Clinical Studies of HS in Epilepsy

At least a dozen clinical studies support the hypothesis that HS has an antiseizure effect on MTLE. Three studies described a decrease in interictal epileptiform discharges in response to HS. Eight studies (n = 49) reported the effect of HS on seizures. In 4 (n = 38), continuous or cyclic HS abolished seizures entirely in 14 patients and improved seizures by 50% or more in 16, without any adverse effects. Two studies have reported on responsive HS in 3 patients (ie, seizure detection software triggers the delivery of HS). Seizures were aborted effectively in 1 patient, and improved by 58% in another. There was no effect in 1 patient with bilateral HS. In most studies, HS was performed on a short-term basis (<1 month) during preoperative intracranial EEG assessment. Patients then underwent resective surgery. The resected tissue showed no evidence of electrical injury. One study blinded stimulation status for the first month only, and then applied stimulation to all patients and assessed outcomes at 18 months. They found greater than 95% seizure reduction in 5 patients with normal MRIs, whereas 4 patients with hippocampal sclerosis achieved only 50% to 70% seizure reduction. There were no deleterious neuropsychological effects.

Studies of HS for epilepsy vary substantially in surgical technique and stimulation parameters. Some studies insert a single 4-contact electrode along the hippocampal longitudinal axis through a parieto-occipital burr hole ( Fig. 1 ). Others use 2 quadripolar electrodes, implanted through 2 occipital burr holes. One electrode is placed in the amygdala and the other in the anterior part of the hippocampus on each side ( Fig. 2 ). The choice of contacts to stimulate also varies. Some studies stimulate all contacts in the electrodes, other choose specific electrodes based on a certain number of epileptiform discharges. Stimulation parameters vary from continuous to cycling, with various durations of On and Off periods. Most studies use 90-microsecond pulses in trains of 60 to 200 Hz, applied at constant current (5 mA) or constant voltage (1–10 V), or at subthreshold amplitudes for side effects (usually sensory or visual phenomena).

T1-weighted MRI images showing HS surgical planning ( A ) and implantation ( B ), using a single quadripolar electrode along the longitudinal axis of the left hippocampus inserted through a parieto-occipital burr hole. L, left.

MRI showing susceptibility artifacts of ( A ) left amygdalar electrode in a sagittal plane and ( B ) left amygdalar and hippocampal electrode and right amygdalar electrode in a coronal plane.

( Reproduced from Boon P, Vonck K, De Herdt V, et al. Deep brain stimulation in patients with refractory temporal lobe epilepsy Epilepsia 2007;48(8):1551–60; with permission.)

This article highlights specific HS studies in epilepsy, focusing on the main findings and commenting on some strengths and weaknesses of the studies.

The initial study exploring HS in patients with MTLE was performed by Velasco and colleagues and included 10 patients who were implanted bilaterally and observed for 16 days before undergoing temporal lobectomy. In 7 patients (70%), there was complete cessation of complex partial and secondarily generalized tonic clonic seizures, and cessation of interictal spikes by the sixth day. Three patients had improvement in seizures and spikes. The most evident and fast antiepileptic responses were found in 5 patients whose stimulation contacts were located at either the pes hippocampus close to the amygdaloid nuclei or at the anterior parahippocampal gyrus close to the entorhinal cortex, or the anterior perforate space. Histopathologic analysis of the stimulated areas revealed no changes related to the HS. Velasco and colleagues concluded that HS is a safe and effective procedure. The proposed mechanism was the activation of the perforant pathway through a polysynaptic inhibitory influence on epileptogenic neurons in CA1 to CA4, and a change in benzodiazepine and opioid receptors produced by HS.

This initial study supporting the use of HS in patients with temporal lobe epilepsy showed that short-term HS is safe and exerted a remarkable beneficial effect on seizures. Weaknesses of the study include a small sample size, a short period of observation after implantation, and the lack of controls, randomization, or blinding.

The second study supporting the use of HS in temporal lobe epilepsy was performed by Vonck and colleagues. This study analyzed 3 patients with normal MRI but electroclinical findings consistent with mesial temporal epilepsy. Two quadripolar deep brain stimulation electrodes were implanted in each hemisphere trough 2 occipital burr holes. All patients had unilateral stimulation of specific electrodes selected by a given number of epileptiform discharges. At 7 days of follow-up, 2 of the 3 patients had fewer seizures, and all had a reduction in the number of interictal spikes. After a mean follow-up of 5 months, all patients had a greater than 50% reduction in seizure frequency, and none experienced side effects. The investigators suggested that long-term HS in the ictal zone reduces spikes and seizures, but commented that implantation alone could produce improvement in seizure frequency.

