Abstract
Objective
This study was undertaken to evaluate the efficacy and safety of deep brain stimulation (DBS) of the anterior nucleus of the thalamus (ANT) compared with best medical treatment (BMT) in patients with drug‐resistant epilepsy (DRE).
Methods
This randomized, open‐label, phase 3 controlled trial was conducted across 14 specialized epilepsy and DBS centers. Eligible participants were adults with focal or multifocal DRE who had previously failed vagus nerve stimulation (VNS). Sixty‐one patients were randomized 1:1 either to receive continuous bilateral ANT‐DBS (n = 30) or to continue BMT, including VNS (n = 31), for 12 months. Afterward, patients in the BMT group were offered delayed ANT‐DBS and followed for an additional year. The primary outcome was the change in monthly severe seizure frequency (defined on the modified Chalfont Scale) at 12 months. Safety outcomes were also recorded.
Results
Among 67 screened patients, 61 were enrolled. At 12 months, median seizure reduction was greater in the DBS group (−44%, interquartile range [IQR] = −67 to 0) versus the BMT group (−6%, IQR = –56 to 20; p = .09); 44.5% of patients in the DBS group achieved a ≥50% reduction in seizure frequency compared to 27% in the BMT group. Within‐group analyses showed significant seizure reductions in the DBS group at both 12 months (−44%, p < .001) and 24 months (−46%, p < .001), as well as in the delayed DBS group at 12 months (−36%, IQR = –73 to −4; p < .001). No significant differences in quality of life were observed between groups, and no major DBS‐related adverse events were reported.
Significance
Our study suggests a potential benefit rather than demonstrating superiority of ANT‐DBS over medical therapy. However, within‐group improvements and favorable safety profile support the use of ANT‐DBS as a palliative treatment option in patients with DRE who have failed VNS.
Keywords: anterior nucleus of the thalamus, deep brain stimulation, drug‐resistant epilepsy, vagus nerve stimulation
Key points.
The efficacy and safety of ANT‐DBS in patients who have also failed VNS remain insufficiently explored, and the effect compared to BMT is not known.
In 61 patients, 45% of patients in the ANT‐DBS group achieved a ≥50% reduction in severe monthly seizure frequency compared to 27% in the BMT group (p: not significant). ANT‐DBS showed a statistically significant reduction in seizure frequency relative to baseline at 12 months (−44%) and 24 months (−46%), which was not achieved in the BMT group.
However, quality of life was not improved in the ANT‐DBS group.
No major adverse events related to DBS were noticed.
Our study confirms that ANT‐DBS may serve as a palliative therapeutic option for patients with drug‐resistant epilepsy.
1. INTRODUCTION
Epilepsy affects approximately 1% of people throughout the world, and for approximately one third of these patients, their disease is inadequately controlled by antiseizure drugs (ASDs). 1 For some of these patients, surgical resection or ablation can result in seizure freedom. However, not all individuals with drug‐resistant focal epilepsy are suitable candidates for epilepsy surgery. 2 For these patients, neuromodulation‐based interventions are a particularly promising avenue. 3 For example, vagus nerve stimulation (VNS) has proven effective for some patients and is currently the most widely used neurostimulation method, typically implemented as a first‐line approach before considering more advanced options. 4
Deep brain stimulation (DBS) delivers stimulation directly to a specific area of the brain. For epilepsy treatment, the area thought to be involved in the control of seizures is targeted; this targeting is key to the success of DBS in this indication. The neural circuits linked to the anterior nucleus of the thalamus (ANT)—in both the frontal and temporal lobes, the insula, and, to a lesser extent, the parieto‐occipital area—respond to electrical modulation by producing changes on electroencephalographic traces. 5 Stimulation of these circuits was shown to inhibit chemically induced seizures in laboratory models. 6 , 7 , 8 In patients with epilepsy, ANT‐DBS has been extensively studied in unblinded studies, 9 producing an approximately 50% reduction in seizure frequency. In the randomized blinded SANTE study, 10 bilateral ANT‐DBS reduced median seizure frequency by 42% at 3 months (28.7% in the unstimulated control group), achieving freedom from seizures in 12.7% of the study population for at least 6 months. The longer term follow‐up showed an increase in responder rates, which reached 68%. 11 The cohort for that study included some participants who had previously not responded to VNS. As DBS is a significantly more invasive procedure than VNS, it is important to determine whether it provides a clear added benefit.
Here, we report the outcomes of the French Anterior Nucleus Stimulation Cohort in Epilepsy (FRANCE) study. This randomized trial was designed to assess the efficacy of ANT‐DBS in patients with drug‐resistant focal epilepsy not eligible for or who failed resective surgery or after the failure of both pharmacological treatments and VNS.
2. MATERIALS AND METHODS
2.1. Trial design
In this randomized, controlled trial, we recruited patients from 14 DBS centers specialized in epilepsy surgery and DBS between June 2014 and April 2019. Sixty‐one patients were included and followed up for at least 2 years.
Eligible patients were randomly assigned in a 1:1 ratio to the DBS or best medical treatment (BMT) group. In line with the CONSORT guidelines (http://www.consort‐statement.org), randomization by minimization was used to assign patients to groups. Minimization was carried out on the following two factors: the duration of the disease (more or less than 15 years), and the number of seizures during the baseline period (more or fewer than 40 seizures). Due to the nature of the treatments, no masking was applied.
