Personalized multichannel transcranial direct current electrical stimulation (tDCS) in drug‐resistant epilepsy: A SEEG based open‐labeled study

Abstract

Objective

To evaluate the effects of personalized multichannel tDCS on seizure frequency, severity, quality of life, and psychiatric comorbidities in patients with drug‐resistant focal epilepsy. Secondary goals include assessing the safety and feasibility of this approach.

Methods

This open‐label pilot study involved 16 patients with drug‐resistant focal epilepsy. Patients underwent 3 cycles of personalized multichannel tDCS over 6 months, targeting the EZ defined by stereoelectroencephalography (SEEG). Each cycle consisted of five consecutive days of tDCS, with two daily sessions of 20 min each. The primary endpoint was a reduction in seizure frequency, with secondary endpoints addressing quality of life (QOLIE‐31 scores), seizure severity (NHS3 scores), and psychiatric comorbidities (NDDI‐E and GAD‐7 scales).

Results

Across all participants, a statistically significant 20% reduction in seizure frequency was observed (p = 0.044). Six patients (37%) were identified as responders (≥50% seizure reduction), with one achieving seizure freedom. The mean seizure reduction among responders was 68%. Significant improvements were noted in overall quality of life (QOLIE‐31, p = 0.009), with greater benefits for patients with poorer baseline scores. No overall significant changes were observed in depression, anxiety, and seizure severity scores, though individual variability was noted. The treatment was well tolerated, with mild adverse events, primarily skin‐related.

Significance

Personalized multichannel tDCS shows promise as a noninvasive therapeutic option for drug‐resistant focal epilepsy, with benefits in seizure reduction and quality of life. Although results were variable, the method's safety and feasibility support further exploration through randomized controlled trials to refine protocols, better select potential responders' patients, and validate findings.

Plain Language Summary

This study tested a personalized brain stimulation technique called tDCS in people with difficult‐to‐treat epilepsy. The treatment led to fewer seizures in some patients and improved their quality of life. The approach was safe and caused only mild side effects. These results suggest that this type of noninvasive brain stimulation may be a helpful new option for people who do not benefit from medication or surgery.

Key points.

• Personalized multichannel tDCS reduced seizure frequency by 20% overall; 37% of patients were responders, with one seizure‐free case.

• Significant improvements in quality of life (QOLIE‐31), particularly in patients with lower baseline scores, were observed.

• The study confirmed the safety of personalized tDCS, with mostly mild adverse events, including skin‐related issues and headaches.

• SEEG‐based targeting improved the precision of tDCS, emphasizing the potential for individualized epilepsy treatment.

1. INTRODUCTION

Epilepsy is one of the most prevalent chronic neurological conditions, affecting about 70 million people worldwide. ¹ Approximately, one‐third of them experience seizures that do not respond to pharmacological treatment. ² Drug‐resistant epilepsy poses severe challenges, significantly impairing quality of life. Surgical intervention to remove the epileptogenic zone (EZ) is often not feasible, especially for patients with multifocal or diffuse disease or those with an EZ that overlaps the eloquent cortex. ³ Therefore, alternative treatments that use noninvasive brain stimulation techniques have attracted attention as promising methods to decrease cortical excitability.

Transcranial direct current stimulation (tDCS) is a noninvasive technique for brain stimulation that delivers a low electrical current (typically 1–2 mA) through the cerebral cortex via at least two electrodes placed on the scalp. This method modifies cortical excitability painlessly and reversibly. It has been shown that the effects of tDCS depend on the polarity of the electrodes: the anodal electrode enhances cortical activity, while the cathodal electrode reduces excitability. ⁴ , ⁵

Several human and animal studies have shown the efficacy of cathodal tDCS in reducing epileptiform discharges by inhibiting epileptic brain regions. ⁶ , ⁷ , ⁸ Sham‐controlled randomized studies ⁷ , ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ have highlighted the safety and potential utility of tDCS in epilepsy, showing encouraging results in reducing epileptiform discharges. However, these studies utilized simple tDCS configurations (one anode and one cathode) that lacked precision and personalization to the specific epileptogenic networks. ¹⁴

