Deep Brain Stimulation Works for Drug-Resistant Epilepsy, But How?

Commentary

High-frequency stimulation of anterior nucleus of thalamus desynchronizes epileptic network in humans.

Yu T, Wang X, Li Y, Zhang G, Worrell G, Chauvel P, Ni D, Qiao L, Liu C, Li L, Ren L, Wang Y. Brain 2018;141;2631–2643.

Epilepsy has been classically seen as a brain disorder resulting from abnormally enhanced neuronal excitability and synchronization. Although it has been described since antiquity, there are still significant challenges achieving the therapeutic goal of seizure freedom. Deep brain stimulation of the anterior nucleus of the thalamus has emerged as a promising therapy for focal drug-resistant epilepsy; the basic mechanism of action, however, remains unclear. Here, we show that desynchronization is a potential mechanism of deep brain stimulation of the anterior nucleus of the thalamus by studying local field potentials recordings from the cortex during high-frequency stimulation (130 Hz) of the anterior nucleus of the thalamus in nine patients with drug-resistant focal epilepsy. We demonstrate that high-frequency stimulation applied to the anterior nucleus of the thalamus desynchronizes ipsilateral hippocampal background electrical activity over a broad frequency range, and reduces pathological epileptic discharges including interictal spikes and high-frequency oscillations. Furthermore, high-frequency stimulation of the anterior nucleus of the thalamus is capable of decoupling large-scale neural activity involving the hippocampus and distributed cortical areas. We found that stimulation frequencies ranging from 15 to 45 Hz were associated with synchronization of hippocampal local field potentials, whereas higher frequencies (> 45 Hz) promoted desynchronization of ipsilateral hippocampal activity. Moreover, reciprocal effective connectivity between the anterior nucleus of the thalamus and the hippocampus was demonstrated by hippocampal-thalamic evoked potentials and thalamic-hippocampal evoked potentials. In summary, high-frequency stimulation of the anterior nucleus of the thalamus is shown to desynchronize focal and large-scale epileptic networks, and here is proposed as the mechanism for reducing seizure generation and propagation. Our data also demonstrate position-specific correlation between deep brain stimulation applied to the anterior nucleus of the thalamus and patients with temporal lobe epilepsy and seizure onset zone within the Papaz circuit or limbic system. Our observation may prove useful for guiding electrode implantation to increase clinical efficacy.

Earlier this year the U.S. FDA approved deep brain stimulation (DBS) of the anterior nucleus of the thalamus as a therapeutic option for adult patients with medically refractory focal-onset epilepsy. The pivotal randomized controlled trial of 110 enrolled patients, cleverly acronymized as SANTE and originally published in 2010, showed that by approximately 2 years after surgical implantation, DBS was associated with a median seizure frequency reduction of more than 50% and a significant improvement in quality of life (1). Long-term follow-up data over subsequent years have shown that most patients experience further gradual improvement, with the average reduction in seizures reaching nearly 70% by 5 years (2). Implantation-related complications have not been trivial, however, including site infections in 12% of patients, electrode malpositioning in 8%, and intracerebral hemorrhage in 5%. Accumulated data from this cohort and from other case series have led some investigators to suggest that there may be multiple components to the therapeutic improvement from DBS—an immediate insertional effect, an immediate stimulation effect, and a delayed stimulation effect (3).

But how does DBS of the anterior nucleus of the thalamus work for epilepsy? There has long been evidence from animal models suggesting that the anterior nucleus plays a role in epileptic networks and that lesioning or electrically stimulating this nucleus can reduce epileptiform activity on EEG or raise the threshold required to generate experimentally induced seizures (4). Medical school neuroanatomy lessons from our past remind us, of course, that the anterior nucleus is part of the Papez circuit and has close connections to the mesial temporal lobe and other limbic system structures. Indeed, patients whose seizures originate in the temporal lobe have shown greater improvement from DBS than those with extratemporal onset.

In the article by Yu et al., nine patients with medically refractory focal epilepsy underwent intracranial stereo-EEG recording for the clinical purpose of presurgical localization of the epileptogenic zone. For research purposes, one of the depth electrodes was extended slightly into the anterior nucleus of the thalamus, and a number of experiments were then performed, testing the effect of stimulating the anterior nucleus at a range of different frequencies on local field potentials recorded from the other implanted electrodes.

Several important findings emerge from this work:

1. High-frequency stimulation of the anterior nucleus desynchronizes the background activity from the ipsilateral hippocampus (but not from extrahippocampal seizure-onset zones), and stimulation of other thalamic nuclei fails to do so;

2. High-frequency stimulation of the anterior nucleus reduces interictal epileptiform discharges and high-frequency oscillations from the ipsilateral hippocampus (but not from neocortical foci), and decouples larger neural network activity according to a measure of global density;

3. Low-frequency stimulation of the anterior nucleus seems to synchronize (rather than desynchronize) hippocampal activity;

4. Effective connectivity between the anterior nucleus and the hippocampus can be demonstrated through evoked potentials recorded during stimulation.

Taken together, these results expand our understanding of the mechanism of action of anterior nucleus DBS in focal epilepsy and may help to explain why temporal lobe epilepsy is particularly amenable to DBS therapy. They nicely support the growing appreciation of focal epilepsy as a disorder of widespread brain networks rather than just a single pathologic source, and add to the body of translational and clinical research that shows how these abnormal networks can be modulated, invasively or noninvasively, for excellent therapeutic effect, using a variety of different devices and stimulation paradigms.

As treating clinicians, we already know DBS “works” for epilepsy, but further studies, such as this one, will allow us to optimize its usage by identifying those who are likely to be helped most and indicating targets and parameters that are likely to be most efficacious. In the bigger picture, exactly when anterior nucleus DBS should be considered in patients with medically uncontrolled seizures is still yet to be determined, but it will surely take its rightful place among multiple options for those in whom resection of a single seizure focus is not a possibility.