The study was conducted in accordance with the Declaration of Helsinki and Good Clinical Practice guidelines; the protocol was approved by a French ethics committee (CPP Sud Est V) on July 10, 2013 and authorized by the French National Agency for the Safety of Medicines and Health Products on August 23, 2013 (ID RCB: 2013‐A00880‐45).
The clinical trial is registered at ClinicalTrials.gov under ID NCT02076698. All subjects gave their written informed consent for participation in the study before enrollment.
2.2. Trial participants
Patients diagnosed with focal or multifocal seizures with or without focal to bilateral tonic–clonic seizures who were ineligible for surgery or for whom resective surgery had failed were eligible for inclusion. Further inclusion criteria were as follows: (1) patients had to have experienced intractable seizures for at least 4 years, with at least four seizures/month for at least 3 months during the baseline period (a minimum of 12 seizures during the 3 months); (2) treatment with at least two antiseizure medications for at least 2 years had to have failed; and (3) patient's disease had to have failed to respond to VNS over 2 years of stimulation.
Exclusion criteria were generalized epilepsy, focal aware seizures with only auras, any contraindications for DBS (vascular malformation, hemostatic disorder) or a contraindication to magnetic resonance imaging (MRI), evolutive disease such as tumor, risk of committing suicide over the previous 6 months (score ≥2 on item #10 of the Montgomery–Åsberg Depression Rating Scale), and women at risk of becoming pregnant during the study.
2.3. Trial interventions and procedures
At each neurosurgical center (see Figure 1), recruitment and follow‐up were assessed by the same investigators (neurologist and/or neurosurgeon) throughout the study, using a standardized protocol. All patients underwent a preinclusion assessment, including a 3‐month observation period.
FIGURE 1.
Trial timeline (A) and flowchart (B). M, month.
At each visit, the following elements were used by the neurologist and/or neurosurgeon to assess outcomes: patient and family seizure diary and validated French‐language version of the Quality of Life in Epilepsy Inventory‐31 (QOLIE‐31) questionnaire. Over the 3‐month preinclusion monitoring phase, the number of seizures was recorded for each patient. These records were used to define the eligible population, requiring an average of at least four seizures/month, and to minimize differences between the groups at randomization (minimization randomization described above). The information was also used to validate the severity of the seizures. Each patient was asked to rate the severity of their seizures based on the modified Chalfont Scale (see Table S1). 12 The Chalfont Scale was used because it provides a measure of seizure‐related severity and functional impairment, rather than simple event counts, allowing clinically meaningful changes to be detected even when seizure frequency remains unchanged. Patients with only type A seizures were not included in the study. Types B, C, and D were considered as severe.
All patients were clinically assessed for 1 month after electrode implantation, before turning on the stimulators. Numbers of seizures were compared between the baseline period (3 months before surgery) and the last 3 months of follow‐up (M10–M12).
At the end of the 12‐month period, patients in the BMT group were offered DBS. Monthly follow‐up of all patients was pursued until 24 months.
2.3.1. Surgery and lead localization
See Figure S1, supplementary material.
2.3.2. Stimulation parameters
At 1 month (see Figure S1), stimulators were turned on, and the parameters were set for the 12‐month study period with monthly assessments. Over the first 3 months, stimulations parameters were set to 130 Hz, 60 μs for the pulse width, and the voltage was adjusted to optimize response; the choice of the active contact was based on the postoperative imaging showing the contact located into the ANT. Thereafter, parameters were maintained unchanged until Month 12 (M12). If no obvious effect of stimulation was observed in the 3‐month titration period, the parameters were set as follows: frequency at 130 Hz, 3.5‐V maximum, 60‐μs pulse width applied to the contact located in the ANT, based on the anatomical, MRI definition.
When the patient's VNS device was ON at inclusion, the stimulation parameters were kept the same throughout the study. For patients fitted with a VNS device that was inactive at the time of inclusion, the device remained inactive for the duration of the study.
ASDs were maintained unchanged in the DBS and BMT groups from the inclusion to the end of the study. However, in the case of status epilepticus or severe seizures worsening, therapeutical adjustments were still possible.
2.4. Outcomes
The primary outcome sought was a 50% reduction in the frequency of severe seizures (defined as types B, C, and D) at 1 year that was compared between the two groups. In addition, we performed intragroup post hoc analysis of the median reduction of severe seizure compared to baseline. To do so, average frequency of severe seizures was assessed between the 3 months presurgery (M−3 to M0) and the last 3 months of follow‐up (M10–M12), and M22–M24 for the delayed DBS group. The average reduction in severe seizure numbers was also measured at 1‐year follow‐up for the delayed DBS group and at 2 years for the DBS group.
