Intracranial electrical stimulation can be considered for patients with drug-refractory epilepsy when the epileptogenic focus is not clear or when the epileptogenic focus is clear but cannot be surgically removed. The targets of DBS for epilepsy are many, including the anterior nucleus of the thalamus (ANT), subthalamic nucleus (STN), cerebellum, and hippocampus. nucleus (STN), cerebellum, hippocampus, caudate nucleus, etc. The possible treatment mechanisms are as follows.
(1) Interference with seizure-related neural circuits.
(2) High-frequency stimulation inhibits epileptiform discharges in the cortex.
Among the targets of DBS, ANT is the most studied, and the following focuses on anterior thalamic nucleus electrical stimulation (ANT-DBS).
I. Overview
In 1937, Papez JW described a classical loop: the hippocampus sends fibers via the fornix and papillary somatic nucleus to the ANT, and the ANT sends fibers back to the hippocampus via the cingulate bundle, the lateral ventricular wall periaqueduct, and the parahippocampal cortex [8]. Alterations in imaging of the structures involved in the Papez loop are seen in patients with medial temporal lobe sclerosis as well as other forms of epilepsy, thus speculating that structures in the Papez loop, including the ANT, are involved in seizures. Animal experiments have shown that high-frequency stimulation of the ANT can increase seizure thresholds and that the ANT nucleus is small and suitable as a target for DBS, so ANT-DBS has been attempted for the treatment of epilepsy.
Patient selection
Patients with epilepsy treated clinically with ANT-DBS generally have the following characteristics: partial seizures with frequent falls and significantly impaired quality of life and a poor response to at least two antiepileptic drugs for 12 to 18 months of treatment. ANT-DBS may also be considered in patients with epilepsy who have failed vagus nerve stimulation or epileptogenic focus resection.
II. Surgical approach
After the patient is fitted with a stereotactic head frame, an MRI scan is performed and the ANT is positioned on the reconstructed MRI image at 5 mm parasternal and 12 mm superior to the cerebral origin according to the standard brain atlas. There was no cell discharge when the microelectrode crossed the lateral ventricle, and cell discharge could be recorded until it reached the ANT surface. When the microelectrode is clearly ANT, the microelectrode is removed and the stimulating electrode is placed, with the following stimulation parameters: frequency 5-10 Hz, pulse width 90-330 μs, voltage 4-5 V, and stimulation time 3-10 s. The stimulation-related EEG changes can be observed on the intraoperative scalp EEG, and it should be noted that besides ANT, stimulation of other nuclei of the thalamus can also show similar EEG changes, so Such EEG changes do not prove that the electrode is in the ANT. After removal of the stimulation electrodes, the Medtronic Model 3387 electrodes were implanted so that each contact of the electrodes was located within the thalamus, the electrodes were fixed, and the stimulation generator was implanted under the clavicle.
III. Postoperative treatment
Postoperative CT or MRI review the position of the electrodes. The stimulation generator is mostly turned on within 10 days after surgery, and the initial stimulation parameters are: frequency 90-130 Hz, pulse width 60-90 μs, and voltage 4-5 V. The stimulation can be chosen as intermittent stimulation, because intermittent stimulation has similar efficacy as continuous stimulation, but intermittent stimulation is more energy-efficient. The specific stimulation parameters should be adjusted according to the efficacy, where the intermittent stimulation can be referred to Litt B[s article: 1 minute on, 5 minutes off.
IV. Clinical efficacy
The effectiveness of ANT-DBS in the treatment of epilepsy may be related to the following three aspects.
1. the micro-destructive effect: Velasco F found that the seizure frequency of patients decreased significantly immediately after electrode implantation, and the later adjustment of stimulation parameters had no further effect on seizure control, but a recent study found that the micro-destructive effect did not reduce seizure frequency.
2. the direct stimulation effect of ANT-DBS: the study by Kerrigan et al. found a significant decrease in seizure frequency in 4 out of 5 patients and a significant increase in seizure frequency when the stimulator was turned off.
3. Long-term stimulation effect of ANT-DBS: A study found that 1 patient with DBS turned off 3 years after surgery was still able to maintain more than 1 year of efficacy.
Although studies on ANT-DBS for drug-refractory epilepsy are progressively being conducted, most have small sample sizes. Table 1 shows 6 previous studies suggesting that ANT-DBS can reduce seizure frequency by at least 50%. Epilepsy with bilateral medial temporal lobe sclerosis responds better to ANT-DBS with a 75% reduction in seizure frequency compared to extratemporal lobe epilepsy or epilepsy that is difficult to pinpoint.