Recently, two pilot studies have explored personalized multichannel and multisession tDCS protocols. In the study by Kaye et al., ¹⁵ each participant underwent 10 sessions of 20 min of inhibitory tDCS over 2 weeks. Multi‐electrode montages were designed using realistic head model‐driven approaches to deliver an inhibitory electric field to the targeted cortical EZ and surrounding cortex, reducing excitability and seizure rates. A median seizure reduction of 44% relative to baseline was observed, with no significant differences between adults and children. Daoud et al. ¹⁶ reported the effects of personalized multichannel tDCS in 10 patients with various types of epilepsy and prior SEEG recordings. The multichannel tDCS (Starstim, Neuroelectrics) was administered during three cycles, with one cycle every 2 months. Each cycle consisted of five consecutive days, during which patients received two daily tDCS sessions of 20 min each, separated by a 20‐min break. The electrode montages were personalized to target each patient's EZ, as defined by SEEG recordings. After the final tDCS session, five patients (50%) experienced a seizure frequency reduction of 50% or more compared with baseline.

However, these studies did not assess the impact of tDCS on quality of life or psychiatric comorbidities. Therefore, we proposed a pilot study to evaluate the effects of a patient‐specific multichannel tDCS protocol on seizure frequency, quality of life, and psychiatric comorbidities in a population of patients with focal epilepsy. This study included patients considered inoperable or those who have failed prior surgical interventions, with clearly identified epileptogenic networks based on SEEG recordings.

2. MATERIALS AND METHODS

2.1. Study design

The study is a prospective, open‐label, monocenter pilot trial designed to evaluate the therapeutic and neurophysiological effects of personalized multichannel transcranial direct current stimulation (tDCS) in patients with drug‐resistant focal epilepsy. The protocol (N°ID‐RCB 2020‐A02861‐38, NCT04782869) obtains approval from the ethics committee. The study was conducted over 8 months and involved participants who acted as their own controls. Eligible patients are screened and observed during a 2‐month baseline, during which seizure frequency and characteristics were recorded. Following baseline, participants underwent three cycles of treatment, each consisting of five consecutive days of tDCS, administered every 2 months. Each session included two 20‐min periods of cathodal stimulation separated by a 20‐min break. Personalized stimulation parameters were based on pre‐treatment stereoelectroencephalography (SEEG) data, with cathodal electrodes targeting the EZ and anodal electrodes placed over regions non‐involved during seizures. The positions and currents of the electrodes were determined via personalized biophysical head models combined with dose parameter optimization algorithms (Stimweaver algorithm, see ¹⁵ , ¹⁶ , ¹⁷ for a more detailed explanation). Anodal electrodes were positioned over non‐involved cortical regions, as identified by SEEG and optimized via the Stimweaver algorithm to minimize current density in epileptogenic tissue while maximizing the targeted electric field over the EZ.

The primary objective of this study was to evaluate the effectiveness of personalized multichannel transcranial direct current stimulation (tDCS) in reducing seizure frequency and severity in patients with drug‐resistant focal epilepsy. The study aimed to demonstrate a significant reduction in seizure frequency after each tDCS treatment session compared with the baseline period, thereby establishing the therapeutic potential of this technique. Secondary objectives aimed to measure improvements in patients' quality of life using the QOLIE‐31 questionnaire and evaluate changes in psychiatric comorbidities, such as depression and anxiety, through the NDDI‐E and GAD‐7 scales. Modifications in seizure severity and semiology were analyzed using the NHS3 scale, ¹⁸ particularly focusing on disabling seizures. The study sought to identify the proportion of patients achieving a reduction in seizure frequency greater than 50% and determine the number of seizure‐free patients. Lastly, the safety and tolerability of the tDCS protocol were closely monitored, with all adverse events recorded to ensure the method's feasibility and patient well‐being.