Additionally, we evaluated seizure outcomes at 1 year based on seizure onset location, categorized into four groups: Case 1, temporal or bitemporal with or without insular involvement; Case 2, frontal or bifrontal with or without insular involvement; Case 3, multifocal; and Case 4, posterior quadrant, defined as parietal and/or occipital onset. Secondary outcomes were related to effects on quality of life (QoL), mood, and memory at 1 and 2 years. In particular, we used the validated scales of the Beck Depression Inventory (BDI) to determine mood, the Wechsler Adult Intelligence Scale IV (WAIS‐IV) for intellectual efficiency, the Trail Making Test and Stroop to test executive function, and the Wechsler Memory Scale (MEM) and Recall on Rey–Osterrieth Complex Figure to evaluate memory. We used the validated French‐language versions of two QoL questionnaires: one specific to epilepsy, QOLIE‐31; and the generic Health Utilities Index Mark 3 to measure any impact on QoL.
2.5. Adverse events
All adverse events (AEs) related to surgery or to chronic stimulation were monitored for all patients and are reported at 1 and 2 years of follow‐up. AEs were coded according to the Medical Dictionary for Regulatory Activities, version 24.1. Serious AEs (SAEs) were defined as any AEs that led to death, disability, or prolonged or new hospitalization due to a serious health problem.
2.6. Sample size justification
The power calculation was based on the estimation that 50% of patients in the DBS group would show at least 50% improvement in seizure numbers, whereas in the BMT group 10% of patients would show a 50% improvement. We designed the study to have an overall power of 90% with an alpha‐risk of 5%. According to calculations performed using nQuery Advisor 6.01 software, a sample size of 62 patients (31 per group) would provide this statistical power.
2.7. Statistical analysis
Statistical analyses were all undertaken by the statistics department at Grenoble Alpes University Hospital.
All analyses were performed on an intention‐to‐treat basis. Categorical variables were presented using counts and frequencies, whereas continuous variables were presented using means, SDs, medians, and interquartile ranges (IQRs). The threshold set for statistical significance was p < .05. No adjustments were made for repeated testing. Analyses were performed using Stata (v15.1). Normality of continuous variables was assessed graphically on histogram. In accordance with the statistical analysis plan, missing data were not imputed. Complete case analysis was performed.
2.7.1. Analysis of primary efficacy endpoint
The superiority of neurostimulation in terms of success rate, that is, the proportion of patients with at least 50% reduction in seizures, between the mean preintervention seizure frequency (M−3 to M0) and the mean seizure frequency at 1 year (M10–M12) or 2 years (M22–M24), was tested using a χ 2 test (Fisher exact test for small‐sized samples).
2.7.2. Analysis of secondary efficacy endpoint
Qualitative variables in each group were compared using a χ 2 test (Fisher exact test for small‐sized samples). Quantitative variables in each group were compared using Student t‐test (nonparametric Wilcoxon test in the case of non‐Gaussian distribution).
A post hoc analysis was conducted to assess the impact of seizure onset location on the primary outcome, using a logistic regression model that included both seizure onset location and treatment group as covariates.
2.7.3. Description of study population
The primary analysis population included all patients randomized to the DBS or the BMT group.
2.8. Role of the funding source
This study was funded by the “Direction Generale de l'Offre de Soins,” which is part of the French Ministry of Health. The funder of the study played no role in study design, data collection, data analysis, data interpretation, or writing of the report. Devices were purchased from Medtronic; the company played no role in the initiation of the study, in study design, data collection, data analysis, data interpretation, or writing of the report. The corresponding author had full access to all the data in the study and had final responsibility for the decision to submit for publication.
3. RESULTS
3.1. Study population
Between June 16, 2014 and April 3, 2019, we assessed 67 patients for eligibility (see Table 1), enrolling 61 in the study (Figure 2). Thirty patients were randomly allocated to the DBS group and 31 to the BMT group. One patient in the DBS group died from probable sudden unexplained death in epilepsy (SUDEP) before surgery could be performed; consequently 29 patients completed the study at 12 months in this group. One patient from the BMT group was lost to follow‐up; as a result, 30 patients completed the study at 12 months in this group. In the BMT group, after 12 months of follow‐up, patients were offered DBS (delayed DBS group). However, one patient died before implantation could be performed, one patient declined to pursue participation in the study, and three patients refused the implantation. Consequently, 25 patients of the BMT group received DBS at 12 months.
TABLE 1.
Clinical characteristics at baseline.
| Characteristic | DBS group | BMT group |
|---|---|---|
| Age, median (IQR) | 34 (25–40) | 34 (28–45) |
| Sex [male], n (%) | 21 (72.4) | 11 (36.6) |
| Level of study [diploma], n (%) a | 20 (68.9) | 24 (80) |
| Marital status [single], n (%) | 24 (82.7) | 18 (60) |
| Occupational status [active employment], n (%) b | 8 (27.5) | 9 (30) |
| Disease duration > 15 years, n (%) | 18 (62) | 21 (70) |
| >40 seizures/month, n (%) | 7 (24.1) | 6 (20) |
| IQ, median (IQR) | 68 (61–83) | 79 (72–89) |
| Type of seizures [temporal], n (%) | 4 (13.7) | 11 (36.6) |
| Seizures/month | ||
| Median (IQR) | 11 (7–20) | 12 (6–21) |
| Mean (SD) | 16.7 (15.7) | 15 (11.7) |
| Beck Depression Inventory score | ||
| Median (IQR) | 6 (4–10) | 6 (2–9) |
| Mean (SD) | 7.8 (6.7) | 6.4 (5) |
| Working memory score/19 | ||
| Median (IQR) | 6.5 (3–8) | 6 (4–7) |
| mean ± SD | 5.9 (3.6) | 6.5 (2.7) |
| Processing Speed score/19 | ||
| Median (IQR) | 3 (1–5.5) | 6 (5–8) |
| mean ± SD | 3.8 (3) | 6.2 (2.6) |
| VCI score | ||
| Median (IQR) | 81 (69–88) | 81 (67–88) |
| mean ± SD | 74.3 (26.9) | 74.1 (24.6) |
| PRI score | ||
| Median (IQR) | 74 (62–80) | 83.5 (72–98) |
| mean ± SD | 66.4 (25.1) | 77.5 (29.3) |
Abbreviations: BMT, best medical treatment; DBS, deep brain stimulation; IQ, intelligence quotient; IQR, interquartile range; PRI, perceptual reasoning index; VCI, verbal comprehension index.