V. Complications and side effects
Complications and side effects of ANT-DBS are mostly mild or transient. Surgery and hardware-related complications include small hematomas in the frontal lobe, hardware-related skin breakdown and infections, and stimulation-related side effects include delusions, nystagmus, hallucinations, anorexia, and lethargy.
1. Central nucleus DBS
The use of the central nucleus as a target for the treatment of epilepsy is based solely on Penfield’s conjecture that the central nucleus may be involved in the pathophysiology of generalized epilepsy because it is an integrated structure of the subcortical superior system, receiving fibers from the brainstem and mesencephalon and projecting to the extensive cerebral cortex. The results of an open study showed an 80% reduction in seizure frequency and a significant improvement in quality of life in 13 patients with Lennox-Gastaut syndrome who underwent DBS of the central nucleus, but another double-blind crossover controlled study showed no significant efficacy of DBS of the central nucleus for seizures. Overall, central nucleus DBS may be effective in generalized epilepsy, but it remains unclear whether it is effective in partial epilepsy[.
2. STN-DBS
Animal experiments have demonstrated that drug or electrical stimulation to suppress STN suppresses seizures. A case study with a 30-month follow-up showed that bilateral STN-DBS reduced seizure frequency in partial epilepsy by 81%. STN-DBS not only reduced the frequency of seizures by 64.2%, but also reduced the severity of seizures.
3. Caudate nucleus DBS
The caudate nucleus loop exists between the caudate head, the thalamus and the neocortex. Activation of the caudate head is associated with hyperpolarization of cortical neurons, and this inhibition of the cortex may suppress epileptiform discharges. Studies have shown that low frequency stimulation (4-8 Hz) of the ventral aspect of the caudate nucleus head suppresses interictal epileptiform discharges, amygdala hippocampal discharges, and generalization of epileptiform discharges, but further clinical studies are lacking.
4.Electrocerebellar stimulation
The mechanism of cerebellar electrical stimulation may be as follows: activation of the Pakeno’s fibers inhibits the excitatory projections from the cerebellum to the thalamus, which reduces the excitatory projections from the thalamus to the cortex, and ultimately reduces the excitability of the cortex. Two other studies also showed good seizure control with EEG. However, a controlled double-blind study with 14 patients showed good results in only 2 patients. Therefore, the clinical efficacy of cerebellar stimulation needs to be confirmed by further studies.
5. Hippocampal electrical stimulation
The hippocampus plays a very important role in the development of temporal lobe epilepsy. In one study, after deep hippocampal electrodes in 10 patients with refractory temporal lobe epilepsy, short-term hippocampal stimulation resulted in a significant reduction in partial seizures and secondary generalized tonic clonic seizures in 7 patients, while long-term hippocampal stimulation resulted in the disappearance of seizures in 3 patients, with no effect on short-term memory. In a recent study including 7 patients with a 14-month follow-up, hippocampal electrical stimulation resulted in the disappearance of seizures in 1 patient, a reduction of more than 50% in 3 patients, a reduction of 25% in 2 patients, and no effect in 1 patient.
6.Electrical stimulation of the cerebral cortex
The stimulation mode used in DBS mentioned above is mainly open-loop stimulation, i.e., fixed pulses of electrical stimulation are generated according to the set parameters. The RNS consists of a pulse generator that can be embedded in the skull, electrodes with four contacts (cortical and deep electrodes), and a programmable controller. The RNS records and stores EEG information, which allows the physician to set the threshold for electrical stimulation and the stimulation parameters based on the EEG information. The following stimulation parameters can be set: frequency 1-333 Hz, current 1-12 mA, pulse width 40-1000 μs, and the amount of electricity generated by the stimulation is less than 25 μC/cm2, which is very low intensity and does not cause damage to brain tissue, and the risk of neural tissue damage from this intermittent stimulation is relatively low.
The study by Fountas et al. showed a mean follow-up of 9.2 months in 8 patients and a reduction in seizures of more than 45% in 7. Anderson et al. also showed that RNS reduced seizure frequency by 50%-75%, with no exacerbation of seizures or serious hardware-related complications, and that stimulation-related side effects were rare and disappeared with parameter adjustment. It should be noted, however, that the RNS battery depletes rapidly and needs to be replaced in 11-20 months.
VI. Summary
When patients with drug-refractory epilepsy are not suitable for epileptogenic focal resection, intracranial electrical stimulation may be considered for treatment. Although current studies have shown that intracranial electrical stimulation, especially ANT-DBS, STN-DBS, and hippocampal electrical stimulation, can control seizures, the effectiveness and safety of intracranial electrical stimulation for epilepsy needs to be further investigated due to the lack of prospective randomized controlled double-blind studies in large samples. In addition, the focus of future studies may include the types of epilepsy that respond well to intracranial electrical stimulation, the targets with the best efficacy, and the optimal stimulation parameters and modalities.