2.2. Patients' selection

To be included, patients had to be aged 12 years or older and have a diagnosis of focal drug‐resistant epilepsy that was either inoperable or had not responded to prior treatments such as vagus nerve stimulation or deep brain stimulation. All patients had previously undergone SEEG to define their EZs. After SEEG, patients did not undergo surgery for functional reasons (EZ affecting an eloquent cortex) or refused the operation. Some failed the operation. Patients with bilateral or multifocal EZs were not operated on. The Table S1 describes the main characteristics of the patients. A minimum of three seizures per month over at least 3 months before the baseline period was required, along with prior SEEG to define the EZ. All patients had a stable antiseizure medication regimen for at least 1 month before baseline monitoring and remained unchanged throughout the study. A clinical or research MRI suitable for navigated brain stimulation and electrical field modeling had to be available, and participants needed to have a stable antiseizure medication regimen before and during the study. Additional requirements include a total IQ greater than 65, proficiency in French, written informed consent from the patient or legal representative, and affiliation with a health insurance plan. For women of childbearing potential, a negative urinary pregnancy test is mandatory. Exclusion criteria included generalized epilepsy, serious intercurrent pathology, progressive brain tumors, contraindications to MRI, cranial metal implants, pacemakers, cochlear implants, or other devices incompatible with tDCS, as well as skull defects exceeding 5 mm in radius from prior surgeries. Patients experiencing fewer than three seizures per month, those with scalp conditions or severe headaches, pregnant or breastfeeding women, and individuals unable to understand or comply with study requirements due to cognitive or language barriers were excluded. Furthermore, individuals deprived of liberty by judicial decisions, those participating in another clinical trial within the previous month, and any condition deemed unsuitable by the investigator were also grounds for exclusion. Concomitant use of stable‐dose antidepressant or antipsychotic medications was permitted in patients with comorbid psychiatric conditions, and all such medications remained unchanged throughout the study duration.

2.3. Statistical analysis

All safety and efficacy variables have been summarized descriptively. Continuous variables have been described as means (standard deviation) or medians (quartiles) according to their distribution. Categorical variables have been described as frequency distributions (percentages). Comparisons used parametric or non‐parametric tests according to the variables' nature and distributions. Unless otherwise specified, statistical significance was defined as p < 0.05. Statistical analyses have been performed using RStudio (version 4.2.2). Demographic and baseline characteristics data have been summarized for all patients in the sample. Medical history findings, previous and concomitant medications, adverse events (AEs), and other pertinent information have also been summarized. The change (expressed in %, relative change) of the number of seizures (mean number per month) during the 8 weeks following the last cycle of tDCS (V4) has been compared with the 2‐month baseline period using a Wilcoxon signed‐rank test (paired). Categorical variables (severity of epilepsy, quality of life, scores of depression and anxiety) as estimated at V4 have been compared with those estimated at V1 using the Friedman test and/or Wilcoxon signed‐rank test. The number of responders (>50% of seizure reduction) and seizure‐free patients has been calculated.

3. RESULTS

3.1. Patients' characteristics

Seventeen patients were included. One patient (P8) did not undergo the protocol for other medical reasons. Finally, 16 patients (10 female, six male) were included in the analysis. The patients' mean age was 29.6 years [12–53]. The main clinical characteristics have been reported in Table 1. The etiology of epilepsy was unknown in 7 cases, a malformation of cortical development in 4 cases, hippocampal sclerosis in 1 case, and other etiologies in 5 cases. The epileptogenic zone (EZ) localization data per participant is now provided in Table S1.

TABLE 1.

Clinical characteristics of the study cohort.