Diploma = “middle school completed,” “general or technical high school diploma,” “2 years of third level,” “professional diploma,” or “More than 2 years third level training.”
Active employment: “Occupies a profession full‐time” or Part‐time employment.
FIGURE 2.
Temporal evolution of median seizure frequency. Graph shows the percent change in the median number of seizures for deep brain stimulation (DBS) and best medical treatment (BMT) groups at 1 and 2 years. The delayed surgery group corresponds to patients included in the BMT group who accepted to be implanted and stimulated at 1‐year follow‐up. M, month; ns, not significant.
There was an imbalance in terms of gender between our groups. This variable was not part of our randomization algorithm. Regarding our primary endpoint, an adjustment for gender showed no difference in the success of reducing severe seizures (male: odds ratio [OR] = 1.26, p = .696), according to a logistic regression model that also adjusted for group. This adjustment did not alter the significance of the treatment group on the primary outcome either (BMT: OR = .48, p = .223).
The other clinical characteristics were well balanced between the two groups, except intelligence quotient, which was significatively higher in the BMT group, and except cases of temporal epilepsy, which tended to be more frequent in the BMT group: four (13.7%) versus 11 (36.6%) in the DBS and BMT groups, respectively (p = .07).
The BDI scores for the two groups were no different. During the first part of the study (up to 12 months), no patients changed their treatment.
At inclusion, there were some differences between the two groups regarding executive functions as measured using the WAIS. Specifically, the score for processing speed and the perceptual reasoning index were higher in the BMT group. With regard to memory, no difference between the two groups was found at the time of inclusion.
3.2. Efficacy
At the 12‐month follow‐up (see Table 2), the proportion of responders was higher in the DBS group (44.83%, 95% confidence interval [CI] = 26.4–64.3) compared with the BMT group (26.67%, 95% CI = 12.3–45.9), but the difference did not reach statistical significance (p = .14).
TABLE 2.
Effects on seizures at 1 and 2 years.
| DBS group | BMT group | Delayed DBS group [1 year postop] | p | |
|---|---|---|---|---|
| Patients with 50% seizures reduction at 12 months, % | 44.83 | 26.67 | ns, .14 | |
| Change of seizures at 12 months [post hoc], median % (IQR) | −44.44 (−66.66 to 0) | −5.95 (−56.10 to 19.64) | ns, .09 | |
| Change of seizures at 12 months [post hoc] compared to baseline, median % (IQR) | −44.44 (−66.66 to 0)** | −5.95 (−56.10 to 19.64) a |
<.0001** ns a |
|
| Change of seizures at 24 months for DBS and 12 months for BMT [delayed DBS group; post hoc] compared to baseline, median % (IQR) | −46.43 (−66.66 to 0)* | −35.71 (−72.73 to −4.17)** |
.001* <.0001** |
Abbreviations: BMT, best medical treatment; DBS, deep brain stimulation; IQR, interquartile range; ns, not significant.
p < .001.
p < .0001.
p not significant.
Regarding seizure frequency, DBS produced a median reduction of 44% (range = −66.66 to 0), whereas BMT resulted in a median reduction of 5.95% (range = −56.10 to 19.64); this between‐group comparison approached but did not achieve significance (p = .09).
Within‐group analyses showed that the reduction from baseline to 12 months was significant in the DBS cohort (44%, p < .0001), whereas no significant change was observed in the BMT cohort (5.95%, p = .28).
Monthly seizure frequency decreased from a median of 11 (IQR = 7–20) to 6 (IQR = 4–19) in the DBS group, and from 12 (IQR = 6–19) to 7 (IQR = 4–18) in the BMT group over the 12‐month period, indicating a clinically meaningful improvement with DBS, but formal between‐group significance was not reached.
Two patients in the DBS group were completely seizure‐free at 12 months, as were four in the BMT group (p: not significant).
Regarding the results of the delayed DBS group at 1 year, we observed a 35.71% (IQR = −72.73 to −4.17) median decrease of seizures that reached significance when compared to baseline (p < .0001).
At 24 months, the median percent reduction in seizure frequency, compared to baseline, was maintained at 46.63% (IQR = −66.66 to 0) for the DBS group; this result was still significant (p = .001).