This study showed a potential benefit and safety of hippocampal HS after a few months of treatment. Weaknesses include a small sample size, and the lack of controls, randomization, or blinding. Participants were selected from a large initial pool of patients with intractable temporal lobe epilepsy, suggesting that HS may be applicable only to a small, select group of patients.

The same group studied 12 patients with refractory mesial temporal epilepsy who required intracranial EEG recording because of normal MRI or incongruent results of the noninvasive presurgical evaluation. Ten had seizure outcomes assessed at 12 to 52 months (mean 31 months). One of 10 patients had a greater than 90% reduction in seizure frequency, 5 had a seizure frequency reduction of 50% or more, 2 had a seizure frequency reduction of 30% to 49%, and 1 did not improve. One patient (8%) had asymptomatic hemorrhages along the trajectory of the depth electrodes. Two of the 12 patients with the poorest response to HS subsequently had resective surgery and were rendered seizure free. In most patients, the number of AEDs was lower at the end of follow-up. The investigators suggest that the mechanism of HS is related to the hypothesis of a reversible functional lesion. In this theory, local inhibition affects the areas that are involved in triggering, sustaining, and propagating epileptic activity.

As in the previous study by this group, HS was safe and had a beneficial effect after few months of treatment. However, the study also indicates that observation of larger numbers of patients can identify complications that are not apparent with limited sample sizes. The weaknesses of the study include small sample size, and the lack of controls, randomization, and blinding. Two patients who initially were considered for HS eventually had a temporal lobectomy and were rendered seizure free.

Velasco and colleagues performed a second study in 9 patients to evaluate the safety and efficacy of HS. All patients had intracranial EEG recordings because of potential bitemporal seizure onset. During the first month of the study, 5 patients were randomly assigned to the stimulator being turned off, and 4 patients to the stimulator turned on. The stimulator was turned on in all patients after 1 month, and the main analysis was done at 18 to 84 months (average 37 months). Three patients had skin erosion and a local infection after 24 months’ implantation. Five patients had normal MRIs and a seizure reduction of greater than 95%. Four patients had hippocampal sclerosis on MRI and their seizure reduction was only 50% to 70%. Neuropsychological testing before and after HS showed no changes. Patients randomized to 1-month stimulator turned on had both an immediate (1-month) and a late marked improvement in seizures, whereas those randomized to 1-month stimulator turned off had only late improvement in seizures and these were not as homogeneously favorable as those of the patients whose stimulator was on during the first month. The investigators concluded that HS is an alternative therapy for patients with mesial temporal epilepsy with normal MRI who are not candidates for epilepsy surgery.

There are several strong aspects of this study, including the longer follow-up, before-and-after neuropsychological assessment, and the analysis based on MRI findings. The brief, single randomized period and the small sample have negligible power and contribute little to the analysis. The investigators report that delaying stimulation by 1 month (stimulator off group) results in less favorable long-term seizure outcomes. This finding can hardly be attributed to the intervention and is more likely an artifact of the small sample size and perhaps inadequate long-term blinding. The investigators indicate that HS is useful where surgery is not indicated.

In a truly double-blind, randomized controlled trial, Tellez-Zenteno and colleagues studied 4 patients with unilateral MTLE and hippocampal sclerosis who were not candidates for resective surgery because of substantial risk of memory loss. One quadripolar electrode was inserted along the longitudinal axis of the hippocampus through a parieto-occipital burr hole. A within-patient, double-blind, multiple crossover, randomized controlled design (N-of-1 randomized trial) was used, with paired randomization of On and Off treatment sequences in each patient. The On and Off treatment periods lasted 1 month each, and each patient completed 3 pairs of treatment periods (6 treatment periods of 1-month duration each). Outcomes were assessed at monthly intervals in a double-blind manner, using standardized instruments and accounting for a washout period. The comparison between On and Off periods showed a median percent reduction in seizure frequency of 15% during the On periods. Comparison with seizure frequency at baseline showed a median reduction of 26% when the stimulator was on, and an increase of 49% when the stimulator was off ( Fig. 3 ). The absolute difference from baseline between On and Off groups was 75% favoring HS. One patient had no improvement in seizure frequency with HS. Neuropsychological function was comprehensively assessed at the end of each treatment period (6 times per patient), using alternate forms of the Boston Naming Test, Digit Span Test, Hopkins Verbal Learning Test, Brief Visual Memory Test, and Memory Assessment Clinic Self-Rating Scale. No differences were found between On and Off periods, or before and after HS in any of the tests. One patient who had a previous failed right temporal lobectomy improved dramatically with contralateral HS and showed no change in any of the cognitive tests. There were no differences between On and Off periods in outcomes such as depression, quality of life, seizure severity, impact of epilepsy, and individual symptoms. The investigators concluded that the seizure control with HS was substantially more modest than in all previous studies, likely related to the true randomized, blinded nature of the study, compared with largely open-label, nonrandomized comparisons in other studies showing larger improvements. The short duration of treatment periods could also have contributed, and an implantation effect could not be excluded. A large-scale, double-blind, parallel-group, randomized controlled trial to adequately assess the effect of HS was recommended.