	Overall	Non‐responder	Responder
	N = 16	N = 10	N = 6
Sex
Female	10 (63%)	5 (50%)	5 (83%)
Male	6 (38%)	5 (50%)	1 (17%)
Age at Epilepsy onset	9.5 (8, 0–26)	12.8 (7.8, 5–26)	4.0 (4.8, 0–10)
Age	29.6 (12.9, 12–53)	36.1 (11.2, 23–53)	18.8 (7, 12–31)
ASM number (mean, range)	8.8 (4.2, 3–17)	9.0 (4.5, 3–17)	8.3 (3.9, 4–15)

Note: This table presents the demographic and clinical characteristics of the 16 patients included in the study. Patients are categorized into responders (≥50% seizure reduction) and non‐responders. Variables include sex distribution, age at epilepsy onset, current age, and the number of anti‐seizure medications (ASMs) taken. Data are expressed as means with standard deviations and ranges for continuous variables and as frequencies (percentages) for categorical variables.

3.2. Effect on seizure frequency and severity

Six patients (37%) were responders, with more than 50% seizure reduction after the three cycles of tDCS, including one patient who became seizure‐free (Figure 1A). The average seizure reduction among responder patients was 68%. We noted varying individual responses in terms of seizure reduction (Figure 1B,C). Overall, the average reduction in seizure frequency from baseline to V4 was 20%. A Wilcoxon signed‐rank test indicated a significant decrease in seizure frequency between baseline and V4 (V = 107, p = 0.044).

FIGURE 1.

Impact of tDCS on seizure frequency. (A) Individual changes in seizure frequency (V4 compared with baseline). Note that six patients are responders. (B) Seizure frequency (expressed as the logarithm of the 28‐day average) at baseline, V2, V3, and V4 visits during the protocol. (C) Individual 28‐day seizure frequency (expressed in log scale) from baseline to the V4 visit.

Comparison between responders (n = 6) and non‐responders (n = 10) was performed using monovariate analysis. We found that responders were younger (median NR: 31.5 (range 23–53); median R: 16 (range 12–31), Wilcoxon p = 0.009) and had essentially an extratemporal EZ (R: 100%; NR: 40%; Fisher's exact test p = 0.03).

3.3. Impact on quality of life and other questionnaire measures

There was a significant improvement in the QOLIE global score from baseline (median = 47) to V4 (median = 58, Wilcoxon signed‐rank test: V = 19, p = 0.009). There was no significant difference in the change in the global QOLIE score (from baseline to V4) between responders and non‐responders (W = 32, p‐value = 0.8707). Regarding the different items of the QOLIE 31 testing, the items with significant increases between V4 and baseline were “overall quality of life” and “cognitive functioning” (p < 0.05, Table 2). The only item for which the V4 score was lower than the baseline score is “energy/fatigue.”

TABLE 2.

Quality of life in epilepsy (QOLIE‐31) results.

QOLIE items	Baseline median	V4 median	Wilcoxon signed‐rank test
Seizure worry	4.05	4.95	NS
Overall quality of life	7.53	9.1	p < 0.05
Emotional well‐being	7.8	9.0	NS
Energy/fatigue	5.1	4.8	NS
Cognitive functioning	14.21	16.95	p < 0.05
Medication effects	1.0	1.46	NS
Social functioning	5.04	9.03	NS
Overall score	47.23	57.7	p < 0.01

Note: This table summarizes the changes in QOLIE‐31 scores between baseline and the final evaluation (V4). Items assessed include overall quality of life, cognitive functioning, and energy/fatigue. Statistical significance for changes in specific domains is indicated.

A Pearson correlation analysis revealed no significant relationship between changes in the global QOLIE score and seizure frequency at baseline, V4, or in their change. As expected, strong correlations were observed between baseline and V4 QOLIE scores on the one hand, and between baseline and V4 seizure frequency on the other hand. Notably, a significant negative correlation indicated that patients with lower baseline QOLIE scores experienced greater improvements, emphasizing the impact of initial quality of life on treatment outcomes (Figure 2).

FIGURE 2.