3.3. Seizure outcome based on seizure onset location
Data were available for 58 patients (see Figure 3), with 29 in each treatment group. Patients with frontal or bifrontal seizure onset (with or without insular involvement) exhibited the highest response rates, with seven of 10 patients in the DBS group achieving a ≥50% reduction in seizure frequency compared to three of nine patients in the BMT group (p = .10). Temporal or bitemporal onset cases (± insular involvement) showed a trend toward better outcomes in the DBS group (3/6) than in the BMT group (3/10), although the difference did not reach statistical significance (p = .23).
FIGURE 3.
Median seizure frequency according to seizure onset location. Case 1: posterior quadrant. Case 2: frontal/bifrontal ± insula. Case 3: multifocal. Case 4: temporal/bitemporal ± insula. One patient had no seizure type. BMT, best medical treatment; DBS, deep brain stimulation.
3.4. Effects on QoL
The secondary outcomes focused particularly on QoL (see Table 3). At 12 months, no difference in QoL was observed between the two groups, nor was there a difference among the QOLIE‐31 subscores.
TABLE 3.
Effect on quality of life.
| DBS group | BMT group | p | |
|---|---|---|---|
| Seizure worry | |||
| Median (IQR) | 59.5 (49.7; 78.3) | 67 (31.3; 80) | |
| mean ± SD | 61.6 (21.2) | 56.8 (29) | .474 |
| Overall quality of life | |||
| Median (IQR) | 67.5 (55; 77.5) | 67.5 (50; 77.5) | |
| mean ± SD | 65.4 (15.2) | 63.6 (23.4) | .723 |
| Emotional well‐being | |||
| Median (IQR) | 64 (52; 74) | 66 (52; 76) | |
| mean ± SD | 61.6 (17.8) | 63.5 (18.4) | .692 |
| Energy/fatigue | |||
| Median (IQR) | 50 (45; 65) | 55 (35; 60) | |
| mean ± SD | 52.3 (16.8) | 52 (17.6) | . 944 |
| Cognitive functioning | |||
| Median (IQR) | 62.8 (43.5; 87.4) | 60.3 (39.5; 70.5) | |
| mean ± SD | 62.9 (25.8) | 54.3 (25) | .204 |
| Medication effects | |||
| Median (IQR) | 58.3 (44.4; 77.8) | 62.5 (44.4; 83.3) | |
| mean ± SD | 60.6 (26.4) | 61.6 (26.8) | 919 |
| Social functioning | |||
| Median (IQR) | 38.5 (17; 61) | 41.5 (27; 66) | |
| mean ± SD | 39.7 (25.6) | 43.5 (26.9) | .624 |
| QOLIE‐31 | |||
| Median (IQR) | 56.2 (46.3; 69.3) | 54.2 (46.9; 64.2) | |
| mean ± SD | 56.7 (15.6) | 54.8 (16.7) | .66 |
Abbreviations: BMT, best medical treatment; DBS, deep brain stimulation; IQR, interquartile range; QOLIE‐31, Quality of Life in Epilepsy Inventory‐31.
3.5. Effects on depression scores and memory
BDI scores for depression showed that 88.46% of patients treated with DBS significantly improved their score at 12 months (Table S3), compared to 55.56% of patients in the BMT group (p = .014). It should nevertheless be noted that patients from both groups had low scores for depression at inclusion, with median BDI scores at baseline of 6 (IQR = 4–10) for the DBS group and 6 (IQR = 2–9) for the BMT group. These scores were not significantly different within each group at 1 year (4 [IQR = 1–6] for DBS group vs. 4 [IQR = 2–7] for BMT group).
Memory assessments based on MEM scores revealed no effect of DBS on memory and no significant differences between the two groups at 1 year.
3.6. Locations of contacts
Data on anatomical locations of the DBS leads were available for 48 patients (Figure S1C). At least one contact was located in the ANT in 91% of patients, and two or more contacts were in the ANT in 79% of patients.
3.7. Adverse events
Forty‐five SAEs were reported in the DBS group at 1 year (see Tables S4a and S5), compared to 19 in the BMT group. One patient in the DBS group died before any surgical procedure due to SUDEP. Seven of the AEs were related to surgery or stimulation, with one small asymptomatic intracranial hemorrhage at the entry point of a lead, three lead migrations (requiring surgery for repositioning), and three infections requiring removal of the stimulator and electrodes. In addition, one patient suffered a severe, life‐threatening event due to a reactivation of brain herpes simplex virus after turning on the device. 12 Other AEs related to the device or surgery were mild, transient, and in some cases resolved after voltage adjustment (this was notably the case for anxiety).
At 2 years of follow‐up, a sum total of 70 SAEs were recorded in the DBS group, compared to 46 in the BMT group. One patient in the delayed DBS group died by drowning before any surgical procedure. Other SAEs were similar to those described at 1 year.
4. DISCUSSION
4.1. Key results
In this randomized, controlled trial of 61 patients with severe, intractable epilepsy, the primary objective was to compare the rate of responders between ANT‐DBS and BMT groups at 12 months. Only severe seizures types B, C, and D of the modified Chalfont Scale were taken into account. Although there was a greater number of responders in the DBS group, this result did not reach significance (45% vs. 27%, p = .14). However, DBS resulted in a median 44% decrease in the frequency of seizures, compared to a median decrease of 5.95% with BMT (p = .09).