Median percent change in seizure frequency in a multiple crossover, double-blind, randomized controlled trial by Tellez-Zenteno and colleagues. The differences between On and Off periods were statistically significant.

Although the sample size was small, the within-patient multiple randomized crossovers allowed for meaningful comparisons. The blinded, randomized comparison is the main strength of this study, including an analysis between On and Off periods. The 2 main weaknesses are the short treatment periods and the crossover design. HS periods of only 1 month may not be sufficient to obtain an effect. The crossover design is problematic because it is impossible to know with certainty whether the basic tenet to perform crossover studies is fulfilled in HS, that is, that all parameters return to baseline between treatment periods. There could be a cumulative effect of HS that continues after HS is turned off. This effect would decrease the difference between On and Off periods, spuriously decreasing the true effect of HS.

McLachlan conducted a blinded, single crossover, randomized trial in 2 patients using bitemporal HS. The 2 patients were poor candidates for resective surgery because of the presence of independent bitemporal originating seizures. After a 3-month baseline period, patients were randomized to a 3-month period of HS on or off, followed by a 3-month washout period and a crossover 3 months to the opposite treatment (on or off). During the On period, the mean monthly seizure frequency decreased by 33% compared with the Off period. During the washout period, seizure frequency was still 25% lower despite no ongoing HS, suggesting an enduring effect of HS, and confirming that crossover designs are problematic in HS. Visuospatial memory during the On period worsened in 1 patient (decline from 21st to 1st percentile) and improved in the other (increase from 8th to 34th percentile), but there was substantial variation in this test during washout and Off study periods as well. Verbal and subjective memory did not change. No serious adverse effects were found in the 2 investigated patients. This study could not reproduce the impressive results reported by Velasco and Vonck. The patient with the better response had a normal MRI, suggesting that HS could be more useful in nonlesional patients, as suggested by Velasco.

The study limitation of small sample size is self-evident. Strengths include the duration of the treatment period and the 3-month washout period, which suggests a persistent effect of HS, lasting several months after it is turned off. The study also suggests that implantation of the electrode itself did not decrease seizure frequency, which is important in the design of future studies (ie, On-Off comparisons are sufficient and no medical arm without implantation is needed). As in the study of Tellez-Zenteno and colleagues, the seizure effect was less impressive than in unblinded, nonrandomized studies. The changes in neuropsychological function are of unknown relevance.

Ongoing Randomized Trials of HS

Two ongoing randomized controlled trials of HS are registered in Clinicaltrials.gov , the trials registry and results database of the National Institutes of Health.

One is a Canadian multicenter study led by Wiebe and colleagues, exploring the effect of HS on disabling seizures, with a secondary analysis of cognitive outcomes, quality of life, mood, and surgical complications. The design consists of a 3-month baseline period followed by a parallel-group, double-blind, randomized trial of on versus off HS for 7 months in patients with MTLE who may be candidates for resective surgery but whose memory function precludes surgery. The eligibility criteria for this study include patients with unilateral or bilateral mesial temporal epilepsy, 18 years of age or older, global intelligence quotient (IQ) greater than 70, failure of adequate trials of 2 or more AEDs, ability to self-complete questionnaires, and patients’ preference for nonresective surgery or not being a candidate for mesial temporal resection. The study is in progress and no results have been released.

