SUDEP - brain stimulation
[Part 2: Questions; Rugg-Gunn F]
Brain stimulation therapy was first evaluated in epilepsy over 70 years ago but careful trials subsequently yielded mixed results and a high complication rate, and interest in this technique waned for several decades. More recently, following a refinement in surgical technique and technological advances in stimulation technology this approach was re-evaluated. Numerous brain stimulation targets have been assessed, including the cerebellum, hippocampus, subthalamic nucleus, caudate nucleus, centromedian nucleus, anterior nucleus of the thalamus and cortical targets. The anterior thalamic nucleus represents an attractive target due to its widespread connections to other brain regions and this has been the subject of the SANTE (Stimulation of Anterior Nucleus of Thalamus for Epilepsy) trial published in 2010 (Fisher et al., 2010). A total of 110 patients underwent electrode implantations into the left and right thalami, and continuous electrical stimulation was introduced. At the end of the three-month evaluation period a 40.4% decrease in median seizure frequency was seen in the stimulated group compared with a 14.5% decrease in the control, no-stimulation, group. The effect was particularly evident in patients with temporal lobe epilepsy. During long-term follow-up, there was a 41% decrease in median seizure frequency at 13 months, and 56% decrease at 25 months. Fourteen patients were seizure free for at least six months during the entire study. Nine patients had an increase in seizure frequency at 25 months. The most common side effects were tingling, pins and needles type sensations, reported in 18.2% of subjects. In addition, depression and memory impairment occurred more frequently in the stimulation group, although most were transient and resolved during long-term follow-up.
In the SANTE trial the thalami were stimulated continuously, whether epileptic activity was present or not. An important development recently is the introduction of responsive stimulation techniques where stimulation is provided to areas of the brain cortex where seizures are thought to arise and only when epileptic discharges are detected. Electrocorticographic activity is continuously monitored and recorded by implanted electrodes. When abnormal, epileptic, activity is detected, electrical stimulation is delivered, disrupting the activity and stopping the seizure. The results of a large, multicentre trial for patients with refractory focal seizures using this technique, called Neuropace, were published in 2011 (Morrell, 2011). In 191 adults with focal epilepsy, electrodes were implanted at one or two pre-specified cortical foci, from which seizures were considered to arise. During the first month post implantation, there was an improvement in seizure frequency in both the treatment and control, no stimulation, groups. This was similar to the effect seen in the SANTE trial and which has been attributed to an “implantation effect”. It is possible that this is due, in part, to very small lesions at the site of the implantation or a placebo effect. Subsequently however, over the evaluation period, there was a clear response to stimulation with a 37.9% reduction in seizure frequency in the treatment group compared with a 17.3% reduction in the control group. 102 patients were followed up for at least two years subsequently and 46% showed a significant reduction in seizure frequency. Side effects included implant site pain and headache. Generalized tonic-clonic seizures were seen to increase in frequency in 4.7% of patients during the first year of implantation. The rate of serious adverse events over the first 28 days was 12%, which is similar to the rate for the implantation of intracranial electrodes for seizure localization and epilepsy surgery. Over a longer term, the serious adverse event rate was 18.3%, which is lower than the 36% rate seen after implantation and treatment of Parkinson’s disease with deep brain stimulation. Intracranial bleeding was seen in 9/191 subjects (4.7%). Six subjects died during the study. One from lymphoma, one subject committed suicide, and four deaths were attributed to SUDEP, three of whom were in the stimulation group. The four SUDEP deaths over the 340 patient-years (11.8/1000 patient-years) was considered to be within the expected range for this population. Taken together, the SANTE and Neuropace trials show that direct electrical stimulation of brain structures, and specifically, continuous stimulation of the thalami and responsive stimulation of cortical sites can improve seizure control but with a small but definite risk of potentially harmful and long standing adverse effects such as intracranial bleeding.
The potential impact of intracranial stimulation systems on SUDEP rates is uncertain. There is a clear reduction in seizure frequency with stimulation in some patients but the specific data on convulsive seizures which correlates with the risk of SUDEP is not yet available. The seizure freedom rate, a key factor with respect to SUDEP, was low, at 7.1%, as might be expected in this refractory population. An increase in convulsive seizures was seen in less than 5% of patients but there was no overall increase in SUDEP rate. Further information regarding this may be derived from data collected from larger cohorts of patients in the next phase of evaluation and with more widespread use. Theoretically, it may be possible to stimulate regions of the brain critically involved with the control of heart rate and breathing in response to heart rhythm or breathing difficulties during seizures. Additionally, whether responsive cortical stimulation may be programmed and trained to behave as a brain “pacemaker” and thus prevent or treat EEG suppression after seizures, another proposed mechanism of SUDEP, is highly speculative and a number of key issues remain unresolved. These include the accurate and reliable detection of the absence of electrical activity by the device, the focal nature of the stimulation to treat a process affecting the whole brain, the uncertainty regarding the relationship of EEG suppression after seizures with SUDEP and the possibility that a stimulus may exacerbate the progressive electrical suppression.
In summary, brain stimulation has been shown to improve seizure frequency in patients with focal epilepsy, and there is the potential to act as an electrocortical “pacemaker” but the role of stimulation of this kind in preventing SUDEP is far from established. Further study into peri-ictal physiology and SUDEP mechanisms is necessary to provide a better understanding of which interventions should be recommended and which patients may be most likely to benefit.
Dept of Clinical and Experimental Epilepsy
National Hospital for Neurology and Neurosurgery, London, UK
How to cite:
Rugg-Gunn F. SUDEP - brain stimulation. In: Hanna J, Panelli R, Jeffs T, Chapman D, editors. Continuing the global conversation [online]. SUDEP Action, SUDEP Aware & Epilepsy Australia; 2014 [retrieved day/month/year]. Available from: www.sudepglobalconversation.com.