This lack of significance when comparing the number of responders is probably linked to a lack of statistical power. When calculating sample size, a 50% reduction in seizures in 50% of patients was targeted. The calculations were based on the estimation that only 10% of patients in the control group would show a 50% reduction in severe seizures. This estimation was derived from published reports such as Rheims et al. 13 Surprisingly, almost 30% of patients in the BMT group, in our cohort, achieved this level of seizure reduction. It is therefore possible that our study design lacked the statistical power to demonstrate the targeted level of improvement. We do not have a definitive explanation for the unexpectedly high responder rate observed in the BMT group; however, enhanced medical support and follow‐up may have temporarily influenced patients' seizure reporting in their diaries.
In contrast, the median change in seizure frequency was significantly reduced in the DBS group compared to baseline (44% vs. 6% at 1 year, p < .0001, post hoc analysis). Moreover, this improvement relative to baseline remained significant after 2 years for the DBS group. Furthermore, between 12 and 24 months, the trend for reduced seizure frequency was confirmed in patients from the BMT group who elected to undergo implantation surgery (delayed DBS group). In this group, the median change in seizure frequency was also significantly improved.
4.2. Interpretation
The results are in line with those of previous controlled studies in terms of the 50% responder rate at 1 year (37%, 43%, 44%, and 44.4%) for VNS, ANT‐DBS, responsive neurostimulation, and the FRANCE study reported here. 4 , 14 , 15 , 16 Our outcomes are also slightly better than those of the recently published MORE registry, 17 where a median reduction of seizures of 33.1% was reported at 2 years, although our study represents the worst case scenario, applying ANT‐DBS in patients whose disease failed to respond to both common medications and VNS.
In the MORE registry, 43% of patients had failed to respond to VNS before application of DBS. In the FRANCE study, at least one electrode contact was accurately positioned within the ANT in 91% of patients, which may have contributed to enhancing the therapeutic efficacy of DBS.
Only two of our ANT‐DBS patients achieved seizure freedom at 1 year (and four in the BMT group). This rate of success confirms previously reports 4 that ANT‐DBS is not a curative treatment, but rather a palliative option, offering meaningful reductions in seizure frequency.
Regarding the effect of DBS based on seizure onset location, we found that patients with frontal seizure onset (with or without insular involvement) were twice as likely to achieve a ≥50% median reduction in seizure frequency. A greater reduction in seizures was also observed in patients with temporal or bitemporal onset (± insular involvement) in the DBS group compared to the BMT group. However, due to the limited sample size, these differences did not reach statistical significance.
The location of the epileptogenic zone (EZ) and its relation to DBS effects is an important consideration. From our data, we were able to perform a coarse analysis at the lobar level and explore potential correlations with DBS response. However, a more detailed localization of the EZ within a given lobe was not feasible. In all cases, the EZ was often bifocal, multifocal, or spanning several lobes, thus precluding well‐defined resective targets. Consequently, we cannot provide more precise correlations between EZ location and DBS outcomes. Moreover, the small number of patients in each subanatomical subgroup further limits the statistical validity of such analyses.
Thalamic DBS is known to induce effects on mood, 10 , 11 , 16 , 18 memory, and psychiatric symptoms, 19 with the extent varying depending on the area stimulated. For our cohort, mood assessment mainly focused on depression, using the BDI. The scores obtained revealed improvements for a greater proportion of patients in the DBS group than in the BMT group, but the differences in absolute scores were not clinically relevant. Based on these results, we can conclude that, in our cohort, ANT‐DBS induced no worsening of mood. This contrasts with the original SANTE trial report 10 and the long‐term report by Salanova et al., 16 where approximately 15% of patients developed significant levels of depression. Although symptoms seemed to improve over time, for approximately 37% of affected patients, they persisted at long‐term follow‐up, with 10% reporting suicidal ideation.
Previous reports linked memory impairment to chronic ANT‐DBS. 11 , 17 In our cohort, no memory impairment was observed in the DBS group, as measured using the MEM III scale.
One possible explanation for these apparent discrepancies is the voltage limitation in our study; voltage was capped at 3.5 V, potentially reducing the risk of current spread to the mammillothalamic tract. In preliminary tests in some patients, we noticed an acute mood worsening that typically occurred within 1 month of initiating stimulation when the voltage was set around or above 4 V (data not shown). Adjustment of stimulation parameters has previously been reported to successfully reduce mood disorders linked to use of ANT‐DBS. 20 The difference might also be related to the type of lead used (3389 model in the FRANCE study versus 3387 in the SANTE trial). The 3389 lead is shorter, with a smaller intercontact space; it therefore helps to avoid creating numerous contacts outside the ANT nucleus, particularly in the dorsal thalamic nuclei. Although the exact anatomical location of the active contacts was not fully evaluated in the present study, at least one contact was located in the ANT in 91% of patients. In 79% of patients, two contacts were appropriately placed in the ANT. This level of appropriate lead placement is higher than achieved in previous studies and can probably be linked to the use of transventricular trajectories whenever possible. The transventricular route is reported to be a better trajectory to approach the ANT than classical stereotactic placement. 21 Anatomical location of leads and its effects will be fully assessed in a separate study.