The other ongoing study, led by Boon and colleagues in Belgium, is a single-blinded, parallel-group, 3-arm, randomized controlled trial comparing traditional anteromesial temporal resection, hippocampal electrode implantation with HS on, and hippocampal electrode implantation with HS off for the first 6 months. The primary outcome is mean seizure reduction, whereas secondary outcomes include neuropsychological function, quality of life, mood, and safety. Eligibility criteria include pharmacoresistant MTLE shown by video-EEG, age greater than 18 years, IQ greater than 80, ability to adequately report seizures, and MRI suggestive of hippocampal sclerosis.

In summary, the available evidence for HS is weak ( Table 1 ). Many of the studies are small and of short duration, without controls or blinded outcome assessment. There seems to be a discrepancy in results from nonrandomized and randomized blinded studies, the latter showing significantly poorer effects of HS. The limited evidence suggests that the effects of HS seem to be cumulative and accrue with time, and the procedure seems to be safe. Neither the optimum target (amygdala, hippocampus, pes hippocampus, parahippocampal gyrus) nor the optimum stimulation parameters are known. However, variable improvement in seizures is reported in all studies of HS regardless of technique, and in virtually all studies of brain electrical stimulation for seizures. Subgroups of patients with MTLE who may benefit most from HS have not been identified, but there is a suggestion that patients with normal MRI may benefit more than those with mesial temporal sclerosis.

Table 1

Studies exploring hippocampal stimulation in MTLE

Study	Sample Size	Stimulation Parameters	Follow-up	Controlled Study/Randomization	Outcome	Study Limitations
Velasco et al, 2000	10	Pulse width: 450 microseconds Frequency: 130 Hz Output: 200–400 mA (subthreshold) Electrodes: monopolar, all electrodes Cycle: 23 h of every 24 h	16 d, before temporal lobectomy	None/none	In 70% of patients, seizure were abolished in the period of observation	Small sample size; short period of observation; lack of controls, randomization, or blinding
Vonck et al, 2002	3	Pulse width: 450 microseconds Frequency: 130–200 Hz Patients 1 and 2 responded to the initial stimulation Output: 1–3 V Electrodes: bipolar, selected by number of spikes Cycle: continuous	3–6 mo	None/none	Seizure reduction: 50% in all patients	Small sample size; lack of controls, randomization, and blinding
Boon et al, 2007	12	Pulse width: 450 microseconds Frequency: 130–200 Hz Patients 1 and 2 responded to the initial stimulation Output: 2–3 V Electrodes: bipolar, selected by number of spikes Cycle: continuous	12–52 mo, mean 31 mo	None/none	Seizure reduction: >90% in 1 patient ≥50% in 5 patients 30%–49% in 2 patients 0% in 1 patient	Small sample size; lack of controls, randomization, and blinding
Velasco et al, 2007	9	Pulse width: 450 microseconds Frequency: 130 Hz Output: 200–400 mA (subthreshold) Electrodes: monopolar, all electrodes Cycle: intermittent 1 min On, 4 min Off	18–84 mo, mean 37 mo	Yes/yes On vs Off during first month of the study, then analyzed as single group	Seizure reduction: 95% if normal MRI 50%–70% if MTS on MRI	Small sample size, brief and single randomized period, which precludes strong conclusions
Tellez-Zenteno et al, 2006	4	Pulse width: 90 microseconds Frequency: 190 Hz Output: 1.8–4.5 V (subthreshold) Electrodes: monopolar, all electrodes Cycle: continuous	6 mo	Yes/yes Double blind, On vs Off every month for 6 mo with multiple crossover trials in each patient	Seizure reduction: 15% On vs Off 26% On vs baseline Seizure increase: 49% Off vs baseline	Small sample size, short treatment periods, and crossover design
McLachlan, 2010	2	Pulse width: 90 microseconds Frequency: 185 Hz Output: 1.8–4.5 V (subthreshold) Electrodes: monopolar, all electrodes Cycle: continuous	9 mo	Yes/yes Double blind, On vs Off single crossover trial with washout period	Seizure reduction: 33% On vs Off 25% during washout	Small sample size

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Electrical Stimulation for Epilepsy: Experimental Approaches Thalamic Stimulation for Epilepsy Responsive Neurostimulation for the Treatment of Epilepsy Responsive Neurostimulation Suppresses Synchronized Cortical Rhythms in Patients with Epilepsy Focal Cooling Devices for the Surgical Treatment of Epilepsy Implanted Subdural Electrodes: Safety Issues and Complication Avoidance

Stay updated, free articles. Join our Telegram channel

Join