Relationship between baseline QOLIE scores and changes in QOLIE scores between baseline and V4. This figure explores the correlation between initial QOLIE‐31 scores and subsequent improvements. Patients with lower baseline scores demonstrated greater enhancements in quality of life at V4.

Regarding the psychiatric comorbidities, NDDI‐E (depression) and GAD‐7 (anxiety) scores showed variable individual responses, with no overall significant changes from baseline to V4.

Regarding the impact on seizure severity, the NHS3 scales did not show differences between V4 and V1 scores (V = 81, p‐value = 0.2423). Responder patients had a non‐significant tendency for decreasing seizure severity (V = 17.5, p‐value = 0.1718).

3.4. Safety assessment and side effects

A total of 192 adverse events were recorded (85% mild and 15% moderate). And 48% involved the skin (pruritus 27%, paresthesia 19%), then 20% fatigue and 16% headaches. Most (60%) of the moderate adverse events were seizures. A total of seven patients had seizures during the tDCS sessions (five during stimulation, two during the break between the two sessions). While seizures are expected in this population, we classified seizures occurring during stimulation cycles as adverse events when their frequency or severity increased in close temporal proximity to tDCS.

On average, each patient presented 12 adverse events [min: 6; max: 21]. In five patients (31%), seizure frequency increased between baseline and V4 by an average of 41% [19%; 78%] (Figure 1A).

4. DISCUSSION

This open‐label study demonstrated that personalized, multichannel tDCS effectively reduced seizure frequency in certain patients with drug‐resistant focal epilepsy after three cycles of treatment and a 6‐month follow‐up. Among the 16 patients analyzed, six (37%) were classified as responders (≥50% seizure reduction), showing an average seizure reduction of 68%, and one patient achieved complete seizure freedom. Furthermore, notable enhancements were seen in the quality of life (QOLIE‐31 scores), especially in aspects like overall quality of life and cognitive functioning. There was a tendency to lessen the severity of seizures.

4.1. Therapeutic effect

Building on our previous feasibility work, ¹⁶ the present study extends the protocol to a broader cohort and incorporates multidimensional outcome measures (QOLIE‐31, NDDI‐E, GAD‐7, NHS3), providing a more comprehensive evaluation of personalized multichannel tDCS. The number and spacing of stimulation sessions were based on a prior pilot study, ¹⁶ but the optimal dosing schedule for multichannel tDCS remains to be determined. Previous studies have shown the importance of repeating the tDCS sessions to achieve a cumulative efficacity. ¹³ , ¹⁹

The study confirmed the therapeutic efficacy of tDCS, with 37% of participants achieving a clinically meaningful response (seizure reduction of at least 50%) and a mean seizure reduction of 20%. This rate of seizure reduction was inferior to previous studies. Indeed, Simula et al. ¹⁴ conducted a systematic review of 17 clinical trials and 6 case studies examining the effects of cathodal tDCS in patients with drug‐resistant epilepsy. Most of these studies reported significant reductions in seizure frequency, with several trials showing decreases of over 40%. Two previous open‐label studies have used a personalized multichannel protocol. Kaye et al. ¹⁵ reported a median seizure reduction of 44% after 10 daily sessions of cathodal tDCS, supporting the efficacy of repeated, spatially optimized stimulation protocols. Daoud et al. ¹⁶ reported a 48% seizure reduction in 10 patients receiving the protocol described in the present study. Given the great interindividual variability, the global seizure reduction rate is not necessarily the most pertinent criterion. Indeed, some patients expressed an increased seizure frequency, while others had a marked seizure reduction.

Although the small size of the population limits the results, it appears that the responders are younger and have an extratemporal epileptic zone. The characteristics of the EZ, in particular, its extension and depth, are major determinants for limiting the efficacy of the tDCS and probably contribute to the low efficacy of most patients in our cohort.