4.3. Limitations
Our study was underpowered, primarily because of its design. The original protocol aimed to compare the DBS group with a control group receiving BMT. However, the response rate in the BMT group was higher than anticipated, reducing the study's ability to detect a between‐group difference. Consequently, we cannot draw definitive conclusions regarding the superiority of DBS over BMT. An alternative design—comparing the reduction in seizure frequency from baseline within each group—might have been more appropriate and more likely to yield statistically significant results without increasing the sample size.
4.4. Impact on clinical care
Despite a greater reduction in severe seizure frequency in the DBS group, QoL was not improved compared to the control group. This result aligns with those from the MORE registry, also reporting no improvement in QOLIE. It was to be expected because of the palliative nature of ANT‐DBS, with few patients achieving seizure freedom. Seizure freedom is the most relevant factor influencing QOLIE. 22 However, neuromodulation therapies may also confer QoL benefits through mechanisms that are at least partly independent of seizure reduction, particularly within emotional and psychosocial domains. Notably, one study reported improvements in QoL in patients treated with responsive neurostimulation despite the absence of seizure freedom; these gains were primarily driven by improvements in mood rather than seizure control. 23 In the same vein, VNS therapy has been associated with improvements of QoL not associated with seizure freedom. 24 , 25
In our cohort, although we observed some improvement in mood measures, this did not translate into a statistically significant improvement in overall QoL.
When recommending ANT‐DBS, clinicians should acknowledge that although seizure reduction may occur, the overall impact on the patient's QoL is likely to be limited.
5. CONCLUSIONS
The findings from the FRANCE study show that ANT‐DBS is beneficial for approximately 45% of patients with severe, intractable epilepsy, resistant to both drugs and VNS, who can expect 50% reduction of severe seizures on average. However, our study suggests a potential benefit rather than demonstrating superiority of ANT‐DBS over medical therapy. We acknowledge that the limited statistical power of our small cohort warrants caution in interpreting our conclusions, particularly given that the effect observed in the BMT arm was greater than anticipated.
Although no impact on QoL could be demonstrated at 1‐year follow‐up, the low incidence of severe, clinically relevant AEs is a strong argument to use ANT‐DBS as a palliative option that should be included in the therapeutic arsenal for patients with drug‐ and VNS‐resistant epilepsy.
AUTHOR CONTRIBUTIONS
Conception and design of the study: Stéphan Chabardès, Fabrice Bartolomei, Claire Haegelen, Sylvain Rheims, Philippe Kahane, Sandra David‐Tchouda, and Jean Regis. Acquisition and analysis of data: all authors. Drafting a significant portion of the manuscript or figures: Stéphan Chabardès, Fabrice Bartolomei, Claire Haegelen, Sylvain Rheims, Sandra David‐Tchouda, and Jean Regis. Statistical analysis and interpretation: Corentin Leroy, Sandra David‐Tchouda, and Stéphan Chabardès. Review of the final version of the paper: all authors.
CONFLICT OF INTEREST STATEMENT
S.C. has served as a paid consultant for Medtronic and Boston Scientific. F.B. has served as paid consultant for Medtronic and LivaNova. S.R. has served as paid consultant for LivaNova, UCB Pharma, Angelini Pharma, Neuraxpharma, and Jazz Pharma. A.N. has received honoraria for lecturing for Angelini Pharma, Eisai, Jazz Pharma, LivaNova, and UCB. The remaining authors have no conflicts 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
Data S1.
ACKNOWLEDGMENTS
Grenoble Alpes University Hospital was the promoter of this study, providing technical resources and ensuring regulatory approvals. The authors wish to thank all collaborators who participated in this study. The “Direction Générale de l Offre de soins–DGOS,” French Ministry of Health, funded the study. Open access publication funding provided by COUPERIN CY26.
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available on reasonable request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
REFERENCES
- 1. Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med. 2000;342(5):314–319. [DOI] [PubMed] [Google Scholar]
- 2. Wiebe S, Jette N. Pharmacoresistance and the role of surgery in difficult to treat epilepsy. Nat Rev Neurol. 2012;8(12):669–677. 10.1038/nrneurol.2012.181 [DOI] [PubMed] [Google Scholar]
- 3. Englot DJ. A modern epilepsy surgery treatment algorithm: incorporating traditional and emerging technologies. Epilepsy Behav. 2018;80:68–74. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Ryvlin P, Rheims S, Hirsch LJ, Sokolov A, Jehi L. Neuromodulation in epilepsy: state‐of‐the‐art approved therapies. Lancet Neurol. 2021;20:1038–1047. [DOI] [PubMed] [Google Scholar]
- 5. Kerrigan JF, Litt B, Fisher RS, Cranstoun S, French JA, Blum DE, et al. Electrical stimulation of the anterior nucleus of the thalamus for the treatment of intractable epilepsy. Epilepsia. 2004;45(4):346–354. 10.1111/j.0013-9580.2004.01304.x [DOI] [PubMed] [Google Scholar]
