ghobadi-Azbari et al., 2021). Looking at individual studies, one study targeting the sensorimotor network in healthy controls through applying cathodal tDCS of 1 mA for 5 min, using the same montage as in our study, resulted in decreased activation in the sensorimotor cortex (Baudewig et al., 2001), concurring with our finding. A study investigating numerous brain networks including the sensorimotor network after the application of 5-min stimulation periods at 2 mA, found connectivity near the applied field and also with remote nodes decreased during tDCS (Leaver et al., 2022). In our research, a similar finding was achieved in patients despite potential differences owing to pathology and medication. This demonstrates network modulation with tES is feasible. Conversely, it has been shown that stimulation for 20 s at 1 mA did not produce a detectable BOLD signal change (Antal et al., 2011). These variable results can among other factors be explained by differences in anatomy (i.e., scalp and skull thickness), placement of electrodes and current intensity (Liu et al., 2018) and analysis approaches. The network effects of tDCS are also dependent on brain state, with cathodal tDCS having greater effects during a task while anodal tDCS has greater effects during rest (Li et al., 2019). Epileptic brain activity, both seizures and IEDs, are often more prevalent during certain states of arousal such as sleep in both focal epilepsy and in GGE (Bernard et al., 2023). This shows that the probability of epileptic activity is modulated by the global state of the brain which relates to cortical excitability.

There is evidence for significant clinical benefit in GGE from the use of VNS and DBS (Haneef and Skrehot, 2023), but a downside of these techniques is their invasiveness, both needing surgery, making non-invasive approaches like tES attractive alternatives if efficacy can be established. Additionally, electric stimulation-driven, non-invasive approaches such as temporal interference have also been shown to reduce epileptiform activity in mouse models and it would be beneficial to analyse how temporal interference affects sensorimotor connectivity with a paradigm like ours (Acerbo et al., 2022). Our preliminary evidence of reduced mean degree centrality of the sensorimotor network supports previous literature about the modulatory effect of tES on the brain. Further confirming our hypothesis, the sham condition was not significantly different from the rest condition and cathodal stimulation significantly reduced the mean degree centrality of the nodes of the sensorimotor network, indicating reduced excitability. In line with previous studies anodal tDCS showed the same results as cathodal tDCS in reducing synchrony (Li et al., 2019; Kurtin et al., 2021).

One key limitation of our research is the sample size of participants. Recruiting patients with JME was cut short by the global COVID-19 pandemic, though the effects of tES on this small group were still powerful enough to produce statistically significant results. While there is a statistically significant difference between the stimulation conditions, a larger group size would add further power and validity to these findings. A further limitation is that the healthy controls were scanned in a different location to the patients, though both were scanned at 3 T. The small sample size and different scanner types precluded conducting a meaningful group comparison because the differences in baseline might be due to either scanner or population. Parameters were matched, however, and the same overall main trend of reduced sensorimotor cortex connectivity was observed within subjects between stimulation conditions in both healthy controls and patients which is not affected by scanner type. Another potential limitation is the effect on SNR. TES has been shown to affect image quality only minimally, with a minor effect on image SNR (Antal et al., 2011). One further limitation is the intake of different anti-seizure medications (ASM) by the patients. Due to the small number included here we could not perform a statistical analysis to account for possible pharmacological effects. Again, although this factor might change overall network connectivity in individuals, the directional reduction in sensorimotor degree centrality between conditions is likely to exist regardless of medication.

This study provides initial evidence that tES can be safely applied during fMRI in patients with JME. Here, we have also demonstrated sensorimotor network alterations in mean degree centrality that was used as a measure of network connectivity related to overall network synchrony. This preliminary finding appeared to be unrelated to the polarity of the applied stimulation. Further work is required to determine the reliability of this finding in a larger cohort, understand the interaction between current distribution and individual brain structures and establish if the modulation of motor network synchrony can modulate epileptogenicity.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by London – West London and GTAC & Health Research Authority and Health and Care Research Wales (HCRW): REC reference: 19/LO/1668. The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

ZC: Formal analysis, Writing – original draft. MS: Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Writing – original draft. RP: Formal analysis, Writing – review & editing. CT: Writing – review & editing. MR: Conceptualization, Writing – review & editing. DS: Writing – review & editing. IV: Conceptualization, Data curation, Funding acquisition, Methodology, Project administration, Writing – review & editing. DC: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Writing – review & editing.