Analysis of the EZ and its influence on response to tDCS has been analyzed in more detail in a recent study using the cohort from this and the previous study. ²⁰ This study showed that patients with more profound or widespread EZs had significantly poorer outcomes following multichannel tDCS. Non‐responders tended to exhibit deeper EZs and broader epileptogenic networks, limiting the inhibitory effect of the electric field due to current dispersion and reduced field strength in deeper regions. Taken as a whole, the impact on seizure reduction is superior to the placebo effect, reported to account for 6.5–11% seizure reduction in randomized studies. ⁷ , ¹¹ , ¹³ In addition, we observed a tendency for a decrease in seizure severity (NHS3 scores) in responders. It is worth noting that the assessment of epilepsy severity remains complex. The NHS3 is a valuable tool for evaluating the severity of seizures, particularly regarding seizure types and loss of awareness, but may in some cases, may not fully capture some other aspects of the condition (i.e., subjective symptoms, modification in specific semiology). ¹⁸

4.2. Effect on quality of life and psychiatric scales

Patients reported significant improvements in quality of life as measured by QOLIE‐31 scores, with notable gains in the “cognitive functioning” item and overall quality of life. Interestingly, patients with poorer baseline quality of life experienced greater improvements, suggesting that tDCS may particularly benefit those with substantial initial impairments. This effect aligns well with the benefits of quality of life observed in other neuromodulation techniques in this population of patients, in particular, vagus nerve stimulation (VNS) ²¹ or deep brain stimulation. ²² However, there was no significant relationship between changes in seizure frequency and overall quality‐of‐life scores. This indicates that factors beyond seizure reduction might contribute to quality‐of‐life enhancements (such as seizure intensity, which was not measured in this study). Changes in depression (NDDI‐E) and anxiety (GAD‐7) scores varied across participants, and no overall significant improvement was observed. This suggests that while tDCS can reduce seizure frequency and improve quality of life, its direct impact on psychiatric comorbidities remains unclear. Future studies with larger samples may provide more conclusive insights into this aspect.

4.3. Limitations

Several limitations should be acknowledged. The study's open‐label design may introduce bias, and the absence of a control group limits the ability to attribute observed effects solely to tDCS. This design was chosen for initial safety and feasibility evaluation and should be followed by a sham‐controlled randomized trial to confirm efficacy.

The sample size was also relatively small, and individual response variability was notable. The lack of significant changes in seizure severity (NHS3 scores) and psychiatric scales suggests that these domains require further investigation. Given the small sample size (n = 16), the statistical power is limited, and findings should be interpreted with caution. This study aims to inform the design of future larger‐scale, controlled trials. Another limitation is that although the study included adolescents, the findings may not be generalizable to younger pediatric populations.

5. CONCLUSION

This study provides evidence that personalized multichannel tDCS can be an effective and safe treatment for drug‐resistant epilepsy, offering significant seizure reduction and improved quality of life for a subset of patients. Despite its limitations, the results support the continued exploration of tDCS as a viable, noninvasive therapeutic option with few reversible side effects. Further randomized controlled trials with larger sample sizes, sham procedures, and extended follow‐up periods are needed to validate these findings and refine the protocol for broader clinical application.

FUNDING INFORMATION

The European Research Council supports this work: Galvani project, ERC‐SyG 2019, grant agreement No 855109.

CONFLICT OF INTEREST STATEMENT

Neuroelectrics Corp. and Starlab Neuroscience produced the noninvasive brain stimulation device used in this study. Dr. Giulio Ruffini is a co‐founder of these companies, and Dr. Salvador works for this company. All the other authors state no conflict of interest. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

Supporting information

Table S1.

Bartolomei F, Daoud M, Delourme M, Tardoski S, Makhalova J, Bourguiba E, et al. Personalized multichannel transcranial direct current electrical stimulation (tDCS) in drug‐resistant epilepsy: A SEEG based open‐labeled study. Epilepsia Open. 2025;10:1034–1042. 10.1002/epi4.70055

DATA AVAILABILITY STATEMENT

The datasets used in the present study are not publicly available due to privacy regulations of patient health information but may be available from the corresponding author upon reasonable request.