- 6. Mirski MA, Rossell LA, Terry JB, Fisher RS. Anticonvulsant effect of anterior thalamic high frequency electrical stimulation in the rat. Epilepsy Res. 1997;28:89–100. [DOI] [PubMed] [Google Scholar]
- 7. Sherdil A, Chabardès S, David O, Piallat B. Coherence between the hippocampus and anterior thalamic nucleus as a tool to improve the effect of neurostimulation in temporal lobe epilepsy: an experimental study. Brain Stimul. 2020;13(6):1678–1686. [DOI] [PubMed] [Google Scholar]
- 8. Sherdil A, Coizet V, Pernet‐Gallay K, David O, Chabardès S, Piallat B. Implication of anterior nucleus of the thalamus in mesial temporal lobe seizures. Neuroscience. 2019;418:279–290. [DOI] [PubMed] [Google Scholar]
- 9. Osorio I, Overman J, Giftakis J, Wilkinson SB. High frequency thalamic stimulation for inoperable mesial temporal epilepsy. Epilepsia. 2007;48(8):1561–1571. 10.1111/j.1528-1167.2007.01044.x [DOI] [PubMed] [Google Scholar]
- 10. Fisher R, Salanova V, Witt T, Worth R, Henry T, Gross R, et al. Electrical stimulation of the anterior nucleus of thalamus for treatment of refractory epilepsy. Epilepsia. 2010;51(5):899–908. [DOI] [PubMed] [Google Scholar]
- 11. Salanova V. The SANTÉ study at 10 years of follow‐up: effectiveness, safety, and sudden unexpected death in epilepsy. Epilepsia. 2021;62(6):1306–1317. [DOI] [PubMed] [Google Scholar]
- 12. Hamdi H, Robin E, Stahl J‐P, Doche E, Azulay J‐P, Chabardes S, et al. Anterior thalamic stimulation induced relapsing encephalitis. Stereotact Funct Neurosurg. 2019;97(2):132–136. 10.1159/000499072 [DOI] [PubMed] [Google Scholar]
- 13. Rheims S, Cucherat M, Arzimanoglou A, Ryvlin P. Greater response to placebo in children than in adults: a systematic review and meta‐analysis in drug‐resistant partial epilepsy. PLoS One. 2008;5(8):1223–1237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Morris GL, Mueller WM. Long‐term treatment with vagus nerve stimulation in patients with refractory epilepsy. Neurology. 1999;53(8):1731. [DOI] [PubMed] [Google Scholar]
- 15. Heck CN, King‐Stephens D, Massey AD, Nair DR, Jobst BC, Barkley GL, et al. Two‐year seizure reduction in adults with medically intractable partial onset epilepsy treated with responsive neurostimulation: final results of the RNS system pivotal trial. Epilepsia. 2014;55(3):432–441. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Salanova V, Witt T, Worth R, Henry TR, Gross RE, Nazzaro JM, et al. Long‐term efficacy and safety of thalamic stimulation for drug‐resistant partial epilepsy. Neurol Int. 2015;84(10):1017–1025. 10.1212/WNL.0000000000001334 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Peltola J, Colon AJ, Pimentel J, Coenen VA, Gil‐Nagel A, Ferreira AG, et al. Deep brain stimulation of the anterior nucleus of the thalamus in drug‐resistant epilepsy in the MORE multicenter patient registry. Neurology. 2023;100(18):E1852–E1865. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Tröster AI, Meador KJ, Irwin CP, Fisher RS. Memory and mood outcomes after anterior thalamic stimulation for refractory partial epilepsy. Seizure. 2017;45:133–141. [DOI] [PubMed] [Google Scholar]
- 19. Hariz MI. Psychosurgery, deep brain stimulation, and the re‐writing of history. Neurosurgery. 2008;63:E820. [DOI] [PubMed] [Google Scholar]
- 20. Voges BR, Schmitt FC, Hamel W, House PM, Kluge C, Moll CKE, et al. Deep brain stimulation of anterior nucleus thalami disrupts sleep in epilepsy patients. Epilepsia. 2015;56(8):e99–e103. [DOI] [PubMed] [Google Scholar]
- 21. Lehtimäki K, Coenen VA, Gonçalves Ferreira A, Boon P, Elger C, Taylor RS, et al. The surgical approach to the anterior nucleus of thalamus in patients with refractory epilepsy: experience from the international multicenter registry (MORE). Clin Neurosurg. 2019;84:141–150. [DOI] [PubMed] [Google Scholar]
- 22. Wiebe S, Blume WTB, Girvin JP, Eliasziw M. A randomized, controlled trial of surgery for temporal lobe epilepsy. N Engl J Med. 2001;345:311–318. www.nejm.org [DOI] [PubMed] [Google Scholar]
- 23. Meador KJ, Kapur R, Loring DW, Kanner AM, Morrell MJ. Quality of life and mood in patients with medically intractable epilepsy treated with targeted responsive neurostimulation. Epilepsy Behav. 2015;45:242–247. [DOI] [PubMed] [Google Scholar]
- 24. Ryvlin P, Gilliam FG, Nguyen DK, Colicchio G, Iudice A, Tinuper P, et al. The long‐term effect of vagus nerve stimulation on quality of life in patients with pharmacoresistant focal epilepsy: the PuLsE (open prospective randomized long‐term effectiveness) trial. Epilepsia. 2014;55(6):893–900. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Kawai K, Tanaka T, Baba H, Bunker M, Ikeda A, Inoue Y, et al. Outcome of vagus nerve stimulation for drug‐resistant epilepsy: the first three years of a prospective Japanese registry. Epileptic Disord. 2017;19(3):327–338. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data S1.
Data Availability Statement
The data that support the findings of this study are available on reasonable request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.